I I 1 1 Mai\yland Weather Service MARYLAND WEATHER SERVICE VOLUME TWO MARYLAND WEATHER SERVICE VOLUME TWO BALTIMORE THE JOHNS HOPKINS PRESS 1907 Z^i Boxi (§aitimovt (prcee BALTIMORE, MD., V. S. A. BOARD OF CONTROL W.A[. BULLOCK CLAKK, Director. REPRESEM'INC; THE JOII>"S HOPKIXS UMVKKSITY. W. T. L. TALIAFERRO, . . Secretary axd Treasurer. REPRESENTING THE MARYLAND AGRICULTURAL COLLEGE. OLREK LAXARt) FA.SSIG,, .... Meteohologist. REPRESENTING THE U. S. WEATHER BUREAU. TJie Maryland Weather Service is conducted under the joint auspices of the institutions above mentioned, the Central OflBce being located at the Johns Hopkins University. The meteorological work is under the immediate supervision of the Meteorologist who is detailed by the Chief of the U. S. Weather Bureau. Other lines of investigation are carried on in co-operation with various State and National organizations. LETTER OF TRANSMITTAL To His Excellenc}', Edwin Warfield, Governor of Maryland, Sir: — I have the honor to present herewith the second volume of the new series of reports of the Maryland Weather Service. The first volume contained a general account of the physiography and meteorology of the State while the present volume is chiefly devoted to a special study of the climatic features of Baltimore and vicinity. I am, Very respectfully, Wm. Bullock Clark, Director. Johns Hopkins Univebsitt, December 1, 1907. CONTENTS ^ PAGE PREFACE 17 INTRODUCTION, OPERATIONS OF THE SERVICE. By Wm. Bullock Clark 21 Physiography and Climate of the State 21 Climate and Weather of Baltimore 22 Climate of the Counties 22 Distribution of Plant Life in the State 23 SL'B\^y of the Swa:mp Lands of the State 23 Other Lines of Work 25 THE CLIMATE AND WEATHER OF BALTIMORE. By Oliver L. Fassig 27 THE CLIMATE OF BALTIMORE 29 Introduction 29 The Geographic Horizon of Baltimore 30 Atmospheric Pressure 31 The Diurnal Variations of the Barometer 34 The Normal Diurnal Variation at Baltimore 34 Phases of Diurnal Oscillation 35 Diurnal Variations of Pressure on Clear and Cloudy Days 40 The Diurnal Barometric Wave 41 Corrections for Reduction to true Mean Pressure 43 The Annual March of Atmospheric Pressure 44 Average Monthly and Annual Pressure 47 Annual and Secular Variations of Pressure 50 The Average Variability of Pressure 53 Extremes of Pressure Sfi Temperature of the Atmospheke 56 Introduction o6 Average Temperatures 57 The Normal Hourly Temperature 59 Phases of the Diurnal Variation 63 Diurnal Variation as Affected by Clouds and Rain 66 Effect of a Snow Covering 69 The Effect of Wind Velocity on Temperature 70 CONTENTS PAGE Range of Temperature on Calm and Windy Days 72 Reduction to the True Mean Temperature 72 The Hourly Rate of Change 73 Mean Daily Temperature 76 Average Inter-diurnal Changes of Temperature 79 Average Daily Range 83 Diurnal Variability of Temperature 83 The Probable Error of the Mean Daily Temperatures 90 Mean Monthly, Seasonal, and Annual Temperatures 91 The Normal Temperature 95 The Variability of the Monthly and Annual Mean 96 Warm Months and Seasons 97 Frequency of Stated Departures from the Monthly Seasonal and Annual Mean Temperatures 99 The Probable Error of the Monthly and Annual Means 101 Succession of the Seasons 103 Daily Extremes of Temperature 104 The Greatest Daily Range of Temperature 109 Monthly and Annual Extremes 112 The Greatest Monthly Range 114 Frequency of Days with Frost 115 The Frequency of Cold Waves 126 Killing Frdsts 129 The First and Last Occurrence of a Minimum of 32° 131 Light Frosts 135 The Period of Effective Temperatures for Plant Growth 136 The Frequency of Warm Days in Summer 137 Time of Occurrence of Annual Minimum and Maximum Temperatures 145 Temperature of the Water in the Harbor 147 Humidity 148 Introduction 148 Hourly Variation in Humidity 152 Phases of the Diurnal March of Relative Humidity 154 Mean Monthly and Annual Relative Humidity 156 Absolute Humidity 158 Mean Vapor Pressure 159 Pkecipitation 159 Introduction 159 The Causes of Precipitation 161 The Geographical Distribution of Rainfall 161 The Influence of Wind Direction 162 The Influence of Topography 162 The Influence of Atmospheric Pressure 163 The Seasonal Distribution of Rainfall 164 Hourly Amount of Rainfall 165 MARYLAXD WEATHER SEKYICE 7 PAGE Hourly Rainfall Frequency 167 Duration of Precipitation 170 Frequency of Precipitation of Stated Amounts 174 Average Daily Rainfall 178 Daily Rainfall Frequency 182 The Probability of Rain 183 The Monthly Precipitation 185 The Seasonal and Annual Precipitation 190 Monthly and Annual Departures 195 Excessive Rains 197 Greatest Rainfall in 24 Hours 199 Excessive Rates of Precipitation 205 Dry Spells 214 Wet Spells 219 The Distribution of Precipitation in Normal, Dry, and Wet Years. . . 223 Snowfall 227 Dates of First and Last Snow 231 The Frequency of Days with Snowfall 232 Heavy Snowfalls 235 Duration of Snowfall 236 Fogs 237 Sunshine and Cloudiness 239 Sunshine 239 Average Daily Sunshine 243 Sunshine Phases 244 Cloudiness 245 Clear, Partly Cloudy, and Cloudy Days 246 Frequency of Clear Days 248 Frequency of Partly Cloudy Days 250 Cloudy Days 251 The Winds 251 Introduction 251 Average Hourly Wind Movement 252 Average Daily and Total Monthly Wind Movement 255 Maximum Wind Velocities 258 Frequency and Duration of Stated Wind Velocities 261 Average Duration of Storm Winds 262 Gales 263 Prevailing Hourly Wind Directions 265 Prevailing Monthly and Annual Directions 268 Monthly Frequency of Stated Directions 273 The Direction of Upper and Lower Clouds 274 Electrical Phenomena 276 Thunderstorms 276 Thunderstorm Probability 280 8 CONTENTS PAGE Consecutive Days with Thunderstorms 280 Direction of Thunderstorms 281 Pressure Changes during Thunderstorms 282 Hail 284 Auroras 288 Sunspots and Weather 288 General Character of the Seasons 295 Observations and Instrumental Equipment 296 Historical Notes 296 Observers and Observations 301 Instrumental Equipment 301 Hours of Observation 302 Changes in the Location of the Station and Officials in Charge 304 Summary of Average and Extreme Values 306 THE WEATHER OF BALTIMORE 311 Introduction 311 The Synoptic Weather Chart 312 Cyclones and Anti-cyclones 313 Areas of Unsettled Weather (Cyclones) 316 Pressure and Winds 318 Temperature and Wind Direction 319 Distribution of Clouds and Precipitation 320 Areas of Fair Weather (Anti-cyclones) 321 Isobars and Winds 321 The Winds and Distribution of Temperature 323 Distribution of Clouds 324 The Eastward Drift of Cyclones and Anti-cyclones 324 Weather Charts of the Northern Hemisphere 327 Weather of the Principal Climatic Zones 328 The Tropical Zone 328 The Temperate Zones 329 The Polar Zones ; 330 The Seasons 331 AVinter Weather 333 Winter Cyclones 334 The Lake Storm 335 The Storm of December 24-26, 1902 335 The Storm of January 7-8, 1903 341 The Storm of February 27-March 1, 1903 345 The Southwest Storm 350 The Storm of February 3-5, 1903 350 The Storm of December 26-28, 1904 354 The Storm of December 12-13, 1903 359 MARTLAXD W EATHEll SEilVICE 9 PAGE The Gulf Storm 363 The Storm of February 1-3, 1902 364 The Storm of January 5-7, 1905 368 The Storm of February 20-22, 1902 373 The Blizzard 378 The Blizzard of March 11-13, 1888 378 The Blizzard of February 12-14, 1899 382 Areas of Fair Weather (Anti-cyclones) 389 Cold Waves 391 The Cold Wave of December 13-15, 1901 ^ 392 The Cold Wave of February 10-13, 1899 395 The Origin of Cold Waves 396 The Cold Winter of 1903-04 397 The Warm \\'inter of 1889-90 399 The Distribution of Atmospheric Pressure during the Cold Winter of 1903-04 and the Warm Winter of 1889-90 400 The Variability of Winter Weather 401 The Weather of Christmas Day 404 The AVeather of Washington's Birthday 409 Si'Ki.Nc; Wkathek 410 March Winds and Storms 412 Ice Storms 413 The Squall of March 1 , 1907 413 Equinoctial Storms 415 Hail Storms 417 The Storm of May 19, 1904 418 The Storm of April 27, 1890 418 Spring Frosts 421 Ice without Frost 423 Periods of Unsettled Weather 424 The Rainy Period of April 19-25, 1901 425 The Rainy Period of May 16-26, 1894 426 The Variability of Weather in Spring 428 The Weather of March 4 429 The Weather of May 1 432 The Weather of Easter Sundays 432 Sim .mi;k Wkathki! 436 Summer Storms 437 The Thunderstorm of July 20, 1902 438 Ttic 'j'hunderslorm of July 3, 1902 444 The Thunderstorm of July 12, 1904 446 The Tornado of July 12, 1903 447 * Waterspouts 452 Summer Hot Spells 453 The Summer of 1900 454 General Weather Conditions 459 10 CONTENTS PAGE The Summer of 1901 462 The Hot Periods of August, 1900, and July, 1901, Compared 463 Days with a Maximum Temperature of 90° or above 466 The Cold Summer of 1816 467 Distribution of Pressure during the Cool June of 1903 469 Distribution of Pressure during the Normal June of 1902 470 The Variability of Summer Temperatures 470 The Weather of July 4 472 West Indtan Hurricanes 475 Frequency of Hurricanes 476 The Hurricane of October 13, 1893 476 Autumn Weather 480 Indian Summer 482 The Variability of Autumn Temperatures 486 The Weather of September 12 486 The Weather of October 1 489 The Weather of Thanksgiving Day 489 The Heavy Rains of September 24-26, 1902 492 FOEETEtLING THE WEATHEB (HISTORICAL) 493 Introduction 493 Natural Signs 494 Astro-Meteorology 496 Symbolic Days 498 Early Books on Weather Proverbs 498 Forecasts Based on Average and Extreme Values 499 Temperature Variability 500 Rainfall Probability 500 Special Days 501 Recurring Periods 502 The Method of the Synoptic Weather Chart 504 The Indian Seasonal Forecasts 507 Index 511 ILLUSTRATIONS PLATE FACIXG PAGE I. The Diurnal Barometric Wave 42 II. Typical Barograms 44 III. Daily March of Temperature and Pressure 80 IV. Daily March of Temperature 82 V. Typical Thermograms 86 VI. Departures of Mean Monthly Temperature from Normal for 87 Years 87-88 VII. Departures of Mean Monthly, Seasonal, and Annual Tempera- ture from Normal for 87 Years 92 VIII. Selected Relative Humidity Curves 158 IX. Precipitation Probability 184 X. Monthly, Seasonal, and Annual Departures from the Normal Precipitation (1817-1904) 194 XI, Average Hourly Wind Direction 266 XII. Sunspots, Solar Prominences, and Weather Conditions 294 XIII. General Character of the Seasons. — Winter 296 XIV. General Character of the Seasons. — Spring 296 XV. General Character of the Seasons. — Summer 296 XVI. General Character of the Seasons. — Autumn 296 XVII. General Character of the Year 296 XVIII. Office of the U. S. Weather Bureau and Maryland State Weather Service 300 XIX. Storm Warning Display Station 304 XX. Hourly Observations at Baltimore during the Blizzard and Cold Wave of Feb. 9-14, 1899 388 XXI. Frost Figures 311 XXII. Effects of Ice Storm of March, 1906 413 XXIII. Depanures in Temperature during the Hot Spell of 1900 462 XXIV. Distribution of Pressure, Winds, and Temperature during Normal, Cold, and Warm Seasons in the United States 470 FIGURE ■ PAGE 1. Hourly Variations of the Barometer 33 2. Isopleths of Hourly Pressure 36 3. Principal Phases of Diurnal Oscillation of Pressure 38 4. Diurnal Variations of Pressure on Clear and on Cloudy Days 40 5. The Diurnal Barometric Wave 42 12 ILLCSTRATIOXS PAGE 6. Mean Monthly Atmospheric Pressure 44 7. Variations in the Mean Monthly Pressure 46 8. Annual Variations of Pressure Expressed as Departures from the Normal Value 50 9. Monthly Means and Extremes of Pressure 54 10. Mean Hourly Temperature 60 11. Isopleths of Hourly Temperature 62 12. Principal Phases of Diurnal Variation of Temperature 64 13. Effect of Cloudiness and Rain on the Hourly Variations of Tempera- ture 67 14. Effect of Snow-covering on the Hourly Variations of Temperature.. 68 15. Effect of Wind Velocity on the Hourly Variations of Temperature.. 71 16. Hourly Rate of Change of Temperature 74 17. Curves Representing the Average Hourly Pressure and the Hourly Rate of Change in Temperature for the Year 76 IS. Inter-diurnal Temperature Changes 80 19. Total Seasonal and Annual Frequency of Stated Diurnal Changes of Temperature 84 20. Diurnal Changes of Temperature of less than 6°, +6°, +8°, and -1-10° each month 85 21. Diurnal Changes of Temperature of —6°, -^6°, +8°, -M0°, -|-20°... 88 22. Frequency of Stated Departures from the Monthly Normal Tem- perature 101 23. Frequency of Stated Departures from the Normal Seasonal and An- nual Temperatures 102 24. Greatest Daily Range of Temperature 109 25. Extreme, Average, and Mean Maximum and Minimum Temperatures. Ill 26. Absolute Maximum and Minimum Temperatures 114 27. Greatest Monthly Range of Temperature 115 28. Longest Period of Consecutive Days with a Minimum Temperature of 32° or Below 119 29. Number of Days with Mean Temperature Below 14° and 32° 120 30. Annual Frequency of Days with a Maximum Temperature Below 32° 121 31. Annual Frequency of Cold Days 122 32. Monthly Frequency of Cold Days 122 33. Interval Between Last and First Occurrence of a Minimum Tem- perature of 32° 132 34. Interval Between Last and First Occurrence of Minimum Tempera- ture of 40° 134 ■ 35. Annual Number of Days with Mean Temperature Above 42° 136 36. Annual Number of Days with Maximum Temperature of 90° and Over 137 37. Time of Occurrence of the Lowest and Highest Temperature of the Year 144 38. Air and Water Temperatures in Baltimore Harbor 144 39. Mean Hourly Relative Humidity 151 :martlaxd weather service 13 PAGE 40. Mean Hourly Relative Humidity 153 41. Phases of the Diurnal Variations in Relative Humidity 155 42. The Mean Monthly Relative Humidity 156 43. Variations in the Mean Annual Relative Humidity 156 44. Average Hourly Precipitation 165 45. Average Hourly Amounts of Precipitation in January and July 167 46. Average Hourly Frequency of Precipitation in January and July. . . . 169 47. Average Hourly Frequency of Precipitation 170 48. The Average Duration of Precipitation 172 49. Variations in the Annual Frequency of Days with Appreciable Pre- cipitation 177 50. Monthly Frequency of Precipitation 182 51. The Monthly Amount of Precipitation 185 52. Mean Monthly Precipitation 188 53. Variations in the Annual Amount of Precipitation for 1817 to 1904. . 191 54a. Departures from Mean Monthly Precipitation (1817-1859) 193 54b. Departures from Mean Monthly Precipitation (1860-1904) 194 55. The Heaviest Precipitation in any 24 Consecutive Hours 201 56. Rainfalls Equalling or Exceeding 2.50 Inches in a Day 202 57. Rainfalls Equalling or Exceeding One Inch per Hour 203 58a. Excessive Rates of Rainfall 210 58b. Excessive Rates of Rainfall 211 59. Dry Periods 215 60. Dry Periods 218 61. Wet Periods 220 62. Total Monthly Precipitation During a Dry Year, a Normal Year, and a Wet Year 224 63. Daily Precipitation During a Dry Year, a Normal Year, and a Wet Year 225 64a. Annual Frequency of Days with a Snowfall to the Amount of One- tenth of an Inch 228 64b. Annual Depth of Snowfall in Inches 229 65. Monthly Frequency and Amount of Snowfall 233 66. Mean Hourly Sunshine 241 67. Mean Hourly Sunshine for the Year 242 68. Average Hourly Cloudiness 245 69. Relative Frequency of Clear. Partly Cloudy, and Cloudy Days 247 70. Hourly and Annual Variations of Wind Velocity 253 71. Average Hourly Variations in Wind Velocity 254 72. The Frequency of Storm Winds 258 73. Average Hourly Wind Direction facing page 266 74. Prevailing Morning and Afternoon Wind Directions in January. . . . 267 75. Relative Frequency of Prevailing Wind Directions 269 76. Prevailing Monthly Directions of the Wind in Warm, in Normal, and in Cold Seasons and Years 270 77. The Frequency and Distribution of Thunderstorms 277 78. The Average Monthly P'requency of Occurrence of Thunderstorms ., 277 14 ILLUSTRATIONS PAGE 79. The Annual Frequency of Occurrence of Thunderstorms from 1871 to 1904 278 80. The Direction of Movement of Thunderstorms 281 81. Some Typical Barograms during Thunderstorms and Squalls 283 82. The Frequency of Occurrence and the Hourly and Seasonal Distribu- tion of Hailstorms 286 83. Barograms during Hailstorms 286 84. Sunspots, Solar Prominences, and Weather Conditions 294 85. Typical Cyclone of Dec. 27, 1904 (Pressure and Winds) 317 86. Typical Cyclone of Dec. 27, 1904 (Complete Chart) 317 87. Typical Anticyclone of April 4, 1904 (Pressure and Winds) 322 88. T3T)ical Anticyclone of April 4, 1904 (Complete Chart) 322 89. Typical Cyclone and Anticyclone of March 3, 1904 325 90. Pressure Distribution over the Northern Hemisphere, Dec. 4, 1886 . . 327 91. The Lake Storm of Dec. 24, 1902 336 92. The Lake Storm of Dec. 25, 1902 336 93. The Lake Storm of Dec. 26, 1902 337 94. The Lake Storm of Dec. 24-26, 1902 (Diagr.) 339 95. The Lake Storm of Jan. 7, 1903 342 96. The Lake Storm of Jan. 8, 1903 342 97. The Lake Storm of Jan. 6-8, 1903 (Diagr.) 343 98. The Lake Storm of Feb. 27, 1903 346 99. The Lake Storm of Feb. 28, 1903 346 100. The Lake Storm of March 1, 1903 347 101. The Lake Storm of Feb. 27-March 1, 1903 (Diagr.) 348 102. The Southwest Storm of Feb. 3, 1903 351 103. The Southwest Storm of Feb. 4, 1903 352 104. The Southwest Storm of Feb. 5, 1903 352 105. The Southwest Storm of Feb. 3-6, 1903 (Diagr.) 353 106. The Southwest Storm of Dec. 26, 1904 356 107. The Southwest Storm of Dec. 27, 1904 356 108. The Southwest Storm of Dec. 28, 1904 357 109. The Southwest Storm of Dec. 26-28, 1904 (Diagr.) 358 110. The Southwest Storm of Dec. 12, 1903 360 111. The Southwest Storm of Dec. 13, 1903 360 112. The Southwest Storm of Dec. 12-14. 1903 (Diagr.) 361 113. Paths and Rain Areas of Southwest Storms of Jan., 1898 362 114. The Gulf Storm of Feb. 1, 1902 365 115. The Gulf Storm of Feb. 2, 1902 365 116. The Gulf Storm of Feb. 3, 1902 366 117. The Gulf Storm of Feb. 1-3, 1902 (Diagr.) 367 118. The Gulf Storm of Jan. 5, 1905 369 119. The Gulf Storm of Jan. 6, 1905 369 120. The Gulf Storm of Jan. 7, 1905 370 121. The Gulf Storm of Jan. 5-7, 1905 (Diagr.) 371 122. The Gulf Storm of Feb. 20, 1902 373 123. The Gulf Storm of Feb. 21, 1902 374 MARYLAXD WEATHER SERVICE 15 PAGE 124. The Gulf Storm of Feb. 22, 1902 374 125. The Gulf Storm of Feb. 20-22, 1902 (Diagr.) 375 126. Paths of the Gulf Storms of February, 1902 376 127. Diagram of Rainj^ Sundays of the Winter of 1801-2 (Diagr.) 377 128. The Blizzard of March 11, 18SS 379 129. The Blizzard of March 12, 1888 379 130. The Blizzard of March 13, 1888 380 131. The Blizzard of March 11-13, 1SS8 (Diagr.) 381 132. The Blizzard of Feb. 9, 1899 384 133. The Blizzard of Feb. 10, 1899 384 134. The Blizzard of Feb. 11, 1899 385 135. The Blizzard of Feb. 12. 1899 385 136. The Blizzard of Feb. 13, 1S99 386 137. The Blizzard of Feb. 14, 1899 386 138. Snow on the Ground after the Blizzard of February, 1899 387 139. Cold Wave of Dec. 13, 1901 393 140. Cold Wave of Dec. 14, 1901 393 141. Cold AVave of Dec. 15, 1901 394 142. Cold February 11, 1899 402 143. Warm February 11, 1887 402 144. The Weather of Christmas Day (December 25) (Diagr.) 406 145. The W^eather of Washington's Birthday (February 22) (Diagr.).. 408 146. The Squall of :\Iarch 1, 1907 414 147. The Hail Storm of May 19, 1904 419 148. The Hail Storm of May 19, 1904 ( Diagr. ) 419 149. The Frost of May 9, 1906 422 150. Ice without Frost, April 17, 1905 424 151. The Weather of March 4 (Diagr.) 430 152. The Weather of May 1 (Diagr.) 433 153. The Thunderstorm of July 20, 1902 440 154. The Thunderstorm of July 20, 1902 (Diagr.) 441 155. The Movements of the Thunderstorm of July 20, 1902 (Diagr. "i 442 156. The Thunderstorm of July 3, 1902 444 157. The Thunderstorm of July 3, 1902 (Diagr.) 445 158. The Thunderstorm of July 12, 1904 447 159. The Tornado of July 12, 1903 (8 a. m.) 448 160. The Tornado of July 12, 1903 (8 p. m.) 448 161. Chart of August 6, 1900 (during Hot Spell) 460 162. Temperature during Hot Spells of 1900 and 1901 (Diagr.) 464 163. The Cold July 1, 1885 471 164. The Warm July 1, 1901 471 165. The Weather of July 4 (Diagr.) 474 166. The Hurricane of Oct. 13, 1893 (8 a. m.) 478 167. The Hurricane of Oct. 13, 1893 (8 p. m.) 478 168. The Hurricane of Oct. 14, 1893 (8 a. m.) 479 169. The Weather of Oct. 29, 1903 (Indian Summer) 484 170. The Weather on September 12 (Defenders' Day) (Diagr.) 488 PREFACE The present volume is the second of a series of reports dealing with the climatic features of Maryland. The first volume was general in character and presented all that was then known regarding the physi- ography and meteorolog}- of the State. The present and succeeding volumes will be devoted to more special studies within the province of climatological research. The Introduction to the present volume, prepared by Wm. Bullock Clark, is devoted chiefly to an account of the operations of the Service together with the plans for future work. An account is given of the Swamp Lands of the State whicli are atti-aeting wide attention. The writer refers to the Botanical Survey of the State, which has been made under the auspices of the State Weather Service, the results of which will be shortly printed in Volume III of the present series. The Report on the Climate and WeatJier of Baltimore and Vicinity, discussed by Oliver L. Fassig, constitutes the chief portion of the vohune and represents the result of many years of exhaustive study of the Balti- more region. All of the available records both public and private have been employed in this work and the result may be regarded as remarkably complete. It is doubtful if any district luis received as thorough study as Dr. Fassig has given to that of Baltimore and vicinity. The report is divided into two parts. The first deals with the average and extreme values of the meteorological elements recorded in the city of Baltimore. The discussion is based upon careful ol)Servatious extending over a period of nearly a century. The second part deals with types of weather experi- enced in Baltimore and vicinity — hence with the actual physical condi- tion of the atmosphere at stated times, during the prevalence of storms, cold and warm waves, etc. 18 PREFACE The Maryland Weather Service desires especially to extend its thanks to Professor Willis L. Moore, Chief of the U. S. Weather Bureau, who has generously aided the conduct of the investigations discussed in the present volume. Dr. Fassig has had access to the complete records of the U. S. Weather Bureau as well as to those of other official organizations. Mr. E. W. Berry, of the State Geological Survey, has materially aided in editing the manuscripts for the volume. INTRODUCTION OPERATIONS OF THE SERVICE BY WM. BULLOCK CLARK INTRODUCTION OPERATIONS OF THE SERVICE BY WM. BULLOCK CLARK The Maryland Weather Service has been engaged for many years in a study of the climatic features of Maryland. These investigations have resulted in the accumulation of a vast amount of information relating to the meteorology, the physiography, the agricultural soils, and the distri- bution of plant life. Much aid has been rendered the State Weather Service in this work by both the National and State bureaus. Physiography and Climate of the State. The results of the physiographic and meteorological studies of the entire state down to 1899 were presented in Volume I of this series of reports. These investigations were based on all the then existing obser- vations and records, both official and private. The physiographic studies had been conducted largely under the auspices of the State Geological Survey, but as the results were so fundamental to an interpretation of the climatic features of the State, their publication in the very first volume of the new series of reports seems desirable. The meteorological data relating to Maryland climate had been ac- cumulated for over a century, but little attempt had been made hitherto to draw conclusions from them or to seek an explanation for the many variations that are found in the different sections of the state and in the T-i INTRODUCTION same regions at different seasons of the year. These studies may be regarded as preliminary to the more exhaustive investigations which have followed, as well as to those which still await completion. Climate and Weather of Baltimore. The investigations of the climate and weather of Baltimore and vicin- ity, discussed in the pages of the present volume, may be regarded as fully meeting the requirements of such a detailed study. The author has treated exhaustively the elements entering into the interpretation of the conditions found to prevail in the Baltimore region. It is probably the most complete study that has ever been given to the climate and weather of a single city and its environs, and will afford a most important store- house of information for those who may be seeking for an accurate knowl- edge of the exact conditions that prevail in Baltimore and its immediate surroundings. The aid rendered by the Chief of the U. S. Weather Bureau has alone made it possible to secure the results here recorded. Climate of the Counties. Special reports on the climate of Allegany, Garrett, Cecil, Calvert, and St. Mary's counties have been prepared by the Maryland Weather Service and issued under the auspices of the Maryland Geological Survey in its series of county reports. It is the intention of the State Weather Service ultimately to bring together and publish these chapters when complete for all the counties in a single volume of the State Weather Service s'eries. When brought out this report on the climate and weather of the Mar3dand counties will present for each political district of the State an exhaustive discussion of its special features that will be of great benefit to the inhabitants and to those seeking information regarding the special climatic conditions of any particular county. This study will take sev- eral years for its consummation, but with the co-operation so generously furnished by the Chief of the U. S. Weather Bureau it will be finally completed in a form that will be recognized as thoroughly authoritative. MARYLAND WEATHER SERVICE 23 The Meteorologist in charge of the State Weather Service, who has always been the representative of the U. S. Weather Bureau in Balti- more, has been hitherto designated by the Chief of the National Service to prepare these reports and has had access not only to the United States, but to the State records in this worlv. He lias been able to employ the services of a trained body of men who would be otherwise entirely beyond the reach of the State for such an investioation. Distribution of Plant Life in the State. The distribution of animal and plant life, and more especially of the latter, is so intimately associated with the physiographic and climatic conditions that prevail that the Maryland Weather Service has under- taken a Botanical Survey of the State as a part of its climatic studies. For the past three years several trained botanists under the direction of Dr. Forrest Shreve have been engaged in the different sections of the state in making a detailed investigation of the botanical conditions. Not only has the distribution of plant life been found to be dependent on the climate and physiography of the state, but upon the agricultural soils which in turn find their ultimate interpretation in the underlying rocks from which they have been derived, thus bringing the work of the State Geological Survey and State Weather Service into close association. The botanical survey is now completed and a report is at the present time being prepared, which will be issued as Volume III of the State Weather Service. Survey of the Swamp Lands of the State. A- survey of the swamp lands of Maryland has been made in connection with the topographic survey of the state, in which the State Weather Service has participated with the State Geological Survey in its co-opera- tion with the Topographic Branch of the TJ. S. Geological Survey. This survey has shown the following swamp areas. 24 IXTRODUCTION Area of Swamp Lands in the Various Counties Computed from the Maryland Geological Survey Maps. County. Fresh. Salt. Total. Sq. Mi. Acres. Sq. >li. Acres. Sq. Mi. Acres. Baltimore 1.7 1,088 5.4 3,456 7.1 4,544 Anne Arundel 3.3 2,112 1.9 1,216 5.2 3,328 Prince George's ... 8.6 5,504 0.2 128 8.8 5,632 Charles 11.9 7,616 22.1 14,144 34.0 21,760 Calvert 3.2 2,048 1.2 768 4.4 2,816 St. Mary's 0.3 192 1.3 832 1.6 1,024 Harford 0.4 256 11.3 7,232 11.7 7,488 Cecil 0.2 128 6.5 4,160 6.7 4,288 Kent 0.4 256 7.9 5,056 8.3 5,312 Queen Anne's 9.7 6,208 4.5 2,880 14.2 9,088 Talbot 0.3 192 5.3 3,392 5.6 3,584 Caroline 9.7 6,208 2.6 1,664 12.3 7,872 Dorchester 78.3 50,112 123.2 78.848 201.5 128.960 Wicomico 17.1 10,944 22.1 14.144 39.2 25,088 Somerset 7.7 4,928 68.5 43,480 76.2 48,768 Worcester 33.0 21,120 35.4 22,656 68.4 43,776 Garrett 4.5 2,880 4.5 2,880 Other counties 4.0 2,460 4.5 2,560 Total 194.3 124,352 319.4 204,416 513.7 328,768 it will thus be seen that the State of Maryland has 328,768 acres of swamp lands, of which 124,352 acres are fresh-water swamps and 204,416 acres salt-water marshes. The eastern and southern counties of the state bordering the Chesapeake Bay and the Atlantic Ocean have 323,326 acres, of which 118,912 acres are fresh and 204,416 acres are salt. The central and western counties have 5440 acres, all of which are fresh. The agi'icultural soil survey of Maryland, which is being carried on in co-operation with the TJ. S. Bureau of Soils, shows a considerably larger acreage of swamp lands in those counties surveyed than the estimates above given, but in the soil survey the small tracts on individual farms were computed, while the topographic maps show only the larger areas, which would alone be considered in any plan of government reclamation. Counting these small tracts the total area would probably reach 500,000 acres. A fuller study of these swamp lands is now in progress and as their present condition is intimately connected with the climatic conditions of MARYLAND WEATHER SERVICE 25 the State their study, in part at least, falls within the province of the State AVeather Service. Other Lixes oe Work. The far reaching influence of climate on the economic and social development of the state suggests other lines of investigation that require the attention of the State Weather Service. The character of the agricultural soils, although fundamentally deter- mined by the underlying rocks, is also to no small degree dependent on the physiography and climatic features of the State. These factors must be considered in any comprehensive study of the agricultural soils. The health of any community is also to no inconsiderable extent de- pendent on the climate, and this is recognized in the field of investigation known as medical climatology. In Volume I the present writer said in discussing this subject in his " Plan of Operation of the Service "' that " the healthfulness of Maryland as a place of residence is a question of no small importance to those who may be considering the advisability of seeking homes in our midst, and actual facts should be presented in such a manner as to command their attention. The various sections of the state, their marked differences in temperature and rainfall, may be shown to be adapted to the physical requirements of different people, and it is highly important that these facts should be made known. " It is also probable, as the meteorological records over considerable periods are carefully studied, that some districts will be found highly beneficial to people suffering from certain ailments. It is the purpose of the Maryland AYeatlier Service to have some expert upon medical clima- tology carefully study its records and prepare a report upon this subject, and already arrangements to this end have been perfected." The general and special studies earlier enumerated naturally afford the basis for a considei-ation of tlie crop condition? of tlie State and this subject should be taken up in a comprehensive way as the data collected become adequate to the discussion of so great a subject. The agricul- tural products of the State far surpass those in every other line, and tlie State Weatlier Service should give whatever assistance it can in the study of the important problems involved. 26 INTfiODUCTIOJSr It is also a well recognized fact that the character and distribution of forest growth is in no small degree determined by the climatological features which have already been described. Since forestry studies were organized by the State Geological Survey a few years ago a State Board of Forestry has been organized and the investigation of our forests is now well under way. The State Weather Service can aid in various ways in this work. The climate in its various relations touches human life in so many points that the investigations already undertaken and proposed will prove not only of great interest, but of greater value to the people of the state. Eesults of real worth can rarely be obtained quickly, but the investigations now being conducted by the State Weather Service are of a fundamental character, and when completed will cover as fully as possible the jfield of climatology in its various relations to the economic interests of the State. REPORT ON THE CLIMATE AND WEATHER OF BALTIMORE AND VICINITY (Based on the Observations of the U. S. Weather Bureau; Supplemented by Obser- vations of the Maryland State Weather Service, and the U. S. Army Medical Department.) PREPARED BY DIRECTION OF WILLIS L. MOORE V Chief oh U. S. Wkather Bureau BY OLIVER LANARD FASSIG THE CLIMATE OF BALTIMORE I^siTEODUCTIOX For more than thirty years the United States Weather Bureau has maintained a station of the first order in Baltimore City. During all these years the weather conditions have been carefully and accurately noted and recorded at several stated hours of the day by trained observers. In 1893 the instrumental equipment of the station was greatly increased and the value of the records enhanced by the acquisition of additional self-recording instruments by means of which a continuous record has been obtained of all the principal elements of the weather. The records of the Baltimore station now show the local state of the atmosphere dur- ing every hour of the day and night since 1893, barring an occasional brief break in the record due to accidental causes. The factors thus continuously noted are the temperature of the atmosphere, the pressure, rainfall, sunshine, wind velocity, wind direotion and, since 1902, the humidity. This mass of exceedingly valuable raw material for the study of problems in local climatology, supplemented by an almost unbroken series of local observations made since 1817 under the auspices of the United States Army Medical Department and the Smithsonian Insti- tution, has never before been subjected, as a whole, to a critical analysis and reduction. It is evident that such observations, secured at enormous expense of time and money, should yield benefits beyond their immediate uses at the time of recording, however valuable these may be. The weather conditions at Baltimore are typical of conditions within a wide area. Allowing for small differences in amplitude of variation due to local surface conditions, an analysis of the Baltimore observations may with safety be applied to much of that portion of the Middle Atlantic States lying east of the Appalachian Mountains. This area lies about midway between the rigorous north and the mild south, the equable ocean region anrl the region of great variability in the interior of the continent. 30 THE CLIMATE OF BALTIMORE Eainfall is abundant and quite uniformly distributed throughout the 3^ear. Storms of destructive violence are of rare occurrence; tornadoes are almost unknown. The season of safe plant growth is long, and sun- shine is abundant. In the following report the analysis of the observations is divided into two distinct parts. The first part deals with the average conditions of the atmosphere, derived from many years of statistical data relating to temperature, pressure, humidity, rainfall, clouds and sunshine, winds, etc., and to departures from their normal values. In brief, it deals with the climate of the region about Baltimore. The second part is devoted to the iveather, or actual conditions of the atmosphere at any given time as regards temperature, humidity, rainfall, clouds, wind — the sum total of the atmospheric conditions. Hence weather is a passing phase of climate. Attention will be directed largely to storms, cold waves, hot waves, etc., as well as to the gentler phases of the atmosphere Avhich con- stitute the daily routine of weather. This division into climate and weather is necessarily more or less arbitrary, and the lines of demarcation employed by different writers will seldom be foimd in exactly the same places, nor will the strict definition be consistently adhered to in the practical treatment of the subjects by the same writer. But the division is, in the main, logical and- a convenient one for all practical purposes. Without entering unduly into details regarding the plan of Part I, attiention may be directed to the order of discussion of the climatic factors. As far as possible each element has been considered with reference, (a) to its diurnal period, (b) its annual period, (c) its variability, or non- periodic aspects of short and long duration. Tables and diagrams have been freely employed, the statistical tables permitting of greater accuracy of statement, the graphic method affording a readier means of presenting at a glance the salient features of the variability of the climatic elements from hour to hour, or from season to season. The Geographical Horizox of Baltimore. The State of Maryland is situated within three distinct physiographic provinces. The low, flat Coastal Plain, averaging about 60 feet above mean tide and cut up by tidal estuaries, extends from the Atlantic seaboard westward to a line joining Philadelphia. Baltimore and Wash- MARYLAND WEATHER SERVICE 31 ington. where it is separated sharply from the Piedmont Plateau, or middle province. The Piedmont Plateau is an undulating area with elevations rising to 700 feet or 800 feet, and extending westward to the mountainous and high plateau region of the Appalachian Province. The mountains of this latter province form a system of parallel ranges extend- ing from northeast to southwest across the state, rising to heights of 3000 feet. The city of Baltimore is partly on the Piedmont Plateau and partly on the Coastal Plain. The country to the north and west is gently undulating; to the east and south it is level and but a few feet above the adjacent estuaries of Chesapeake Bay. ATMOSPHEEIC PEESSUEE. As a direct climatic factor the pressure of the atmosphere, and varia- tions in this pressure, are of comparatively minor importance. The effect of changes in the height of the barometer upon the human system does not begin to be recognized until the rise or fall is very marked. A diminished pressure causes in most persons a feeling of lassitude with increased difficulty of breathing. But this physiological effect is not ex- perienced, excepting by extremely sensitive persons, until the barometer shows a fall of three or four inches, equivalent to an ascent of three or four thousand feet above sea-level. The extreme variations of pressure at any one place do not often exceed an inch within the period of a few days. At Baltimore the extreme range has been but slightly over two inches in the past thirty-three years. The change in the pressure ex- perienced during the passage of the severest type of cyclonic disturb- ance is less than the permanent diU'crcncc in pressure between the east- ern and the higher western portions of the State of Maryland; and hence less than is experienced by travelers daily in passing from Balti- more to Pittsburg, over the Alleghany ^Mountains. As an indirect fac- tor, however, in the climates of the world, and as a direct agency in causing movements of the atmosphere, the pressure changes are of the highest importance and take rank witli those of temperature and rain- fall. In anotlier part of this report, the more general relations of pressure will be discussed in connection with the consideration of storm move- ments. Tlie following pages are devoted mostly to the local conditions 32 THE CLIMATE OF BALTIMORE and variations of pressure at Baltimore, based upon observations made since 1871 under the auspices of the United States Weather Bureau. Observations were begun in Baltimore on January 1, 1871, and have been maintained in an unbroken series to the present time. Standard 29.000 inches. TABLE I.-MEAN HOURLY BAROMETRIC PRESSURE. [In inches and thousandths.] Local time is 6 minutes slow. 75th mer. time. Jan. Feb. Mar. Apr. [ May June July jAug. Sept. Oct. Nov. Dec. Year 1 A.M. 2 3 .938 .940 .937 .947 .953 8 .965 9 j .978 10 I .981 11 I .974 Noon .953 1 931 .920 .918 .921 .926 .932 .940 .943 .945 10 945 11 943 Midnight 939 Average. .885 .883 .880 .881 .886 .894 .906 .912 .911 .906 .891 .869 .855 .a5i .851 .857 .866 .875 .878 .883 .885 .883 .881 .943 .881 .892 .888 .883 .883 .89] .900 .910 .916 .920 .917 .908 .896 .877 .862 .854 .849 .853 .860 .871 .880 .888 .893 .874 .869 .867 .868 .875 .886 .897 .901 .901 .900 .892 .879 .866 .852 .840 .835 .835 .838 .848 .863 .873 .877 .880 .881 .821 .817 .816 .818 .826 .SSI .845 .851 .852 .850 .843 .832 .819 .806 .794 .787 .784 .786 .795 .806 .817 .820 .832 .823 .820 .840 .836 .835 .840 .849 .858 .866 .871 .871 .868 .863 .854 .841 .830 .819 .810 .807 .810 .817 .836 .836 .843 .844 .843 .841 .835 .831 .830 .835 .844 .853 .861 .866 .867 .865 .859 .850 .813 .805 .802 .804 .810 .819 .830 .835 .836 .835 .850 .847 .846 .848 .855 .864 .874 .880 .883 .883 .878 .868 .856 .842 .831 .835 .823 .823 .831 .841 .851 .855 .857 .857 .923 .920 .930 .932 .929 .938 .948 .954 .959 .958 .949 .938 .933 .908 .897 .891 .891 .894 .903 .914 .933 .926 .938 .936 .959 .955 .952 .954 .960 .993 .994 .9fl3 .985 .969 .951 .937 .931 .929 .933 .939 .946 .954 .957 .961 .960 .958 .959 .9.59 .958 .959 .964 .970 .981 .991 .995 .993 .981 .963 .944 .934 .933 .9.36 .941 .9.50 .957 .964 .965 .966 .965 .963 .966 .967 .967 .964 .963 .968 .978 .990 .996 1.003 .990 .971 .950 .940 .939 .944 .950 .956 .966 .971 .973 .973 .973 .970, 0.895 0.893 0.891 0.893 0.898 0.906 0.916 0.934 0.927 0.937 0.919 0.905 0.889 0.876 0.868 0.865 0.867 0.873 0.880 0.888 0.895 0.898 0.899 0.897 .853 .924 i .959 ' .962 I .968 0.895 Table I contains the average hourly values of the station pressure at Baltimore for the period of ten years from 1893 to the close of 1902. The values are derived from the continuous record of a Richard barograph, cor- rected to agree with personal observations of a mercurial barometer made daily at 8 a. m. and 8 p. m. Each mean hourly value for the year is based on over 3600 observations; hence these values may be regarded as very close approximations to normal averages for each hour of the day for the year. The station elevation has been 123 feet above mean tide since August 1, 1896. The average hourly pressures are also shown graphically in Figs. 1 and 2. mercurial barometers were read at stated hours from two to five times daily. Since 1893 a self-recording barograph has furnished a continu- ous record of the pressure conditions and changes, affording excellent material for an analysis of the diurnal fluctuations of the barometer. The rich and abundant material accumulated by the Weather Bureau during the past thirty-three years has been reduced and di.scussed with ilARYLAND WEATHER SERVICE 33 a view to disclosing the nature of the diurnal and annual periodic changes of pressure, as well as the irregular and secular changes. Mdt. 3 6 9 Noon 3 6 9 Mot Mot. 3 6 9 Noon 3 6 9 Mot. 29.00+ In- .98 92 i 1 ' ■ 1 ! ' ' L/^ / i N ' I 1 ' 1 ■ ; 1 .^^ ' ■ , , 1 ' 1 ' , j 1 ■ ' M 1 1 ! ; , : ' ' ' -90 .84 -n — '^ r Arfl. 1 1 ' 1 ' ' 1 ' _.^-*-i_ ' j : 1 ' , : i "\ "^"*^" V"^*"" i i ' rv 1 i i' ' I ' 1 1 i 1 1 iNi 1 X : 1 1 I T^T' 1 1 , Mi 1 1 ' Inches • 94 2 9 90 1 , n "^ "■ 1 1 ' -^n 1 ' 1 ' 1 ^ ] 1 1 1 M Year. - ' 1 i i ' 1 ' 1 ' 1 ' 1 1 : 1 1 i ' . , 1 I ' ! 1 ! 1 ■ ' n ; ; ' ■ > 1 TT- ..,__. t M ' ' ; 1 ' ' ' ' ' ■ 1 1 i ' ' ' , 1 MM 1 ' 1 — -^ 1 ; ] ^1.1 1 i 1 ! ! 1 ■ .^! 1 1 M ' i : 1 1 : 1 1 ; ' ■ ' 'X' MM ' ' 1 , Zl ' ' i M \ 1 1 ' ' M , ^ ' ' ' . ^ MM ; ' i ./ 1 *■ 1 i 1 Ml. 1 ' ' 1 1 ' \ ' ' ^-^-^TH 1 ^^^T- -T""'^ ■ r^ \ f i l'^ C ' I'M , ' 1 ! ! j I I'll \' Mil' 1 1 j/f dix 1 ! \' 1 , I J ' 1 1 1 ' ' ; ' ■ ;\. ' r~ ■ • 1 1 1 1 , 1 i ' i 1 i > ; 1 ■ 1 , ^ ia-. J-^ [ 1 1 1 1 ■ i j ' : 1 T* 1 ' 1 ' j i -n , ' ■ , ! : ' : 1 1 1 1 ; 1 j j 1 t 1 1 i 1 1 1 pi: - ^ - ' ' : 1 t • ' M 1 Inches .94- ■86 Mot. 3 6 9 Noon 3 6 9 Mdt. mdt. 3 6 9 Noon 3 6 9 Mot. 29.00r In- •88 30.0 j ; 1 j , 1 1 --.„ 1 MI ! 1 ' 1 ' ■ ' ! 1 ! I ■ I 1 y^ ' X . ■ ' i-*''***^**- /^ ' 1 . V ' ' ' / '^'V^T . ^ \^ y ^NtO^' ' * ' ' ' , ! 1 t. 1 . . : ' 1 ■ ! Fig. 1. — Hourly Variatious of the Barometer. Fig. 1. The (trcinue height of the barometer at each hour of the day for the ten years from 1892 to 190.S is shown in the alwve diagrams for the months of January, April, July, Oct., and for the entire year. The height of the column of mercury which the pressure of the atmosphere sustained is expressed in inches and hundredths of an inch. See also Fig. 2, and Table I. 34 the climate of baltimore The Diurxal Variations of the Barometer. Within the tropics, wliere C3'clonic storms are of infrequent occur- rence, the diurnal variation of the barometer is the most marked fea- ture of the barometric curve. So regular in form and distinct in out- line are these changes that it is possible by inspection of the curve to tell approximately the time of day. The amplitude of oscillation near the equator is about one-eighth of an inch. This amplitude decreases with distance from the equator but is still recognizable in the latitude of 70 degrees. Along the parallel of Baltimore the amplitude is quite marked in a curve representing average hourly changes for the period of a month or more, but is detected only by the experienced eye in the daily curve, owing to the relatively large irregular changes due to the passage of the cyclonic storms of the middle latitudes. At times, especially in the summer months when tropical conditions prevail for a considerable period in our latitudes, the diurnal variation is very distinct for days at a time. (See the curve for August 7-13, 1900, on Plate II.) THE normal DIURXAL VARIATION AT BALTIMORE. The mean hourly values of barometric pressure for Baltimore are pre- sented in Table I for each month and for the year. The results for each season and for the entire year are also shown graphically in Fig. 1 on page 33 and Fig. 2 on page 36. In Table II the same values are ex- pressed in terms of departures from the average value for the entire day. These tables and diagrams reveal for Baltimore the characteristic double barometric curve so well known to the meteorologist from the results of analyses of observations in all parts of the world, with perhaps minor peculiarities due to local conditions. The fluctuations are well marked in all months of the year, the amplitude varying from 0.060 inch in August to 0.071 inch in March. In Fig. 2 the distribution of pressure is presented by a method not frequently employed but one which shows clearly and in compact form the successive changes from hour to hour throughout the year. Upon a system of coordinates representing the hours of the day and the months of the year, the curved lines of equal pressure are projected in such manner as to enable one to find the exact pressure at any hour of any month. These curved lines are sometimes called " isopleths.'^ For example, to find the average pressure at noon^ MARYLAND WEATHER SERVICE 35 in Ajiril, you run clown the vertical line marked noon, until the horizontal line marked April is intercei^ted, and find the isopleth of 29.875. This method enables ns also to see at a glance the chief characteristics of the seasonal distribution, further emphasized by differences in shading, the lighter shades indicating the lower pressures of the warm months and the darker shades the higher pressures of the colder months. TABLE II.-HOUKLY DEPARTCKES FROM MEAN DAILY PRESSURE. ,In thousandths of an inchj Local time is 6 minutes slow. 75th mer. time. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year 1 A. M —.004 2 -.005 3 -.003- 4 — .OOT- 5 —.006 .004 .010 .022 .035 .038 .031 .009 -.012- — .023 - — .025- — .022- 5 —.017- 6 -.011 10.... 11.... Noon 1.... 2 3.... 4.... 7... 8. —.003- .000 .002 .003 .000 Midnight —.004 .004 .002 -.001- -.002- .000 .005 .013 .025 .031 .030 .025 .010 -.012- -.026 - -.030- -.030- -.034- -.015- -.006- ■. ft < 11:00 10:30 IhOO 11:30 11:30 11:00 11:30 IIMO 11:0U 11:00 10:00 10:00 10:55 Night minimum. Mid- n't. 12 12 3 4 6| 2. 1. 3. 5 2. 1..I 1 1129582915 1 9 3! 1 3:30 3:30 3:00 2:30 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:05 Table III shows the average hour of occurrence of the daily maximum and minimum station pressure for each month and for the year during a period of ten years, and also the extent to which the times of occurrence of these phases have varied from the average time. For example, examining the afternoon minimum phase, the most variable of all, we find that in 10 years it occurred in January 7 times at 3 p. m., 4 times at 4 p. m., and once at 5 p. m.: and that the average time of occurrence for the year is about 3 p. m.; that in August it occurred at 5 p. m. 6 times and at 6 p. m. 5 times; and as an average time we have about 5.30 p. m., etc. The average values are also shown graphically in Fig. 3. primary maximum and primary minimum with the approach of sum- mer, from five hours and a half in the winter months to eight hours and a half in May and June. This increasing interval is due mostly to variations in the time of occurrence of the primary minimum. These phases are shown in detail in the following table and in Fig. 3. 38 THE CLIMATE OF BALTIMORE INTERVALS BETWEEN' PRINCIPAL PHASES OK PRESSURE. (Expressed in Hours and Minutes.) Intervals between. Jan. Feb. Mar. Apr. May June i July Aug. [Sept. Oct. Nov. Dec. Year A. Primary max. and min. B. Secondary max. and min. 5:30 4:30 5:30 5:00 7:00 4:00 7:30 3:00 8:30 3:30 8:30 4:00 8:00 3:30 8:00 3:30 7:30 4:00 7:00 4:00 6:00 5:00 5:00 5:00 7:00 4:05 FMAMJJASOND ■ '~ ill: ' ' , 1 -- ■^ : 1 ' 1 Mi , 1 V 1 s > 1 1 ' ; \ / s. -* y ^ ;a ' [ ^^ _ > S 1 v, k-j ^ ■" V l^ 'V rf _. ^ , p S 3PM 1 ^ / s 1^ r"~ 1 h-j ' ' ' • 1 1 1 A 9 A. M. - i; L _ . .. t^'j . ^-1 > SJ ,<" -? "n 6A M.- fr/ 1 3AM- " N ! V r' i 1 ^ i ' i 1 r : I 3 A. M. 9P.M. 3 P. M. - 9A.M. 6 A. M. Fig. 3. — Principal Phases of Diurnal Oscillation of Pressure. Fig. 3 indicates the time of occurrence of the maximum and minimum points reached by the barometer in the diurnal oscillation. The upper marginal letters represent the months of the year ; the figures indicate the hours of the day. The curved lines represent i-espec- tively (a) the time of occurrence of the secondary maximum; {h) the primary minimum ; (c) the primary maximum ; (d) the secondary minimum. See also Table III. MARYLAND WEATHER SERVICE 39 AMPLITUDE OF OSCILLATIOX. (In Thousandths of an Inch.) Jan. Feb. . Mar. I Apr. , May June A. Diurnal Amplitude. 63 61 71 66 B. Nocturnal Amplitude. 9 6 10 14 ft* 64 6 9 July, Aug. Sept.' Oct. | Nov. Dec. 60 68 65 63 64 1 11 8 1 9 8 10 Tear 63 TABLE IV.-HOL'RLY VARIATIONS OF PRESSURE ON CLEAR AND ON CLOUDY DAYS. (Expressed In thousandths of an inch as departures from the daily average.) Clear days in To'al (90 days Departure. Cloudy days in Total (90 days) Departure. 75th Meridian Time. Jan. and Feb. (60 days) Departure. July (.30 davs) Departure. Winter (60 days) Departure. Summer (30 days) Departure. Midnight . ... 1 2 3 4 5 6 -.053 -.037 -.029 -.021 -.018 -.005 -.004 + .022 + .041 -.056 + .063 -r.058 -r.040 + .019 + .003 -.008 -.013 — .014 -.014 -.012 -.015 — .019 -.026 -.036 -.051 -.006 -.006 -.007 -.008 • -.002 + .009 + .018 + .028 + .034 + .0.34 + .a32 + .028 + .023 + .009 -.004 -.017 -.026 -.0.38 -.035 -.0:31 -.025 -.016 -.013 -.008 -.006 -.028 — .022 -.018 -.014 -.010 + .002 + .011 + .025 + .038 + .045 + .048 + .043 + .032 + .014 .000 — .013 — .020 -.024 — .024 -.022 — .020 -.018 -.020 -.022 -.028 -.003 + .010 + .013 + .012 + .006 -^.006 -.008 + .015 -.001 -.004 -.011 -.014 -.011 -.004 + .011 + .023 .000 + .003 + .001 -.001 —.002 + .001 + .010 + .019 8 -.024 +.029 4-. 026 9 10 11 Noon 1 2 3 4 5 6 + .032 i +.034 1 +.033 -+-.034 +.030 +.033 + .024 +.027 -J-.026 + .U04 +.020 , -.012 -.016 +.005 -.006 -.025 -.003 -.014 —.027 —.011 -.019 —.027 -.017 -.022 -.023 -.020 -.022 -.025 -.017 -.021 —.011 —.013 —.013 8 — .006 —.004 —.005 9 — .003 +.001 —.001 10 11 Midnight -.003 -^.003 .000 + .003 +.001 +.002 + .001 +.001 .000 Table IV shows the amount of the diurnal variation of pressure on clear days as compared with cloudy days in order to detect any difference due to cloudiness. For this purpose CO clear days in January and February and ?,0 clear days in July were chosen, to be compared with CO cloudy days in January and Feliruary and HO cloudy days in July. The effect of irregular \ariations of the barometer due to the passage of storms was first eliminated from the actual means. The results are also graphically shown in Fig. 4. In individual months the time of occurrence of these phases varies considerably from the average times for the entire ten years, as may be seen in Table III, which shows the frequency of occurrence of the differ- ent phases for each month for the ten -year period. 40 THE CLIMATE OF BALTIMORE DIURNAL VARIATIONS OF PRESSURE ON CLEAR AND CLOUDY DAYS. In Tables I and II and Figs. 1 and 2 the average distribution of pressure is shown for all conditions of the weather during a period of ten years. In order to determine the effect, if any, of cloudiness upon the oscillation of pressure, selection was made of 60 clear days in Janu- ary and February and 30 in July to be compared with a like number of MDT + .05 9 Noon -.0 5 ••-.05 -.05 ♦.05 y"" ^^ 7 \ j/ .'' ■* X, •' — ~ - . ^'^ ^ S j' ■ _i,_.j_^_:^Jl2_. -,,_.,.-,__,.- — .».— -^«»-. --!.--. .ii. J." *. ir ,^ f'> '- " ^^^ "^ ""-^ XL ^■^j^r'^X- xS ^"^i / s i I """ ' : .^sf--^ :" "'-^-^ Tfy' ; ..^>s^ ,„ . v.--^-^-v^ ~'--+' "*?.-'—•,_ "'^ ^ — -"''^ 1 >^ ^jfyf"^ "*-._ --"'' ' *" " ~'^5» 1 >;:, — H^ ^> ^^P 1 "'nj^ V -.. . .>^ > S < ^. ..-.., 1 '•i-jfj;^ ■' "■" -■ ■ ■ - TT-- '-S^ ' -1- \ i ,-r-- \ \ ^ 1 ^"Hn ' •!, "«.._ 1 '•'f\ -t— — 1 1 -j-^ r "t--sii-+«-— T "*■-«, 1 . 1 i i ' 1 i -.05 ud5 /) -.05 +.05 Fig. 4. — Diurnal Variations of Pressure on Clear and on Cloudy Days. Clear. Cloudy. Fig. 4 shows the hourly chanKes of the barometer on selected da3'S approximately similar in all respects excepting as to tlie amount of cloudiness. The cootinuous lines show the movements of the barometer on clear days and the dotted lines on cloudy days, for (a) win- ter months, Q>) summer months, and (c) for the year. The hourly heights of the barometer are expressed in hundredths of an inch, as departures from the average height for the entire day. See also Table IV. cloudy days in the same months, as far as possible. The computed variations are shown in tabular form in Table IV, and graphically in Fig. 4, on page 40, after first eliminating the effect of irregular fluc- tuations of the barometer due to passing cyclonic disturbances. The primary maximum and minimum phases differ but little from those of MARYLAND WEATHER SERVICE 41 the normal curve, although the amplitude on clear days is somewhat exaggerated, especially so during the winter months. The curves for the normal and for the cloudy days coincide very closely for the sum- mer months and for the year, but diverge in the early morning hours of the winter months. The most striking feature of the curve for totally clear days is the wide divergence from the normal curve in win- ter during the night and early morning hours. THE DIURXAL BAROMETOIC WAVE. The diurnal variations of the barometer described in the preceding paragraphs are not simply of local occurrence but are part of a general phenomenon extending over the greater portion of the earth's surface. The maximum and minimum phases pointed out occur in all localities at approximately the same hours of local time. As stated above, this pressure wave, as it may be called, has its greatest development in or near the equatorial belt, and diminishes in amplitude with distance north and south of the equator. It has some resemblance to a double atmospheric wave passing completely round the earth from east to west every twenty-four hours, having a velocity at the equator of about one thousand miles per hour. By plotting upon a map of tlic world the departures from the normal daily pressure for successive hours of the day at a large number of stations uniformly distributed over the north- ern and southern hemispheres, and joining such stations as have equal departures of pressure for the same hour, we have presented to us four systems of pressure-distribution, consisting of two areas of low pressure and two areas of high pressure. These systems completely encircle the globe and closely resemble in form the cyclonic and anticyclonic systems of the middle latitudes, but differ from them, among other things, in covering an area vastly greater, and in moving in the opposite direction. The diurnal fluctuations of the barometer are the local evidence of this vast double atmospheric wave passing round the globe daily. The west- ward propagation of these waves near the equator is represented in Fig. 5 on page 42 ; the curve shows the time of occurrence of the different phases of the double wave, its amplitude, and the direction of propaga- tion along the path of greatest development. The character of these waves is further indicated in fhe diagrams of Plato T, in which the succes- 42 THE CLIMATE OF BALTIMORE sive areas of high and low pressure are exhibited at intervals of two hours in passing from east to west across the North and South American continents/ This double atmospheric wave, or tide, is so intimately associated with the apparent diurnal movements of the sun that the conclusion is almost irresistible that the pressure changes are due primarily to changes of temperature. This relationship has not yet been satisfactorily demon- strated to be that of direct cause and effect, but there seems to be a gen- eral consensus of opinion that the primary maximum and the primary minimum phases of pressure are direct effects of the sun's heat. The N. t om 3am M^dn 9 pm , 6pr, 3om Noon 9am t am *040 t020 fO-IO tO?0 / \ / \ / \ W.o -020 -040 / \ oE -0?0 -040 \ ^ / V / / V y s. """1 Fig. 5. — The Diurnal Barometric Wave. Fi^. 5 shows the direction of movement of the diurnal barometric wave, from east to west around the globe ; also the local time at which the crests and the hollows of the wave pass over any locality along the path of the greatest development of the wave, near the equator. The extent of the diurnal rise and fall of the barometer is shown by the figures to the right and left of the diagram, which express the departures above and below the normal height for the day, in thousandths of an inch of mercury. See also Plate I. theory advanced many years ago to account for the chief maximum and minimum phases seems plausible. At the time of day, between 9 a. m. and 10 a. m., when the atmosphere is being warmed most rapidly and the tendency of the air to rise in consequence is greatest, the upper and colder layers impede this upward movment, resulting in a temporarily increased tension at the surface of the earth. When this tension is re- lieved the barometer begins to fall, reaching its lowest point about the 'Fassig, O. L. The Daily Barometric Wave. BulL No. 31, U. S. Weather Bureau. 8°. Washington. D. C, 1902, pp. 62-65, 12 pis. rHK DHUNAI. IIAKIIMETKIC W A\ !•:. (75T]1 MERIDIAN TIMK.) ?:in pressure of tin- day arc expressed in Ihonsaiidtlis of an inch of mercury.) MARYLAXD WEATHER SERVICE 43 middle of the afternoon when the upward movement of the warm air may be assumed to be least impeded. In this connection, Fig. 17, on page 76, is significant, showing the average hourly rate of change of temperature for the year, compared with the curve representing the average hourly variation of pressure for the year. As has already been stated above, the pressure-wave attains its greatest amplitude in the equatorial belt where the diurnal temperature changes are greatest, and over the continental masses north and south of the equator where the diurnal range of temperature is most marked. (See Plate I.) According to Dr. Hann,^ in seeking an explanation of the diurnal variations of the barometer: "We had better deal with the action of the sun on the upper strata of the atmosphere and treat this as the principal cause. The actinometrical observations show us that these upper strata absorb a considerable amount of heat. The diurnal heat- ing action of the sun on the upper strata would harmonize far better with the general uniformity of the daily barometric oscillation along the different parallels of latitude as well as with its general independence of weather. We need not quite exclude local influences, but these seem to be more of a secondary character." This view is also held by Lord Kelvin, who seems to have been the first to suggest this explanation. CORRECTIONS FOR REDUCTION TO TRUE MEAN PRESSURE. The determination of the daily mean barometric pressure based on 24-hourly observations for ten years enables us to apply the necessary corrections to any given combination of daily observations in order to obtain a true mean. The following table contains the corrections for each month and for the year which must be applied to the series of observations made according to any of the five systems most frequently employed in barometric observations in this country. The average of the three observations made at 7 a. m., 2 p. m., and 9 p. m. approaches most nearly the true 24-hourly mean for the day, when the 9 p. m. ob- servation is given double weight. 'Hann, J. The Theory of the Daily Barometric Oscillation. Quart. Journ. Roy. Met. Soc, London, 1899, p. 40. 4 44 THE CLIMATE OF BALTIMORE CORRECTIONS FOR DIURNAL VARIATIONS OF THE BAROMETER. (In Thousandths of an Inch.) Hours of observation. 7A.+2P.+9P. 7A.+2P.-*-2(9P.) 7A.+3P.+11P. 10 A.+IO P. 8A.+8P. Jan. Eeb. Mar. Apr. May June +4 +4 -1 -3 -3 -3 +2 +2 —1 —2 -1 -1 +5 +5 +1 -1 -2 -20 -17 -18 -18 -15 -14 -11 -11 -12 -11 -8 -8 July Aug. Sept. Oct, Nov. Dec. Year -4 -3 -2 +1 +2 +5 -1 2 —2 -2 +2 +1 +2 -2 -1 +2 +3 +5 +1 —15 -16 -18 -18 -18 -20 -18 —1 -8 —10 -14 -16 -12 -11 The Annual March of Atmospheric Pressure. In order to determine the changes of pressure from day to day during the course of the year, the daily averages of the Baltimore observations covering a period of 30 years were reduced to what may be called normal values for each day of the year. In obtaining these normals the sea-level values were employed, but corrections for diurnal variation were not applied. The results are shown in Table Y on page 45 and graphically by means of curve (d) of Plate III. This curve shows a fall in pres- sure from month to month from January to May. During May, June and July the pressure remains fairly uniform, followed by a compara- tively rapid rise in August and September. During October there is a slight fall followed by a rise to January. The curve of daily changes does not, however, show a steady rise and fall from season to season. J F [V A ^ J J A c c N D 1 1 1 IncTies ; 1 1 1 1 i 1 I 1 ! i , I 1 30.00 1 ! 1 I 1 ~>i -f , 1 P-*- "V 1 ! -<■ r 'V, j 1 ; y " ' — •* ! .92 \ 1 I / I > / s / "i r- -n ^ '~-1 ^ 29.84 1 ( — 1 ' ! i t 1 - i 1 ! 1 .J, .. i i _ 1 1 i 1 _ _J _ _^ _ _ ^ 29.84 Fig. 6. — Mean Monthly Atmospheric Pressure. (See Table VI). MARYLAND WEATHER SERVICE. VOLUME 2, PLATE II. MARYLAXD WEATHER SERVICE 45 The progression is marked by successive waves varying in period from two to eight or ten days' duration, which persist even in the average dailv values for 30 vears. The variation from dav to dav is smallest TABLE V.-MEAN DAILY BAROMETRIC PRESSURE, Reduced to sea level. [In inches and hundredths.] 29.00 Inches. Date. 9.. 10.. 11.. 12. 13.. 14.. 15.. 16.. 17.. 18 19.. 20 . 21.. 22. . h'.'. 24.. 25.. 26.. 27. . 2S'.'. 29.. 30.. 31.. Average Amplitude. Jan. Feb. 1.14 .19 1.14 1.21 1.06 1.10 1.22 1.17 1.14 1.10 1.11 1.17 1.13 1.08 1.04 1.14 1.16 1.12 1.12 1.06 1.08 1.08 1.03 1.04 1.08 1.16 1.05 1.08 1.15 1.14 .96 1.11 .18 Mar. : Apr. 1.07 1.03 1.04 1.06 1.16 1.13 1.09 1.11 1.02 1.05 1.08 1.01 1.03 1.10 1.09 1.04 1.08 1.05 .96 .95 1.00 1.02 1.03 1.09 1.09 .99 .97 .97 1.02 1.04 1.02 1.04 1.03 1.00 1.00 .99 1.02 1.04 1.04 1.05 l.OI 1.00 1.03 1.08 1.03 .98 .96 1.03 1.07 1.04 1.02 1.03 1.08 1.08 1.03 1.03 1.03 1.03 1.03 .96 .96 1.00 1.02 .12 May 1.00 .98 1.02 1.01 .99 .99 1.05 1.03 1.01 1.01 1.01 1.00 .99 1.01 .99 1.00 1.04 1.03 .98 .99 1.00 1.01 1.03 1.01 .97 .97 .94 .96 1.01 .99 1.00 June 1.01 1.03 1.01 .99 .99 l.dO 1.00 1.00 .98 1.00 1.01 1.01 1.01 1.03 1.03 1.03 .97 .97 .98 .99 .96 .95 1.00 1.00 .98 .98 .99 .95 .97 1.02 0.99 .08 July Aug. Sept. Oct. Nov. Dec. 1.04 1.03 1.00 .99 .99 1.03 1.03 .98 .96 .99 1.00 1.00 .96 .98 .96 .96 .99 .99 1.00 1.01 1.01 1.01 1.02 1.03 1.00 .97 .98 1.00 .98 .98 1.00 0.99 1.00 1.00 .99 1.03 1.04 1.03 1.01 1.01 1.01 .99 1.00 1.00 1.00 .99 1.02 1.03 1.03 1.00 1.00 1.02 .99 1.01 1.03 1.03 1.01 1.05 1.06 1.07 1.04 1.02 1.05 1.03 .08 1.09 1.10 1.08 1.08 1.09 1.08 1.08 1.10 1.12 1.13 1.11 1.08 1.07 1.12 1.10 1.04 1.06 1.08 1.02 1.07 1.14 l.!3 1.09 1.13 1.12 1.07 1.09 1.12 1.09 1.10 1.09 1.11 1.13 1.11 1.04 1.05 1.05 1.09 1.09 1.12 1.11 1.11 1.11 1.08 1.07 1.13 1.11 1.11 1.11 1.11 1.11 1.13 1.13 1.07 1.12 1.14 1.08 1.04 1.07 1.05 1.09 1.06 1.09 1.11 1.10 1.13 1.17 1.15 1.17 1.15 1.08 1.06 1.05 1.09 1.13 1.13 1.12 1.10 1.20 1.18 1.13 1.04 1.09 1.13 1.14 1.06 1.12 1.12 1.11 1.11 1.13 1.12 1.17 1.12 .10 1.16 1.15 1.13 1.11 1.08 1.10 1.10 1 14 1.16 1.13 1.10 1.13 1.09 1.10 1.13 1.16 I.IT 1.15 1 22 1.22 1.17 1.15 1.15 1.16 1.15 1.05 1.06 1.17 1.11 1.17 1.17 1.14 .17 Average for the year 30.063 inches. Average amplitude 0.05 inches. Table V shows the average sea-level barometric pressure for each day of the year. The period of observation covers the 30 years from 1871-1900. The daily mean is based on three readings of the mercurial barometer at about 7 a. m., 3 p. m. and 11 p. m., from 1871 to June 1888, and on two readings at 8 a. m. and 8 p. ra., from July 1888 to 1900. The correction for diurnal var- iation has not been applied, but this is extremely small for the observations made at 7 a. m., 3 p. m., and 11 p. m., and about -f .01 inch for the series of readings at 8 a. m. and 8 p. m. "The number 29.00 should be added to each of the figures in the body of the table. The monthly range of the mean daily pressure is indicated in the last line of the table. The figures of this table are also represented graphically in curve D of Plate 3. 46 THE CLIMATE OF BALTIMORE M 1875 Fig. 7. — "Variations in the Mean Monthly Pressure (Expressed as Departures from the Normal Values for the Month, ia Thousandths of an Inch of Mercury). See Table YIL MARYLAND WEATHER SERVICE 47 during the summer months when it is generally less than 0.05 inch, and greatest in the winter months, when it rises to 0.10 inch, and, occa- sionally, to 0.15 inch. To what extent these irregular interdiurnal variations would be eliminated in a longer series of observations is a matter of conjecture. To a marked extent at least they are probably persistent and due to a periodic recurrence of certain types of weather at certain seasons of the year. Of special interest is the comparatively rapid rise in pressure from August to September, and the arrested upward movement in October, more clearly shown in Fig. 6, constructed from monthly averages, than in the serrated curve of daily means. The barometric waves of short period vary greatly in length and are not generally sharply defined, but in most instances they extend over a period of three and a half to four days, and are accompanied by in- verse variations of temperature as is clearly shown in the temperature and pressure curves of Plate III. The individual features of these waves are shown in Plate II, in which actual tracings of the baro- graph are reproduced as representative types for the different seasons of the 3^ear. The great variability of barometric conditions in the winter months and the comparatively uniform conditions in the sum- mer months are here shown in strong contrast. The curve representing the conditions for the week ending August 13th, 1900, is almost entirely free from irregular or non-periodic fluctuations, permitting the diurnal variations to be plainly recognized. AVERAGE MONTHLY AND ANNUAL PRESSURE. In an elaborate report on barometry,' Professor Bigelow has discussed in detail the reduction of barometric observations at Weather Bureau stations in the United States. In this report all observations from 1873 to 1899 have been reduced to the epoch of January 1, 1900. During this long period several different methods of reduction had been em- ployed, resulting in series of observations not strictly comparable. In order to obtain comparable values all reduction^ were recomputed and ' Bigelow, F. H. The Reduction of Barometric Pressure Observations at Stations of the United States Weather Bureau. Vol. II of the Report of the Chief of the Weather Bureau for 1900. 48 THE CLIMATE OF BALTIMORE TABLE VI.-MEAN MONTHLY STATION PRESSURE REDUCED TO THE WEATHER BUREAU SrSTEM FOR THE EPOCH JANUARY 1, 1900. [Inches and thousandths.] Lat. 39° 18' N., Long. 76° 3~'=5 hrs., 6 m. W. of Gr. Elevation above mean sea level 123.3 feet. Gravity corr. — .01.5. 29.000 inches. A^ear. j Jan. Feb. Mar. Apr. May June 1873 ! 0.971 0.886! 0.8841 0.808' 0.866 0.859 1874 ! 1.053 1.0:il .879 .8951 .796 1875 1.083 .986 .947 .833 .835; .874 1876. 1877. 1878. 1879. 1880. 1881. 1882. 1883. 1884. 1885. 1886. 1887. 1888 1889. 1890. 1891. 1892 . 1893. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901 . 1902 . 1903. Mean (1873-99). Corr. for sea- { level ) 1.044 1.042 .963 .963 LOGO 1.023 1.048 1.075 1.025 1.006 .928 .934 1.070 .912 1.080 .921 .917 .847 1.023 .909 1.034 1.009 .895 1.032 .951 .995 .127 1.014 .972 .831 .974 .993 1.063 1.025 1.155 .958 .957 1.080 .958 1.026 .956 .954 .993 .996 .764 .935 .969 .918 .896 .762 .947 .968 .137 .918 .974 .841 .980 .943 .642 .983 .873 .909 .922 .819 .852 .942 .783 .908 .949 .876 .920 .963 .874 .864] .938 1.073 .8161 .855 .851 .809 .891 .907 .889 .756 .906 .870 .983 .819 .965 .874 .943 .891 .895 .9811 .960 .828, .939 .884 .8011 .803 .761 .938 .868 .884! .865 .796: .939 .929 .907 .879 .839 .824 .814 .78l! .884 .843 .816 .819 .9011 .838 .759 .793 .933 I .877 .831 .793 .913 .840 .726 .894 .994 .853 .773 .783 .841 .935 .877 .839 .864 .801 .920 .839 .823 .863 .903 .831 .858 .783 .792 .899, .877 .852 .851 .136 .133 .133 .139 Mean, sea-level 1.122 1.105 1.025 1.010 .985 .990 July Aug. Sept.' Oct. Nov. Dec. 0.888 0.916' 0.959! 0.934' 0.865 1.039 .871 .884 .940' .981 1.056 1.U45 .838 .873, .934' .900 .993 .925 .877i .830 .811 .848 .853 I .817 .894 .856 .744 .833 I .800 .830 .879 .838 .880 .934 .834 .870 .840 .814 .899 .839 .856 .818 .871' .8341 .761 .831 .918: .894! .9001 .905} .9161 .833, .8411 .847! .859 1 .930, .878 .858 .865 1 .834 1 .881 .838; .901 .843 .873 .831 .895 .850 .953! .9771 1.013 .918, .906 .935 .857 1.043 .984' l.OlO 1.1171 .928 1.077 .911 1.0.50 .928 .941' 1.013 1.065! 1.037 .938, .964' 1.053; 1.015 .9.34 1.036 1.0.53 1.004 .968 1.016 .964| 1.047 .936 .8711 .830 .911 .984! 1.028! .976 .9101 .918 .848 .881 .886: .971! .771! .898 1.004 .953, .996 .992! .941 .932 1.022 .936: .928 1.015' 1.0241 .910 .949 .9211 .895! 1.006 .9421 .9211 .941! .955, 1.036 .895' .938 .959! .980' .862' .976! .914 1.033; .910 .990 1.053 1.029, 1.063' .974 .953 .926 .918 .883 .931 1.033 .918 .833 .904 .9411 .951 .855 1.009 .9171 .964 1.054 .940 1.004 1.000 .983 1.053 .947 .928 .962 .966 .950 .971 Year .850 .869 .945 .942 .976, .988 29.917 1 : 1 I .130 .136 .129 .138 .134 .13" .980' 1.005 1.075' 1.080, 1.110 1.125 30.050 .133 Table VI presents the average monthly and annual station pressures for each month and year from 1873 to 1899, as recorded in Professor Bigelow's Report on Barometry, wjth the addition of values for 1900 to 1903. All ob- servations used in this table were reduced to the same plane (123 feet above mean tide), to the true mean of 24 hourly observations 'and corrected for the force of gravity at the Station. The values in the footings of the table are Professor Bigelow's " normals " for the period 1873 to 1899. MARYLAND WEATHER SERVICE 49 uniform corrections for gravity and for diurnal variation were applied by Professor Bigelow. The corrected monthly and annual means for Baltimore as given by Professor Bigelow are reproduced in Table VI on page -iS, with the additional values for the years 1900 to 1903, simi- larly reduced. For the methods employed in the reduction of observa- TABLE VII.-DEPARTURES FROM AVERAGE STATION PRESSURE. [In thousandths of an inch.] Jan. Feb. Mar. Apr. May 1873 !— .034!— .082— .015— .069 +.014 1874 ! + .058 +.053 -.030+. 018 -.05B 1875 | + .088| + .018| + .048-.044-.017 1876 ' + .049 +.046 +.019 +.011 +.076 1877 +.047 +.004 +.075 -.036 +.032 1878 -.033 —.137 -.058 —.188 —.0,56 1879 -.032 +.006 +.081 -.068 +.087 1880 I + .065 +.025 +.044, + .014 +.077 1881 ' + .028+.095— .2.571— .091 +.055 1882 +.053 +.057 +.084 +.030 +.027 1883 I + .080+.187— .026+.012— .013 1884 I + .030— .010 + .010|— .121 —.028 1885 + .0U|— .093: + .023 + .029 —.038 1886. 1887. 1889.. 1890.. 1891. 1892. 1893. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901. 1902. 1903. .— .067 -.011 -.080 +.072— .071 . I— .071 +.112— .047— .007 +.032 . +.0751— .010 +.043 +.106— .010 . — .083' + . 058— .116]— .0.58— .03« . +.085— .0131 + .009 +.088— .033 -.074 -.014 +.050 -.078| + .024-.023 -.148 +.028, + .021 + .028 +.018 +.064 .^.086,— .089 -.026 -.003' + .049 + .065— .014 + .014 -.093 + .018— .060 —.008+. 071 June + .008 —.023 + .023 + .017 + .014 — .063 + .002 + .014 -.078 — .068 — .010 + .084 + .026 —.022 + .013 -.050 + .031 -.006 -.024 -.004 + .015 + .011 + .069 + .038 + .021 -.012 + .027 +.067 — .0P6[— .036 — .020— .045 +.006| — .007 .039! -.108 +.031 —.040 .002— .038 +.066!+. 101 + .002 +.049 —.028, + .042 Means. Dept. '-.0391— .204 —.035 +.104 +.025— .013 + .014 — .0:33 +.039 J-.OKi -.031 -.028 — .100 +.001 +.174 -.049 — .059 +.01S + .0;j7i-.0.50— .OKS +.062 +.0<;0l + .O51 -.044 -.072 —.011 +.0371 -.012 -.020 .107 .020 .129 -.140'-. 113 — .206— .044 — .021 +.169 .078 .076— .126 .074' + .042 .116 +.143 .852 + .00' —.069 -.059 .851 051 —.018 —.040 -.065 -.066 -.067 July Aug. Sept + .047 + .015 + .004 Oct. + .013 -.008 -.006; + . 039 -.023 -.042 .033' + .025 + .044 +.031 + .006 +.036 — .106; + .047 -.027-. 037 -.050— .028 — .020— .022 + .0291 — .010 -.022 +.061 + .030 +.009 + .038 + .074 -.016 + .020 -.011 -.004 — .035 + .012 -.010,-. 041 + .039 +.032 -.036! -.026 + .049 +.004 -.011 —.038 + .006 +.026 -.032 + .021 -.032 .850 + .014 -.03' -.014 -.005 +.071 -.018 +.022 -.012! + . 094 + .022: + .074 -.010-. 071 + .a38 +.086 + .030— .03; -.028! — .094 — .065!— .056 + .025 -.171 Nov. Dec. -.111 +.051 + .080 +.057 + .017|— .063 -.119!— .060 + .020! + .089 -.109— .077 + .064 +.062 + .141J-.060 + .089! + .049 + .076 +.027 + .076 +.016 — .012, + .059 -.146-.07 —.078' + . 016 — .023! + .008 + .016— .047 -.044 +.034 —.040— .060 + .069' + .013+.C60! + .066 + .078— .047 -.038— .048 -.036 +.017 +.004 +.016 + .ro3-.080 .000+. 012 -.025!— .038+.057|-.005 -.0511-. 033 +.087, + .065 + .060 +.048!-. 003 -.041 -.004 +.030 —.023 —.060 -.025 +.111 -.050 —.026 -.005 +.087— .058 -.033 -.015 -.043 + .063 + .091 -.001 -.025 946 .943 .048 .029 .G25 -.058-.038 -.025 -.017 —.012 —.061 .059 .988 .071 Year -.011 .020 + .001 + .001 + .008 .073 + .028 + .033 .004 + .031 + .038 .005 -.033 015 -.001 + .002 024 -.006 + .019 -.001 -.017 + .005 -.009 + .005 + .005 -.001 + .004 —.009 -.048 — .038 -.007 19.917 Table VII presents the average monthly and annual pressures expressed in terms of departures, in thousandths of an inch, from the normal values for the period 1873-1899. The normal for each month and for the year is shown in the first line of lootings, and the departures of the monthly normals from the annual normal are shown in the last line. These departures are also graphically shown in Fig. 7 and Fig. 8. 50 THE CLIMATE OF BALTIMORE tions and for further particulars in reference to the Baltimore pres- sure data the report of Professor Bigelow should be consulted, espe- cially pages 176, 646 and 798. In Table VII the mean monthly and annual pressures from 1873 to 1903 are given in terms of departures from the monthly and annual normal values. The normal monthly and annual pressures derived from the mean of the daily averages (see Table V) differ somewhat from those derived from Professor Bigelow's monthly means (see Table VI) after reducing the latter to sea-level. This discrepancy is due to the fact that the daily means were taken directly from the original record of observations of the Baltimore Office of the Weather Bureau to which the correction for diurnal variation had not been applied. IBTr 1875 Fig. 8. — Annual Variations of Pressure Expressed as Departures from the Normal Value. (See Table VII.) ANNUAL AND SECULAR VARIATIONS OF PRESSURE. The average atmospheric pressure of a year is by no means a constant quantity. The fluctuations in value from year to year are sometimes considerable. This is most readily recognized when the variations are graphically presented as in Fig. 8 on page 50. Here the Baltimore ob- servations are plotted in terms of annual departures from the normal value for the entire period from 1871 to 1903. The amplitude of fluctuation is expressed in thousandths of an inch. The resulting curve presents a series of waves or surges varying in amplitude from a few thousandths of an inch to nearly one-tenth of an inch. The period of oscillation also varies considerably, yet there is a remarkable uniformity in the length of these periods. Measuring from crest to crest and from hollow to hollow of these waves Ave have the following figures repre- senting the periods in years and fractions: MAKYLAXD "WEATHER SERVICE 51 Number of Years. Mean. From crest to crest (from 1871). 4 4 4 4 5.5 4.5 4 3.5 3.5 5.5 4.2 From hollow to hollow (from 1873). 3.5 3.5 4 4 5 5 5 3 4 4 4.1 Since 1871, the beginning of the series of observations at Baltimore, no crest of a wave has been heloiv the normal value for the entire period and no hollow has been al)ove the normal level. In order to fall into harmony with this series the year 1903 should form the crest of a wave and be followed by approximately equal pressure in 1904 and lower pressure in 1905; but we must not overlook the fact that it is the unexpected which is most likely to follow a long-range forecast. Tlie period from 1871 to 1903 includes ten waves with an average length from crest to crest, or from hollow to hollow, of slightly over 4 years, the limits of variability being three years and five years and a half. A conspicuous feature of the annual variation of pressure at Balti- more is the abnormally low pressure of 1878. The departure in this year was nearly five times the average annual variability. Upon first examination it appears suspiciously large. It is, however, substantiated by similar departures, though not so marked, at stations in all parts of the United States. A few of the larger departures occurring in this year are here given : DEPARTURES FROM NORMAL PRESSURE IX 1878. Baltimore — .072 Washington — .057 New York — .052 Cincinnati — .080 St. Louis —.049 New Orleans . St. Paul San Francisco Key West . . . . Boston —.068 —.052 —.058 —.059 —.056 The abnormally low pressure evidently extended over a very large territory in 1878. At Baltimore the pressure was decidedly below the average during every month of the year, excepting September, when it was but slightly above. Usually there is considerable fluctuation above and below the annual average during the course of the year. In 1901 the average pressure was also abnormally low, but not as low as in 1878. Another marked feature of the curve is the steadv diminution in the 52 THE CLIMATE OF BALTIMORE amplitude of fluctuation from 18T8 to 1900, diminishing with consider- able uniformity from nearly one-tenth of an inch to about one-hundredth of an inch. Since 1900 the amplitude has again increased. The curve 30 000 inches. TABLE VIII.-MAXIMUM STATION PRESSUKES. [In inches and hundredths.] Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year 1875 .68 .52 .51 .61 .44 .61 .57 .85 .50 .26 .70 .58 .37 .43 .48 .40 .65 .40 .32 .43 .26 .20 .37 .32 .24 .32 .30 .14 .42 .31 .13 .12 .19 .11 .22 .22 .16 .18 .16 .06 .23 .12 .20 .16 .16 .10 .13 .34 .35 .13 .24 .40 .36 .31 .48 .36 .41 .29 .75 .41 .61 .30 .51 .46 .50 .68 •61 .59 .51 .49 .56 .47 .68 1876 .85 1877 .51 1878 .61 1879 1880 .68 1881 .61 .73 .29 .30 .41 .12 .16 .22 .24 .52 .66 .57 .73 1882 .83 .68 .65 .43 ..38 .18 .23 .26 2'i .29 .51 .48 .83 1883 ..58 .73 .48 .37 .28 .40 .15 .20 .38 .56 .60 .57 .73 1884 .79 .68 .40 .16 .24 .41 .00 .24 .41 .53 .38 .59 .79 1885 .75 .41 ..38 ..50 .10 .21 .09 .15 .27 .24 .29 .70 .75 1886 .77 .49 .36 .43 .21 .18 .13 .25 .34 .41 .38 .46 .77 1887 .57 .81 .65 .55 .23 .26 .11 .19 .35 .39 .74 .83 .83 1888 .73 .57 .44 .51 .19 .16 .15 .19 .44 .31 .63 .53 .73 1889 .51 .78 .45 .39 .28 .35 .16 .18 .23 .28 .59 .73 .78 1890 .62 .47 .49 .54 .21 .28 .19 .15 .27 .24 .28 .51 . .62 1891 .40 .56 .55 .34 .26 .10 .18 .08 .27 .38 .67 .51 .67 1892 .49 .57 .37 .35 .17 .11 .36 .04 .27 .31 .31 .46 .01 1893 .27 .64 .46 .31 .23 .16 .10 .07 .19 .37 ..53 .71 .71 1894 .48 .34 .66 .40 .45 .,38 .28 .48 .30 .17 .15 .32 .06 .10 .05 .12 .38 .23 .23 .40 .59 .51 .60 .66 1895 .60 1896 .38 .39 .49 .38 .24 .12 .22 .20 .22 .36 .62 .78 .78 1897 ..59 .51 .58 .50 .28 .13 .17 .07 .27 .48 .46 .42 .59 1898 .40 .50 .52 .12 .16 .15 .20 .12 .29 .33 .42 .46 .52 1899 .84 .56 .30 .27 .20 .11 .12 .11 .29 .41 .40 .49 .84 1900 .44 .58 .47 .30 .21 .04 .08 .18 .20 .30 .52 .36 .58 1901 .61 .23 .22 .33 .03 .08 .04 .06 .33 .47 .34 .38 .61 1902 .79 .24 .32 .18 .26 .35 .11 .11 .32 .43 .38 .56 .79 1903 .42 .49 .43 .31 .34 .18 .35 .27 .30 .21 .57 .46 .57 Extremes .84 .85 .65 .55 .42 .41 .36 .34 .44 .75 .74 .83 .85 Table VIII presents the highest station pressure observed at any of the regular hours of observation for each month and for the entire year, from 1875 to 1903, together with the absolute extremes for each month and for the entire period of observation. The absolute extremes are also indicated in Fig. 9 curve '(a). The number 30.00 should be added to each of the figures in the table. also shows some suggestions of a wave of greater period. From 1885 to 1899 there seems to have been a gradual rise in pressure. The Baltimore series of observations is, however, too short to place much reliance upon the evidence of a long period of variation. IMARYLAXD "WEATHER SERVICE 53 THE AVERAGE VARIABILITY OF PRESSURE. Some interesting facts regarding the variability of pressure conditions are revealed in Talkie Til showing the departures of monthly average pressures from 18T3 to 1903. The month of greatest variability is March TABLE IX.-MIXIMUM STATION PRESSURES. [In inches and hundredths.] 29.000 inches. Jan. Feb. Mar. Apr. May June July 1 Aug. Sept. Oct. Nov. Dec. Tear 1875 ..55 .36 .36 .52 .35 .66 .61 .64 .59 .49 .52 .37 .35 1876 .49 .31 .39 .29 .48 .20 .38 .28 .43 .27 .21 .02 .18 .34 .26 .48 .36 .22 .52 .58 .46 .55 .64 .73 .64 .52 1 .40 .44 .59 .72 .64 .55 .51 .59 .72 .63 .56 .59 .66 .18 .66 .49 .65 .59 .49 .29 8.74 .54 .42 .48 .38 8.90 .45 .41 .01 .29 8.65 .57 ..59 .01 1877 .02 1878 28.65 1879 .29 1880 .26 1881 .26 .32 .no .46 .56 .54 .60 .60 .77 .50 .64 .34 .00 1882 .29 .26 .48 .31 .46 .47 .57 .56 .50 .74 .69 .60 .26 1883 .54 .65 .29 .56 .29 .60 .64 .61 .50 .44 .69 .48 .29 1884 .17 .25 .45 •1*. .57 .63 .55 .64 .66 .66 .42 .45 .14 1885 .40 .19 ..53 .46 .44 .43 i .61 ..52 .42 .02 ..53 .25 .02 1886 8.91 .24 .20 .29 .47 .47 .63 .59 .&i .73 .41 .44 28.91 1887 .37 .30 .35 .22 .52 .55 .62 .64 .54 .48 .39 .32 •to 1888 .62 .44 .34 .48 .59 .«3 .58 .35 .60 .42 .38 .85 .34 1889 .13 .47 .35 .22 .63 .59 .60 .66 .47 .44 .40 .54 .13 1890 .63 .54 .38 .42 .54 .67 .64 .64 .82 .35 .60 .36 ..33 1891 .12 .37 .37 .38 .66 .53 .61 .56 .76 .00 .28 .48 .12 1892 .14 .17 .16 .52 .44 .50 .63 .59 .59 .47 .47 .41 .14 1893 .07 .12 .27 .35 .26 .48 ..56 .39 .54 .01 .60 .43 .01 1894 .22 .39 .50 .34 .40 .52 .58 .66 .55 .20 .54 .23 .20 1895 .17 .22 .35 .22 .48 .64 .56 .61 .60 .52 .35 .31 .17 1896 .45 8.81 8.99 .57 .55 .43 .59 .70 .54 .00 ..50 .61 28.81 1897 .54 .43 .25 .47 .40 .57 ..50 .00 .70 ..54 .28 .36 .25 1898 .29 .26 .69 .33 .36 .52 .61 .68 ..59 .52 .29 .11 .11 1899 .18 .22 .14 .32 .39 .14 .44 .14 .04 .22 .21 .19 .48 .54 8.97 .12 .56 .30 .42 .58 .66 .57 .63 .34 .68 .46 .66 .64 .60 .75 .72 ..59 .57 .55 .54 .64 .35 .72 ..53 .43 .31 .19 .25 .30 .13 .30 .38 .20 .03 1900 .14 1901 28.97 1902 .12 1903 .21 .17 .46 .21 .76 .50 .49 .61 .62 .48 .52 .26 .17 Extremes 8.91 8.81 8.99 8.97 .26 .34 .46 .35 .18 8.74 8.90 8.65 28.65 Table IX presents the lowest station pressure observed at any of the regular hours of observation for each month and for the entire year, from 1875 to 1903, together with the absolute extremes for each month and for the entire period of observation. The absolute extremes are also indicated in Fig. 9 curve (e). The number 29.00 should be added to all fractional numbers in the table. with an extreme limit of 0.431 inch and an average variability of 0.058 inch. The month of most uniform pressure is June with an extreme amplitude of 0.162 inch and an average variability of 0.029 inch. The montli of December sbows a remarkable freedom from extreme fluctua- 54 THE CLIMATE OF BALTIMORE tions, being next to June in this respect, while at the same time exhibit- ing a fairly large average variability. In the following table the varia- bility of the average monthly and annual pressure at Baltimore is shown by means of the average plus or minus departures from the normal monthly and annual values, the greatest plus and minus departures and the extreme variations of the monthly and annual values. J VI \ M J J K 5 o M 3 J Inches 1 , (a) s / -^ ^ s X V / ' ' \ ^ \ > (b) K, s / " s s. / ^ -" \ / ^ .50 ^ / ^ t N "• ^ ^ / / ff ■>«. / ^ ■ (<=) - >. .- — ■* 30.00 ■4 0^ ' M K ... r ^ _ —i .' s r* N / , L s, .50 / / ' S. ^ S f / (dl . . \ »« ^ ■" — ,>• r^ \ / \ / \ r ^ J V \ (ej / 29.00 f \, / J / k. ^ \ ' \ / \ / ^ , y y /j f .50 _ _ _ Inches 31.00 (^) fb) 30,00 w 2900 Ce) .50 Fig. 9. — Monthly Means and Extremes of Pressure. (See Tables VIII, IX and X.) VARIABILITY OF PRESSURE CONDITIONS AT BALTIMORE. (Expressed as departures from the normal values in thousandths of an inch.) Jan. Feb. Mar. 174 Apr. May June July Aug-. Sept. Oct. Nov. Dec. Year Greatest plus departure (+). 88 187 106 142 84 74 67 78 Ill 141 89 38 Greatest minus departure (— ). 148 206 357 188 126 78 106 108 96 171 146 77 72 Extreme am- plitude. 236 393 431 294 268 162 18 175 174 282 287 166 110 Average de- parture /' + \ 59 .5.5 58 54 45 29 30 33 32 .55 59 46 15 MARYLAXD WEATHER SERVICE 55 The negative departures are far more marked than the positive. Only in May, June and November have the plus departures exceeded the minus. This contrast is particularly strong in the figures representing TABLE X.-SUMMARY OF PRESSURE CONDITIONS. iln inches and thousandths.] Monthly Means. Mean Monthly and Annual Extremes. Absolute Extremes. Highest and Low- i . 1875 -1903. 1875-1903 Means. est Means as De- 1 ^ partures from ! i -^ 1873 to 1899 Normal. . .:: 1873-11)03. ii, , .3 Average of Extremes. Departures from Nor- mal. o til a 2 Max. & Min. a , n 21 Zt ^ |P3 ^ . tx t>l 1 ^i CO . 1 ■ - ' s "^ "S m ^-» d 1 r" 1-5 » s 5 c 2 § if £ 30.000+ 38. 000 4- 29.000+ 1 1 ^.COn-l- 29.000+ 1 1 Jan .995 -h . 088 1 875 - . 1 48 1 893 . 236 . 0.59 .572 ..307 J+.577— .688 1.265 .843 1899^ .908 1886 1.935 Feb. .. .968 . 1 87 188:3 - . 206 Vm. 393 . 055 .557 .298 .589 -.670 1.259 .849 1876; .809 1896 2.040 Mar. .. .899 .174 1S98-. 2.57 1881 .431 .0.58 .443 .291 .544 -.608 1.152 .6.50 1887] .993 1.S96 1.6.58 Apr... .877 .1061,S88 — .1><81S78 .294 .054 .353 .363 .476 -.514 .99C .550 1887! .969 1901 1.581 May... .852 .142 I903-.126 1901 .268 .045 .245 ..50 J .393-.a5C .743 .424 1879 J. 257 1893 1.167 Juno .. .851 .084 1884 - .078 1881 . 162 .029 .191 ..=^42 ' .340 — .309 .64S .414 1884 1.345 1902 1.069 July.. .850 . 074 1 892 — . 1 06 1 884 . 1 80 . 030 .151 1 .594 : .301 -.256 .557 .3,59 1892 1.4.57 1900 .903 Aug. .. .869 .0671876 -.108 1878 .175 .OSJ .159 .608 .290 -.261 ..551 .342 1880 1.3,50 il8S8 .993 Sept... .946 . 078 1 892 - . 0!'6 1 876 . 1 74 . 032 .293 1 .581 .347-. 365 .712 .440 1888 1.181 1876 1.259 Oct.... .943 .11 11899 -.1711.'-90.'J82. 0.55 .381 .435 ' .439 — .507 .946 .753 1879 .739 il878 2.013 Nov. .. .976 .141 1880 — .1461885 .287 .059 .498 .420 .522 -.5.56 1.078 .740 1887 .902 1878 1.8:« Dec. .. .9^8 .089 1877 -.077 1878 .166.046 .540 .353 .552 -.636 1.188 .830 1887 .648 1878 2.183 Year . . .917 .038 1883 -.072 1878 .110 .015 .691 .120 .777-.797 1.574 .849 1876 .648 .878 3.301 Table X presents a summary of pressure conditions, derived mostly from the preceding tables. It shows the mean monthly values; the highest and lowest mean values, with year of occurrence and range; the mean variability of the monthly means; the mean monthly and annual extremes, with their respective departures from the normal value, and their ranges; the amount and year of occurrence of the absolute extreme values, and the ab- solute range of pressure for each month and year. Much of the data con- tained in this table is also shewn graphically In Fig. 9. the annual departures. The extreme amplitudes, or the differences be- tween the highest and lowest average monthly values are about six times larger than the average plus or minus departures. This ratio is remark- ably constant throughout the year, excepting the months of December and Januarv when the ratio falls to 4. 56 THE CLIMATE OF BALTIMORE EXTREMES OF PRESSURE. The extreme range of the barometer at Baltimore from 1871 to 1903, according to the official records of the United States Weather Bureau, is 2.20 inches. The highest ohserred reading, namely 30.85 inches, oc- curred on February 5, 1876, and the lowest, 28.65 inches, on December 10, 1878. The barometer seldom falls below 29.00 inches in the Middle Atlantic states; since 1875 a lower reading has been observed at Balti- more but once in each of the months of January, February, March, April, October, November, and December, and never in the months of May to September. The very low pressures occur only in connection with a severe cyclonic storm of the winter type, or in connection with tornadoes. In the center of the extremely limited area of a tornado the barometer has fallen to 27.00 inches or less for a fcAV minutes, but Baltimore has fortunately been visited but two or three times in the past 30 years or more by these fierce and destructive storms, and then only by a compara- tively mild type. Of the seven occasions referred to above on which the barometer fell below 29.00, three occurred in the year 1878, a year remarkable for low pressures, two in 1896, one in 1886, and one in 1901. The abnormally high pressures likewise occur in the winter months only, the most marked of them in connection Avith the intenser types of cold waves. The highest observed reading of the barometer occurred during the cold wave of February, 1876, when the pressure rose to 30.85 inches. A detailed record of the highest and lowest observed pressures for each month and for the year is given in Tables Till and IX on pages 52 and 53. For a summary of averages and extreme conditions of pressure reference may be made to Table X. In Figure 9 some of the chief features of this table are graphically shown. TEMPEEATUEE OF THE ATMOSPHEEE. Introduction. — There are certain factors which, in the long run, de- termine the average temperature of every locality. Of these the latitude, the position of the place with reference to large land and water areas, the height above sea-level, the nature of the soil, and other factors of minor importance are constant and tend to give to a place a fixed mean temper- M AH Y LAND WEATHER SERVICE 57 ature. Other factors, as wind direction, amount of cloudiness, etc., vary greatly from day to day and from season to season, and tend to produce a variable mean temperature. In some localities, within the tropics for example, these variable factors become fairly constant, and enable us to determine the average temperature by means of a comparatively short period of observations. In others the variable factors are large, as in the temperate zones, especially in the usual paths of cyclonic disturbances. In such regions a long series of observations is often necessary to de- termine the average temperature conditions to within 1° or less. In the smaller islands of the tropics five or six years of carefully made temper- ature records will yield an annual mean value with a probable error not greater than one-tenth of a degree. In the temperate regions an equally accurate annual mean niay require observations covering a period of 50 to 100 years. In the long run the effect of the variable climatic factors is eliminated and a given locality secures a position upon the normal tem- perature chart, due to its geographical and topographical position and the nature of the soil. Baltimore occupies a middle position on the climatic chart with average annual and summer temperatures 3° or 4° below tlie average for the entire globe, and a winter temperature about 10° below. The city lies between a region of equable temperatures, the ocean, and one of great variability, the nortli continental area. The factor to which is due most of the changeable character of the weather of Baltimore, causing a variability greater than is its due on account of latitude, is its proximity to the great transcontinental storm paths. Baltimore is within the influence of the barometric depressions which continually pass from the northwest, across the Lake region and the New England states, and which are accompanied by rapid changes in wind direction from the warm southerly to the colder west and northwest winds. Average Temperatures. For purposes of comparison it is essential to have a standard of refer- ence. In discussing temperature conditions the standard of value is usually assumed to be the average daily temperature. This daily average is derived from observations made hourly throughout the day and niglit. Approximate averages are obtained from two or more observations made at 58 THE CLIMATE OF BALTIMORE such hours of the day as to give a value more or less closely agi'eeing with that derived from hourly observations. Experience has shown that fairly accurate daily averages may be obtained by noting the temperature at 7 a. m., 3 p. m., and 9 p. m., or 7 a. m., 3 p. m., and 11 p. m., or 10 a. m., and 10 p. m., or 8 a. m., and 8 p. m., or from the highest and lowest tem- peratures recorded during the day. In later years automatically record- ing instruments have largely displaced direct observations permitting us to obtain a daily mean temperature to any desired degree of -accuracy with comparatively little personal attention. Monthly, seasonal, and annual means are in turn derived from the daily means. In the discussion of temperatures in succeeding pages we must not lose sight of the nature of average values. They are not real values in the sense of occurring in nature. When we say that the average temperature on the 4th of July in Baltimore is 79°, we mean that by adding together the hourly temperatures on the 4th of July for a great many years and dividing by the total number of hours we obtain the value 79°. This may never have been the real average value for the day on any 4th of July. It is simply an arithmetical mean ; the real temperatures of the day may have had any value from 60° to 100° or more. Average values are sometimes very misleading if sole reliance be placed upon them to characterize the temperature conditions of a day or a season. Two seasons may have the same mean temperature and yet be totally different in character. The summer of 1898 left the impression of an unusually warm season. The official records show a temperature very near the average of a period of thirty years (76°). The average may be obtained from any one of a large series of combinations, and our general impression of the character of the season will be determined by the particular combination of weather experienced. The temperature may lemain uniformly near the average throughout the season; there may be excessively high temperatures of short duration combined with longer periods of moderately low temperatures; or excessively low temperatures combined with longer periods of moderately high; or there may be very high combined with very low temperatures, etc. All of these combina- tions may produce a " normal " average, but the personal effect will be different in each instance, and give rise to a variety of opinions as to the MARYLAXD WEATHER SERVICE 59 character of the season. Disregard of such considerations frequently leads to unfavorable criticism of official records. Hence the figure rep- resenting the average temperature of a period is not of itself a safe cri- terion of the temperature conditions ; the variability of temperature is an essential factor in revealing the character of the period. The XoRiiAL Hourly Temperature. The most familiar, and at the same time most regular, feature of changes in the weather is the rise and fall of temperature between sunrise and sunset. Like the pressure change it is most regular in the tropical regions, and diminishes in amplitude with distance from the equator until it disappears by merging into the annual change within the Arctic Circle. As the amplitude of variation depends very largely upon the character of the surface upon whch the rays of the sun fall, there are marked de- partures from the general law of decrease in amplitude with increased latitude. Over a water surface the daily changes are small: over the interior of the continental areas, especially over sandy soils and in a dry atmosphere, they are enormously increased. The difference between the highest and lowest temperature recorded during an average day a few feet above the surface of mid-ocean is not ordinarily more than 1° or 2°, owing to the relatively large absorbing power of water, and to the large quantity of heat employed in the conversion of water into vapor — the latent heat of evaporation. The surface of the soil, especially when unprotected by vegetation, is rapidly warmed by the sun's rays and attains a high temperature, owing to its comparatively low specific heat. The at- mosphere above such surfaces is in turn heated by contact and bv con- vection currents. In consequence the difference between midday and night temperatures over land surfaces is many times larger than over water surfaces. For any given locality the diurnal variation also varies with the season of the year, following the changes in the altitude of the sun, and hence is greatest in the summer months and least in the winter months. The Baltimore hourly observations of temperature extend over a period of ten years, affording ample data for determining all phases of the L 80 / \ \ k«- ^ \ V . 1 s ■ / s Ti i N / \ '•>< f N"2 ^ y ■" s. ^ m t t; 64 / N / \ J ^ _, \ t f ^ 'S v 60 / / ^ ''" •> s> N k r / / \ V \ d 1 / f / \\s Nfe f / J Sir- S S - / J ^ / f \ s 55 ; i / / / Va S k ^.^ y > y ^ ^ ^ ^ z^ / / / s S ^ V / V V V ^ ^ / / S iO -. 1^ ^-. l/ \3^i / J^^ f <>» / 33 1 .^ / N s. / ^ 3S / N ^ y s sr - , / V, ^ X '<5 fe-^ / -O^/^ / 30 1 P- •ar *f _| Fig. 10. — Mean Hourly Temperature. (See Table XI.) for each hour of the day, for each month, and for the year are given in Table XI, and in Fig. 10 and Fig. 11. The details of changes in temperature from hour to hour are best shown in tabular form from which the exact value for each hour may be readily taken. The graphic form, however, presents advantages in afford- TABLE XI. -MEAN HOURLY TEMPEBATUEE. Hours. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 1 A. M 31.4 31.0 30.6 30.2 29.8 29.6 29.4 29.7 30.6 32.0 33.7 29.9 29.4 28.9 28.5 38.1 27.9 27.8 28.4 29.7 31.3 33.1 34.6 :i5.8 36.9 37.2 37.2 36.5 35.3 34.4 33.6 32.8 32.2 31.6 31.1 40.1 39.4 38.8 38.3 37.9 37.6 37.8 38.8 40.5 42.2 44.0 45.7 47.0 48.3 48.7 48.8 48.0 47.0 45.6 44.5 43.4 42.5 41.7 40.9 49.1 48.3 47.6 47.1 46.5 46.4 47.4 49.5 51.7 53.7 55.6 57.1 58.2 59.0 59.6 59.6 59.0 57.9 .56.2 .54.9 53.6 52.5 51.6 50.5 59.6 58.8 58.1 57.5 57.0 57.4 59.0 61.3 63.4 65.2 67.1 68.4 69.5 70.5 70.7 70.7 70.2 69.0 67.0 65.2 63.7 62.6 61.5 60.9 68.0 67.2 66.4 65.8 65.2 66.1 68.2 70.5 72.8 74.6 76.3 77.5 78.7 79.6 79.9 79.8 78.9 77.6 75.7 74.1 72.6 71.3 70.2 69.2 72.9 72.2 7l'.4 70.8 70.3 70.6 72.4 74.7 77.2 79^1 80.9 82.4 83.4 84.0 84.2 84.1 83.1 81.8 80.1 78.3 76.6 75.6 74.5 73.7 71.3 70.7 69.9 69.3 68.7 68.8 70.4 73.0 75.4 77.7 79.8 81.0 82.1 82.8 82.9 82.6 81.6 80.5 78.8 77.0 75.6 74.4 73.2 72.2 65.2 64.6 63.8 63.1 62.5 62.2 63.1 65.7 68.0 70.5 72.6 74.2 75.3 76.1 76.4 76.3 75.1 73.4 71.6 70.2 68.6 67.5 66.8 65.7 53.9 53.3 52.7 52.1 51.7 51.3 51.6 53.6 56.0 58.4 60.6 62.3 63.4 64.2 64.4 64.0 62.8 61.2 .59.5 58.1 56.9 55.9 55.0 54.2 44.1 35.0 43.7 34.5 3 43.2 ' 34.2 i 43.7 . 33.8 42.3 as. 4 6 42.0 33.0 41.9 32.9 8 9 42.8 33.3 44.8 1 34.5 10 11 46.7 36.0 48.6 37.9 Noon 1 3.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'. 4 5 6 8 9 35.0 36.0 36.8 37.2 .37.0 .36.2 a5.4 34.5 34.0 33.0 32.8 32.1 31.8 50.2 39.4 51.4 40.7 52.0 41.6 53.2 42.0 51.6 41.5 50.4 40.5 49.2 39.5 48.1 : 38.6 47.3 37.9 46.4 37.1 10 11 45.7 36.5 44.9 35.9 44.2 35.2 32.9 32.2 42.8 53.0 63.9 72.8 77.3 75.8 69.1 57.4 46.5 36.9 MEAN HOURLY TEMPERATURE. Spring. Summer. | Autumn. "Winter. Year. 1 A. M 49.6 48.8 48.2 47.6 47.1 47.1 48.1 49.9 51.9 53.7 .55.6 57.1 .58.2 59.2 59.7 .59.7 .59.1 58.0 .56.3 .54.9 53.6 52.5 51.6 50.8 70.7 70.0 69.2 68.6 68.1 68.5 70.3 72.7 75!l 77.1 79.0 80.3 81.4 82.1 82.3 83.2 81.2 80.0 78.2 76.1 74.9 73.8 72.6 71.7 54.4 53.9 53.2 52.6 52.2 51.8 53.2 54.0 56.3 58.5 60.6 62.2 63.4 (U.l 64.3 64.0 62.8 61.3 59.7 58.5 57.3 56.4 55.6 54.7 32.1 31.6 31.2 30.8 30.4 30.3 ai.o 30.5 31.6 33.1 34.9 36.3 37.5 38.4 38.8 38.6 37.7 36.7 a5.8 35.2 34.3 33.8 dii.-Z 33.7 51.7 o 51.1 3 4 50.5 49.9 49.4 49.4 - .50.2 8 51.8 9 53.7 10 55.6 11 .59.0 1 60.1 *> 61.0 3 61.3 4 61.1 60.2 6 59.0 57.5 S .56.3 9 10 11 .55.0 54.1 53.2 52.5 53.2 75.3 .57.7 34.0 55.0 Table XI shows the mean temperature for each hour of the day, based on the continuous record of a Richard thermograph for the ten-year period ending December 31, 1902. The thermograph record was corrected daily by direct observations of a mercurial thermometer at 8 a. m. and 8 p. m., and by the readings of a maximum and a minimum self-registering thermometer. The annual mean (55.0°) is the average value of over 87,000 hourly observa- tions, and may be regarded as a true normal value for the period covered by the observations. The same results are graphically shown in Fig. 10, for January, April, July, October, and the year, and in Fig. 11, for all the months of the year. 62 THE CLIMATE OF BALTIMORE ing a readier means of observing the relative changes from hour to hour and the distribution of temperature within the day, the month, and the year. The method of presentation employed in Fig. 11 is particularly well adapted to a rapid survey of the hourly and seasonal distribution. In construction it resembles the maps prepared for showing the varying topography of an area. In place of the meridians of longitude and par- allels of latitude to arrive at the geographical position of a given locality, we have vertical lines to represent the hours of the day and horizontal lines for the months of the year, the intersection of which gives us the Fig. 11. — Isopleths of Hourly Temperature. Fig. 11 shows the average distribution of temperature throughout the day and j ear, based on observations of ten years of hourly readings of the thermograph. The hours of the day are indicated by the upper line of figures, Avhile the marginal letters indicate the months of the year. The line enclosing the area of lightest shading defines the time of occurrence of the lowest temperature of the year and day ; rise in temperature is indicated by increase in the intensity of .shading. The diagram indicates that the lowest temperatures of tlie year occur in the early morning hours of January and Februarj', and that the highest occur in the eai'ly afternoon hours of Julj% based on the average of a long series of years. The curved lines show the liours of the day and the months of the year when the average read- ings of the thermometer are equal : these lines are called chrono-isotherms, or. isopleths of temperature. The dotted lines marlicd S.R. and S.S. show the time of sunrise and sunset. See also Table Xi. time sought. In place of the contour lines, or lines of equal elevation of the topographic map, we have lines of equal temperature (or isopleths of temperature as they are called when used in this manner) projected upon the plane of the time lines. The rapid detection of the diurnal and annual distribution of temperature is further facilitated by means of a system of shaded areas, increase in the intensity of the shade signifying :MARYLAXD "WEATHER SERVICE 63 increase in temperature. Consulting Fig. 11 we find that for Baltimore a temperature of 8-i° is limited, under average conditions, to the hours from 2 p. m. to 4 p. m., during the month of July; that the temperature of 75° occurs, on the average, from June, between the hours of 10 a. m. and 7 p. m., to September, between 1 p. m. and 5 p. m., etc. In the winter months the line of 32°, or freezing weather for example, is limited, on the average, to the months of January and February from midnight to 10 a. m. We see that the average summer temperature of 76° extends from June to September during the middle hours of the day, while the average winter temperature of 36° is confined to the night and morning hours of December, January and February, and to a few of the early morning hours of March. The lowest temperature of the day occurs, on the aver- age, just before sunrise and hence varies with the advance and retreat of the sun. The time of occurrence of the highest temperature varies less with the season, occurring throughout the year between 3 p. m. and 4 p. m., excepting the month of November when the highest temperature of the day occurs at about 2.30 p. m. The diurnal variation of temperature is represented by a simple curve which rises steadily from a minimum point just before sunrise, attains a maximum in the early afternoon hours, and then descends without inter- ruption to the early morning minimum. In this respect it differs from the curve representing the diurnal variation of the barometer which, as we have seen, has a double period, Ts-ith primary and secondary maxi- mum and minimum points. Phases of the Diurnal Variation. The principal phases of the diurnal variation of temperature are pre- sented in Table XII, containing a summary of the average time of occur- rence of the minimum, the maximum, and the mean temperature for the day, and the varying interval between the occurrence of the minimum and maximum points. In the months of May and June the lowest tem- perature of the day occurs at 5.05 a. m., 75th meridian time, which is six minutes faster than Baltimore local time; the time advances steadily to 6.50 a. m. in January, returning again to 5.05 a. m. in May. Tlie maximum of tlio day sliows less variation in time of occurrence. From 64 THE CLIMATE OF BALTIMORE January to May it remains at about 3.25 p. m., then occurs successively earlier in the day until November, when a maximum is attained at 2.25 p. m. It seems rather remarkable that the maximum temperature of the day should appear earliest in the month of November. The average temperature occurs first at about 9 a. m. in the summer months and at 10.30 a. m. in the winter months, and again between 8.30 p. m. and 9.00 p. m. in summer, and about 10.00 p. m. in the winter months, excepting v^ "-^ "~~~ — " Max .. 1. Mean , ■ ■— ^ MiN. ■ _,^ -^^ _ Fig. 12. — Principal Phases of Diurnal Variation of Temperature. Fig. 12 shows the time of occurrence of the highest and lowest points indicated by the thermometer on an average day for each month ; also the morning and afternoon hours when the mean temperature of the day is most likely to occur. See also Table XII. December, when it occurs as early as 9.20 p. m. The amplitude of varia- tion, or the difference between the daily maximum and daily minimum temperature, is greatest in the month of June (14°. 7) and is smallest in the month of Januarv^ (7°.0) . (See Fig. 12.) The temperatures thus far discussed are average values for a period of ten years. When we examine into the time of occurrence of the prin- cipal phases of the diurnal march of temperature more closely we find MARYLAND WEATHER SERVICE 65 a wide divergence from the average time of occurrence as recorded in preceding paragraphs. The limits of variability in the average time for a single month are shown in the following tabular statement contain- ing the hour and the frequency of occurrence of each phase in each month of the ten-year period. TABLE XII.— TEMPERATURE PHASES. Minimum. Fre- quency. 5' 6 7 8 January February . . March April May June July August September ] 2 October November I 1 December Tear. 4 9 4 4 6:50 6:30 6:15 5:«) 5:05 5:05 5:10 5:25 5:50 6:05 6:25 6:40 5:50 1st Mean. Fre- quency. 9 10 11 5 5 6 3 7 7 3 9 1 9 1 9| 1 4| 6 4 6 l| 8 6 5 5 9:40 Maximum. Fre- quency. p. m. 1 2 3i 4 5 7 5 5 4 41 6! 5 7 4 21 5 9 2 II 4. 3:10 2nd Mean. Fre- quency. p. m. 8! 91011 1 9'.. 2 8., 3 7., ih-m. 8 4 1 9:50 3 10:00 9:40 9:30 8:50 8:50 8:35 8:50 8:40 8:35 8:55 9:20 8-30 9-00 9-10 9-50 10-20 10-05 9-55 9-20 9-00 8-35 8-00 8-20 9:00 9-20 Table XII indicates the average time of occurrence of the lowest and highest temperature of the day for each month and for the year; the morning and the afternoon hours when the mean temperature of the day is most likely to occur; the frequency of occurrence of these phases at given hours; and the average number of hours between the occurrence of the highest and lowest temperatures of the day. The values are based on hourly observations for the ten-year period from 1893 to 1902. See also Fig. 12. The January minimum may occur from 5 a. m. to 8 a. m., the normal time being 6.50 a. m. In May, June, and July the minimum occurs with great regularity at about 5 a. m., and in September and October at 6 a. m. There is more uniformity in the time of occurrence of the max- imum temperature of the day; this does not vary greatly from the hour of 3 p. m. at any season of the year. The earliest occurrence of the maximum is in the month of November. The time interval between the 66 THE CLIMATE OF BALTIMORE minimum and maximum of the da}' increases steadily from the winter months to the summer months. Beginning with eight hours and thirty minutes in January, the interval reaches a maximum in May when it amounts to ten hours and twenty minutes, then decreases regularly to eight hours in November. This difference in time is due mostly to variations in the time of occurrence of the minimum temperature. DiuRXAL Variation as Affected by Clouds and Eain. In considering the diurnal variation of temperature in the preceding paragraphs the character of the day does not enter into the problem. The average values given include all days for a period of ten years. The amplitude of variation of temperature is manifestly largely dependent upon the presence or absence of clouds. On a cloudy day the sun's rays are largely absorbed by the cloudmass and comparatively little of the sun's heat-rays reach the earth's surface directly. To discover to what extent the normal daily variation is affected by the character of the day the diurnal variation of temperature has been determined for selected days in each season. For this purpose the days were grouped as clear, cloud}'', and rainy. Days were regarded as clear during which the per- centage of sunshine exceeded 90 per cent of the possible amount for the day. They were considered cloudy when the sky was overcast the entire day. A day was considered rainy when rain fell for more than four hours, not necessarily consecutive. Each group included approximately 100 days, selected from all seasons of the year. A further restriction was imposed by excluding days with a moderate or a high wind, as it was desired to eliminate the effect of wind velocity upon the diurnal variation in this problem. The results of the above classification are shown in Fig. 13, in which some interesting and instructive relations are revealed. The ampli- tude, or difference between the lowest and highest temperature of the day, is manifestly greatest on clear days, with a maximum in the spring months. Cloudiness reduces the daily range of temperature to less than one-half of that on a clear day. On a rainy day the difference between the maximum and minimum is reduced to 2° or 3°, equivalent to about one-fourth the range on a clear dav in winter and to about one-sixth MARYLAXD WEATHER SERVICE 67 the range in spring, summer, and autumn. The principal phases of the diurnal march of temperature do not materially change in the summer and autumn months. The minimum occurs approximately at sunrise and the maximum of the day in the early afternoon hours. There is a marked deviation, however, from the normal conditions on rainy days in autumn and winter. After attaining the maximum for the dav it is 1 2 ( 4 ■ « 1 i > » 9 1 uoa _ _ ♦ > r i 9 9 1 r 2 1 - - - ' I 1 y \\ ■ 1 / / \ \ \1. 1 / / \ 1 Ov^v \ 1 ^^ :jtjL \ ' "^ /' y n: Sf. /I '' / ■ \ -1. '~ r^ i/ iS ^H 1 i \f - 1 -- -- - - - 7- - - - - - - - - - ■t\ 50 / i» w II L t' , . — ' ' ~~ p> -^ Ft °^--C^*fe-/ <>. //- ; ' ' -i~-r 1 \ ■ ; , :y^, 1 i \-> 1 /i : 1 Sp ri h iS ^■':/-]-V-\ [_ N^y, i ! 1 ^ rn 1 ! 1 1 y \ / / \ ' i / -^ ~H -V, N i ' U V ■ -^ \ A mr i^jL 1 - - - -- - --- /i-< - - - - -- - '^ c / / \ ^■^f' ii ^ \ \ ^ > k'«a it- '^ \ 50 s 1 \ S' 1 s 1 1 II II m n 1 1 V ^ 1 1 _ 2 J 5 ( 9 10 K on J ■! J < r s i o 1 12 i i ' i 1 ! 1 : >^ i ■ ' 1 M i y' ■ 1 s 85 y 1 ^ ! / ! 1 1 \ / / / , 7 j\f-i. 1 ^ 75 J ] 1 /' \ \ y 1 / " — ^^ N / 1 ft / 1 ^^"^ J <5 ^' ^ ,t^ V ! ; ^^ ! jSS" .L^.JAj . _j ^ lJ Xi I 70 N^' :/ , 1 < ' ' \ ' ^- — -^' \l X s ^u mmei- V < ^ \ / r-" 1 1 ' i=?- \r \;_^>^ _L ■ 1 , i' 1 dh 38 — 1 1 1 1 1 M \ • 1 — '_ \ ^ — — — -^ 1 1 : . 1 ■ : ') 35 r,:^^^^,^...-.-/-.. ^ h - - - -^ ^ / s i 1 1/ \ ^t^ 1. 1 ■V, 1 y ■ V ^J ^' J 1 / N N 5-1 t V ^\^^ f» . s f / 23 " :_ Fig. 13. — Effect of Cloudiness and Rain on the Hourly Variations of Temperature. maintained until nearly midnight. One of the most interesting facts revealed in the diagrams is the relative position of the curves for clear, cloudy, and rainy days in the difTerent seasons. In winter the clear day has a temperature decidedly below the normal for the season, while the cloudy and raiiiv (Uiys are well al)ovo tb.e normal. In spring the clear day lias altout the normal temperature: the cloudy day is far above tbe normal : the raiin- dav is decidedlv below tlie normal. In summer the clear 68 THE CLIMATE OF BALTIMORE day is decidedly warmer than the average for the season ; the cloudy day is about normal; the rainy day is much below the normal. In autumn the clear day is somewhat below the average temperature, the cloudy day is about normal, and the rainy day is well above the normal. These dif- ferences in temperature depending upon the extent of cloudiness and pre- cipitation are in some cases very large. In spring the early morning temperatures may be 10° to 15° lower with a clear sky than with an overcast sky. In autumn there is quite as marked a difference between a clear and a rainy day in the early morning hours. In the summer months the midday temperatures may be reduced 10° to 13° by an over- s' 30* Fig. 14. — Effect of Snow-Covering on the Hourly Variations of Temperature, (a) A normal winter day. {b) Average of days with snow on the ground. cast sky, and 15° to 20° during a rain. During the autumn an overcast sky will maintain the average temperature of the day 6° to 8° above that of a clear day. MEAN HOURLY TEMPERATURE ON CLEAR, ON CLOUDY AND ON RAINY DAYS. Winter. Normal Temperature 34.0° Clear days (Departures) —5.2° Cloudy days " i +0.6° Rainy days " +0.5° Spring. 53.2° 1 .70 +3.5° -6.2° Summer. 75.8° +3.4° 4 2° -:!o° Autumn. 57.7': -0.4° +0.9° ilARYLAXD WEATHER SERVICE 69 Effect of a Sxow Coverixg. To determine the effect of a snow covering upon the diurnal variation of temperature the average hourly temperature was calculated for all da3's within the period of ten years from 1893 to 1902 upon which the ground was covered with snow to a depth of half an inch or more. The values for the entire season are shown in the accompanying table and in Fig. 14 in comparison with the normal temperatures for the winter season. The two curves are identical in form and run parallel through- out their extent, but the days with snow on the ground were uniformly about 10° below the normal temperature for the winter months. HOURLY TEMPERATURES ON DAYS WITH SNOW ON THE GROUND. Hours: A. M. 1 2 3 4 5 6 7 8 9 10 11 Noon. Winter normal.. With snow on ground Departure below normal 32.1 21.8 10.3 31.6 21.2 10.4 31.2 20.8 10.4 30.8 20.4 10.4 30.4 20.0 10.4 30.2 19.5 10.7 30.0 19.3 10.8 30.5 19.5 11.0 31.6 33.1 20.7 32.6 10.9 10.5 34.9 24.6 10.3 36.3 26.0 10.3 Hours: P. M. 1 2 3 4 5 6 7 8 9 10 11 Mid- night. Means. Winter normal.. With snow on ground Departure below normal 37.5 27.3 10.2 38.4 28.3 10.1 38.8 28.9 9.9 38.6 28.9 9.7 37.7 27.8 9.9 36.7 26.8 9.9 35.8 25.7 10.1 35.2 24.9 10.3 34.3 24.0 10.3 33.8 23.4 10.4 33.2 22.5 10.7 32.7 23.0 10.7 34.0 23.6 10.4 The temperature is lowered during the night by the intenser radiation from a snow surface ; it is prevented from rising during the day because much of the heat of the sun which would otherwise go to warm tlie atmosphere is spent in melting and vaporizing the snow. The air tem- perature is likewise reduced by the snow preventing the communication of heat from the ground by convection. As observations show that tlie difference between the normal hourly winter temperature and the hourly temperature over a snow-covered ground is practically constant through- out tlie day and night, the daily range is neither increased nor decreased by the presence of snow. The low average temperature of the winter of l!Hi;;-l!J01 was doul)tlos.s largely due to the exceptional d\iration of a snow cover. The depth of snow was not great, in the vicinity of Balti- 70 THE CLIMATE OF BALTIMORE more, but a moderate snow covering persisted during a period of time nearly double the usual length. There is some compensation in the beneficial protection afforded by snow to winter wheat and to vegetation in general by preventing the penetration of frost into the ground. The Effect of Wixd Velocity ox Temperature. Another factor which largely affects the diurnal range of the ther- mometer is the movement of the atmosphere. It is well known that in a quiet atmosphere there may be a great difference in temperature at the earth's surface and a small distance above. In the night and early morn- ing hours of winter the thermometer may register 5° or 10° lower near the ground than on the house tops ; on a hot summer's day the difference at midday may be quite as large but reversed. In either case the lower layers of the quiet atmosphere tend to take on the temperature of the ground. Such differences are particularly common in the lower-lying portions of any locality. The}'' do not occur in an active atmosphere; a breeze will quickly level any marked differences in the temperature of any neighboring strata of air by intermingling of the lower and higher layers resulting in an approximately uniform temperature. The effect of wind movement on the diurnal range of temperature may be clearly shown by classifying a large number of days according to total daily wind movement, days which in other respects have approximate!}^ similar conditions. The results of such a classification are graphically shown in Fig. 15. An equal number of clear or approximately clear days was selected in each of the months of January, March, July, and October. Those having a total daily wind movement of less than 100 miles per day were placed in one group; another group contained days with a total daily wind movement between 200 miles and 300 miles; still an- other group comprised winter and summer days with a wind movement exceeding 400 miles per day. The average hourly temperature was then determined for each group separately and a comparison made between the resulting temperatures. In each case the diurnal range of temperature is seen to be markedly lower with increase in wind movement. In the following tabular statement tlie total daily range for each condition mentioned above is given, while in the .succeeding table the hourly MARYLAND WEATHER SERVICE 71 3 6 9 Noon 3 6 9 Mot. 3 6 9 Noon 3 6 9 Mot Id ' ' ' 1 i : M ' 1 ' ' ! ' , Ml 1 j y-^-, ' ■ 1 ' i ' 1 1 , ■ a X ( ' 1 1 1 1 ' / ; i ^ 'Mil V . ' I 1 tV ■ 1 i 1 ] / X. , ' ' 1 1 / 1 1 ! ' Xh>. ' / ' ' , ' / I ; ; 'i-! 1 ! ' 1 1 / 1 ' i ! ° "V ' 1 Si;: 1 . '/ , > ; ' ' ■ ^' / / x' i [ ^/v / / ^ , ^ N >v^. / >> , xj ■ i v . i>> s\ rv- 4 L/ Sill X X\ [ xj ; V. \\^\ 1 x^ -r ,>V4^ ' ' ' ' ' 1 ' ■ ' j ; ; ' 11 1 [ t 1 in •"X 1 1 j /•^ \ ' 1 1 "SO 1 ' ; ^\^-r^\^ 1 \ ' ' 1 ' : : > \ ' ' VV X '. , ( ' / / [ ,\^ X, i // --^j NX X IX / ' ^^''**^ / ! ! ' '^S*. — .- ; / / i 1 1 ' 1 ' 1 /' / 1 ' / / , ' ' X ' '^'^^' 1 1 1 j 1 V X/ \ ' \ \ \ ' ! 1 ' ' "^^ ■ ' ' , 1 [ ' , ' i 1 M : ' 1 1 ' i 1 ■ I 1 1 ; 1 ' 1 ' * . .. . . 1 . : -cr='=^ ; ' j ^ , , • ' ; ! ' .... ' ' \^ ' 'lit -4' ' ' ' ' * ' I ! • TX 1 iX i ' ; XL ' 1 ' """"^S^ ' ■^ 'Ml i j 1 1 , TX- ■ ' : ! i ; ' , 1 '^ 1 "i '-^~. 1 1 < 1 ; i M , X-Li^ EAR 3 6 9 Noon 3 6 9 Noon 3 6 9 Mo' Fig. 15. — Effect of Wiud Velocity on tlie Hourly Variations of Temperature. (a) On days with a light wind. ib) On days with a moderate wind. (c) On days with a high wind. 72 THE CLIMATE OF BALTIMORE changes are shown for the year, expressed in terms of departures from the normal temperatures for the year. Eange of Temperature on Calm axd AYixdt Days. Total daily wind movement. Jan. March. July. Oct. Year. Less than 100 miles 14.6° 15.4° 19.5° 18.9° 16.8° From 200 to 300 miles.... 5.9° 12.0° 11.0° 8.2° 9.0° Over 400 miles Winter 5.2° Summer 5.7° 5.4° hourly temperature on calm and on windy days. (Expressed in terms of departures from the normal temperature.) Hours : A. M. 1 J 2 3 4 5 6 7 8 9 10 11 Noon. Normal Temper- 51.7° - 2.6 -3.6 -11.1 51.1° - 2.7 -3.5 -11.2 50.5° -2.6 -3.4 -11.2 49.9° - 2.6 -3.3 -11.2 49.4° -2.6 - 3.2 -11.3 49.4° - 2.6 -3.5 —11.9 50.2° - 2.6 - 4.6 -13.2 51.8° - 1.9 -5.0 -14.5 53.7° - 1.0 - 5.4 -15.7 55.6° + 0.1 - 5.7 -16.6 57.5° + 1.0 — 6.4 -17.7 59.0° + 1.7 - 6.2 -18.1 50-100 Miles (De- partures) 200-300 Miles (De- partures) Over 400 Miles (Departures).. Hours : P. M. 1 2 3 4 5 6 7 8 9 10 11 Mid- night. Means. Normal Temper- ature 50-100 Miles (De- partures) 200-300 Miles (De- partures) Over 400 Miles (Departures).. 60.1° + 2.1 — 6.4 -18.5 61.0° + 2.1 - 6.4 -18.7 61.3° + 2.3 -6.5 -18.9 61.1° + 2.2 -6.4 -19.1 60.2° + 2.1 - 6.4 -19.2 59.0° + 1.6 -6.3 -18.6 57.5° + 1.7 - 6.3 -18.1 56.3° + 1.2 — 6.2 -17.5 55.0° + 1-1 - 6.0 -17.0 54.1° + 1.1 -6.0 -16.7 53.2° + 1.0 -6.0 —16.2 52.5° + 1.1 -6.1 -16.1 55.0° + 0.1 — 5.3 —15.8 EeDUCTION TO THE TrUE MeAN TEMPERATURE. As it is often inconvenient or impossible to make daily observations of the temperature at the hours best suited to the purpose of securing an accurate average value, it is desirable to know the corrections to be applied to any selected combination of hours in order to arrive at a true average value for the day for a given locality. This can readily be done whenever hourly observations, or continuous records, have been main- tained somewhere within a hundred miles or so of the locality, pro- vided the physiographical conditions of the two localities do not differ widely from one another. In the following table the necessary correc- tions have been computed for the horizon of Baltimore for some of the ilARYLAXD WEATHER SERVICE combinations of hours of observation employed at different times within the State of Maryland. CORRECTIONS TO REDUCE OBSERVED TEMPERATURES TO THE TRUE DAILY MEAX. Hours of Observation : — (75th Meridian Time.) J (7:37 a. + 4:37 p. + 11:37 p.). . .. 4(7:00 a. + 2:00 p. + 9:00 p.).... i (7:00 a. + 2:00 p. + 3(9:00 p.) . . . . i (7:00 a. + 3:00 p. + 11:00 p.) . . . . J 7:00 a. + 3:00 p. + 10:00 p.) A (10:00 a. + 10:00 p.) J (Maximum + Minimum) H8:00a. + 8:00 p.) i(Ta. + lla.+3p. + 7p. + llp.) +0.2+0.1+0.1+0.1 —0.3—0.3—0.3—0.3—0.5 —0.3—0.4—0.4—0.4—0.3 0+0.1+0.1 -0.3-0.2 +0.5 0.4 +1.1 -0.5 -0.2 +0.4 -0.4 +1.3 +0.4 -0.3 +1 -0.6;-0. 81—1.1 -0.3 -0.1 +0.8 +0 -0.3 +0.1 +0.7 -1.2 0.7 0.4 0.3 -0.2 +0.2 +0.5 -1.3 +0.21+0.3 -0.5 -0.3 +0.3 -0.1 -0.2 0.4 -0.1 +0.3 0.1 -0.1 +0.1 0-0.3 +0.8+1.3+1.1 -1.1-1.2—1.0—0 +0.3 -0.2 +0.3 +0.1 +0.1 +0.5 -0.2 +0.4 +0.1 +0.2 0.4 +1.6 +0.5 -0.3 0.2 +0.3 -0.1 +0.3 -0.5 +1.5 +0.3 -0.3 -0.3 -0.2 +0.7 -0.5 +1.3 -0.6-0.6 0.3 0.4 -0.3 +0.1 0.2 +0.3 0.4 + 1.0 -0.9 In a system of three hours of observation the combination 7 a.m., 3 p. m. and 11 p. m., gives a mean value very close to the 24-hour ly mean, the annual average differing from the latter by only 0.1°, while the maximum departure is but + 0.4° during the month of widest divergence. During four months of the year, namely, January, February, June and December, no corrections need be applied. One of the best combinations of two hours is that of 10 a. m. and 10 p. m., which yields an average but 0.2° above the true annual mean. The maximum and minimum readings of self- registering thermometers require a correction of — 0.4° to the annual average. Considering the great convenience of one observation a day over two or more and the further advantage of showing the highest and lowest temperatures, this is the most desirable system to adopt. The United States Weather Bureau maintains an organization of about 3500 co- operating voluntary observers, all reporting daily maximum and minimum temperatures. The Hourly Eate of Change. While the temperature increases steadily from sunrise to about 3 p. m. and then steadily decreases to sunrise, the rate of warming and cooling 74 THE CLI^[ATE OF BALTIMORE has its own period which differs from that of tlie temperature itself. The temperature rises most rapidly from 8 a. m. to 10 a. m., depending upon the season of the year, and falls most rapidly from 6 p. m. to 7 p m. The hours of least change coincide with those in whieh the maximum and minimum temperatures of the day occur. In selecting a combination of hours for observation it is important to bear in mind this varying rate of change, and to avoid as far as practicable the hours of maximum rate. Consideration of this point is of no consequence when maximum and Fig. 16. — Hourly Rate of Change of Temperature. Fiif. 16 shows the extent of change in the temperature from hour to hour throug-hout the day and year. Tlie values are based on hourlj' records during a period of ten years. The houi s of the day are indicated liy the upper horizontal line of figures, and the months of the year by the marginal letters. The areas without shading show the time of day when the change in temperature is least, the heavj- black line within this area marking the time of change from falling to rising, or rising to falling temperature. The areas with darkest shading show the time of most rapid change. A falling temperature is designated by a minus sign, a rising by absence of sign, before the figure representing the amovnit of change in degrees and tenths. The dotted lines marked S.R. and S.S. show the time of sunrise and sunset. The time of most rapid rise in the temperature is between 8 a. m. and 9 a. m., the time of most rapid fall is between 7 p. m. and 8 p. m. See Table XIII and Fig. IT. minimum thermometers are employed, or when a continuous record of the temperature is maintained. The approximate time at wdiich the rate of change is greatest and least for each month and for the year is shown below in connection with the hours of maximum and minimum temper- ature of the dav. MARYLAXD WEATHER SERVICE VO TIME OF DIURXAL MAXIMUM AND MIXIMUM RATE OF WARMIXG AND COOLIXG. c i o 3:30 >> S 3:.30 05 a 3 3:00 < 3:00 P. ® t/J o 6 a 3 a c < Time of Max. temp. (p. m.). 3:00 3:30 4:00 3:00 3:00 3:00 3:00 3:00 3:00 " Min. rate (p. m.).. 3:00 3:00 3:00 3:30 3:30 3:(K) 3:00 3:(K) 3:(M) 3.00 3:00 3:1 3:00 " " Min. temp. (a. m.). T:(KI T:00 6:00 6:00 5:(J0 5:(10 5:00 5:00 6:(K) 6:00 7:00 7:00 6:00 " " Min. rate (a. m.).. 7:(H) 6;()0 6:00 5:30 5:U0 4:30 5:00 5:30 6:(M) 6:00 6:30 6;.30 5:30 " Max. rate (a. m.).. 1000 10:00 10:00 9:01) 9:(H) H-.m S;30 f<:(H) 9:U0 9:00 9:00 10:30 9:30 " *' Max. rate (p. m.).. 6:30 5:30 6:30 7:00 7:00 7:00 7;30 7:30 6:30 6:30 6:30 6:00 6:30 TABLE XIII.-MEAX HOURLY CHANGE OF TEMPERATURE. (Expressed in degrees and tenths of a degree.) Midn't tol a. m.. 1- 3 2- 3 3- 4 4- 5 5- 6 6-7 7-8 8- 9 9-10 10-. 1 11-Xoon... X^oon-1 p. m.. 1- 3 2- 3 3- 4...'.... 4- 5 5- 6. 6- 7 7- « 8- 9 9-10 10-n 11-Midn't.. — .4 — .O — .4 - .4 — .4 — .4 2 •> .,T — . t — .5 .4 .3 .3 i!o 1.7 1.7 1.8 1.7 1.3 1.2 .5 .1 1.4 - 8, i .6 .1 1.0 2.1 1.3 .8 3.0 1.9 1.5 1.1 — 1. —1.2 -1-0 - — .81 —1.1; — : 1.0 1.4 1.1; 1.1; 9, — .3 — .5 — .8 —1.1 — .6 — .0 .6 1.1 1.7 1.3 1.3 1.1 .9, 1.1 - .5 .4 1.6 2.3 2.1 1.8i 1.9 1.3 1.1 1.0 -1.3 -3.0 -1.8i -1.5i -1.1 -1.1 .6 - .8 - .8 - .6 - .6 .9 2.1 2.3 2.3 1.8 1.7 1.3 1.3 .9 .3 - .1 - .9 -1.3 -1.9 -1.6 -1.5 -1.3 -1.1 -1.0 .3 1.8 2.3 3.5j 1.9; 1.8 1.5 1.0 .6 - .1 -1.0 -1.3 -1.7 - .9 - .6 - .8 - .6 - .6 .1 1.6 3.6 3.4 2.3 3.1 1.2 1.1 !i - .3 -1.0 -1.1 -1.7 -1.8 -1.4 -1.3 -!.& -1.0 - .5 - .6 - .8 - ie - .3 .9 2.6 2.3 2.5 2.1 1.6 l.ll .8 .3 - .1 -1.3 -1.7 -1.8 -1.4 -1.6| -1.1 - .3 - .6 - .6 - .6 - .4 - .4 .3 3.0 2.4 2.4 2.2 1.7i l.l' .8 .2 - .4 -1.3 -1.6 -1.7 -1.4 -1.3 -1.0 •9: - .5 - .3 - .4 - .4 - .4 - .1 .4 1.3 1.5 1.9 1.5 1.3 .9 .4 — .5 -1.0 -1.3 -1.0 -1.1 — .9 - .4 - .3 - .1 .9 3.0| 1.9' 1.9i 1.6 1.31 .6 - .6 - .8 — .« -1.0 -1.1 - .8 — .1 — .6 .8, - .6 - .6 - .5 .0 .8 1.6 1.9 1.9 1.9 1.5 .9 .9 -1.3 -1.5 -1.2 -1.3 - .9 - .9 - .7 Table XIII shows the average amount of change in temperature from hour to hour in each month and in the year. The minus sign preceding a number indicates a fall in temperature; numbers without a sign show a rise in temperature. For example, from midnight to one a. m., the temperature falls, on the average, four-tenths of a degree in the month of January, one and four-tenths in April, and eight-tenths, on the average, for the entire year. The results are also graphically shown for each month in the year in Fig. 10, and for the year, in Fig. 17. The values are based on hourly observa- tions for a period of ten years. In Table XIII, the average amount by wliich the temperature changes from liour to hour througliout the day is rocordeil for t-acli mouth and 6 76 THE CLIMATE OF BALTIMORE for the year. These values are derived from ten years of hourly observa- tions from 1893 to the close of 1902, In Fig. 16, the values are graph- ically shown for each month of the year in terms of departures above and below a line separating the rising from the falling temperatures. In Fig. 17 the curve representing the average hourly rate of change in temperature for the year is drawn in connection with the average hourly pressure curve. As noted above in the paragraph on the diurnal variation of the barometer there is a close resemblance between these two curves, suggesting some causal connection between the diurnal warming and cool- ing of the atmosphere and pressure changes. -to \ 1 — ^ /'''T^S-s i \/A I 1 1 ^J ^ ^^_^ T^f-/-| rV-V ' — H h+- U- jU- -^ ^i,^±^^^ i^-^SU- __ _ .^V _ 2^ _±+ _ _ it - :iiii::ii-±^gii:iiii:^ii#iiiii:+iiiHg :^::==4^:^==="=====+:^+^fci::±^";^itifc - - z^.^^^ - -^t;^ "TV-f-^ /-^t+t ^.^^ :::^_4__^_^_^ _^ ^^s-h- v - ^' -^+ ^J^ J U- -U-.^^ h^^^ J7^*^- == =^=F =i+=T= =" ==d==4=r--^-'f =^ 4+ "=1^=4="=n^4r"'4'4l?r' " . 1_^ i 1 1 M i M ' 1 — ' — — ___ 1 . . ^ ' - . . . ..... 1 . — . — , — _ +015 Fig. 17. — Curves Representing the Average Hourly Pressure (a), and the Hourly Rate of Change in Temperature (b), for the Year. Mean Daily Temperature. Just as the daily change in altitude of the sun causes a daily rise and fall in temperature, so the annual variations in altitude give. rise to an annual rise and fall in temperature. In the. diurnal period, the highest temperature is attained about three hours after the sun reaches the mer- idian; in the annual period the maximum temperature is reached from three to four weeks after the sun attains the greatest elevation. While the lowest temperature of the day occurs about sunrise, the minimum for the vear lags four to five weeks behind the time of lowest seasonal altitude ilAKYLAXD WEATHER SERVICE 77 of the Sim. The steady advance and retreat of the sun in his annual course would probably cause a uniform increase in temperature from day to day from winter to summer, and a corresponding decrease to the winter months, if the character of the earth's surface were uniform. The distri- bution of land and water surfaces is doubtless responsible for the irreg- ular character of the curve representing the annual changes of temperature when constructed from mean daily temperatures. TABLE XIV.— MEAN DAILY TEMPERATURE. (Corrected to hourly mean.) 1. S. i. 5. 6. S. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 2.5. 26. 27. 28. 29. 30. 31. Jan. Feb. 33.5 33.9 34.3 31.5 32.2 33.9 33.2 3:3.8 33.2 30.9 33.0 32.7 34.4 34.3 &3.7 35.3 33.0 34.7 32.2 34.5 31. li .36.1 .34.0 36.7 33.2 36.1 34.8 a5.3 33.1 35.9 34.6 35.7 33.8 36.5 33.2 37.5 33.9 .37.3 34.2 36.3 36.0 37.5 35.5 38.1 35.1 37.1 .32.2 a5.3 32.4 37.4 32.5 37.7 .33.8 36.0 .34.8 35.9 33.6 36.6 33.0 34.3 38.2 38.0 38.7 37.8 36.8 39.0 40.2 39.5 40.3 43.1 41.4 42.8 41.8 40.6 39.0 39.9 39.2 39.8 43.2 41.4 41.2 42.2 41.2 41.0 42.0 42.5 44.0 45.0 43.4 44.5 46.2 47.0 49.2 48.8 48.4 48.7 49.8 49.8 50.6 .50.2 50.2 50.6 51.2 52.4 .55.2 53.2 52.8 52.7 .53.4 55.2 55.4 .55.9 55.6 57.2 57.6 .56.4 .58.0 57.4 59.7 59.3 60.5 58.9 59.5 61.2 60.9 60.7 61.7 64.9 66.2 65.8 64.5 62.7 61.9 63.8 64.0 63.4 63.9 &5.5 66.5 66.4 66.1 65.2 65.7 67.1 67.5 67.3 68.6 67.3 68.8 70.3 rune July Aug. Sept. Oct. Nov. 70.2 75.5 76.6 72.4 62.5 51.5 ' 69.0 75.7 76.4 71.8 61.6 53.4 71.2 78.1 76.6 7'' 7 62.2 50.1 71.4 78.7 76.6 72.0 63.0 47.7 71.2 77.5 76.7 72.4 60.9 48.5 71.6 77.. 5 77.0 73.6 60.7 49.0 , 70.4 77.7 76.8 71.8 59.0 48.8 ' 72.1 77 7 77.0 71.4 .59.4 49.6 72.6 77.9 77.4 70.6 59.0 49.9 ; 72.3 77.9 77.5 70.3 59.0 49.3 72.6 77.7 76.8 68.6 58.7 47.7 72.4 77.6 76.8 69.7 .57.4 47.1 72.4 78.5 76.2 69.6 58.0 46.3 73.3 78.2 75.6 67.6 58.1 44.9 73.8 78.9 75.6 67.1 56.9 45.3 72.8 79.6 75.4 69.0 57.8 46.2 73.9 78.9 75.0 68.8 .57.4 45.7 73.8 79.1 75.6 67.7 57.2 46.3 74.6 77.3 75.4 69.3 56.2 44.3 75.6 77.3 75.4 66.8 54.7 42.0 75.6 77.5 75.8 64.6 54.6 41.3 75.2 76.7 75.2 64.8 54.3 42.8 75.1 77.6 74.3 65.4 55.0 43.7 ' 76.1 76.8 74.6 65.2 54.1 41.7 75.9 7';. 3 74.4 64.8 .53.1 41.5 77.fi 78.7 74.0 65.4 53.4 41.6 76.2 78.2 72.8 64.0 53.8 41.9 76.6 77.1 72.1 64.1 o«.o 40.8 76.2 77.4 72.6 64.0 .52.6 38.1 75.6 77.5 73.4 62.9 52.0 36.2 76.9 73.2 50.6 36.3 38.3 38.5 39.0 37.9 38.3 39.3 38.6 38.7 37.8 40.0 39.9 39.7 38.3 36.9 36.2 36.3 36.3 35.9 33.9 a5.7 37.3 37.8 36.5 35.5 34.3 34.0 33.9 33.7 34.1 33.8 Table XIV shows the mean temperature for each day of the year as derived from the daily maximum and minimum temperatures for 30 years, from 1871 to 1900. To the average daily values derived from these observations, cor- rections have been applied to reduce them to the true mean based on 24 hourly observations. The altitude of the thermometers varied from 40 to 60 feet above the ground. In Table XIV the average temperature for each day of the year is shown, based upon the daily maximum and minimum temperature for a period of thirty years. The corrections wliieli were found necessary in 78 THE CLIMATE OF BALTIMORE the preceding paragraphs, in order to reduce these values to the true daily mean based on 2 4-hour ly observations, have been applied in this table. In Plates III and IV the daily mean temperatures for the same period (1871-1900) of thirty years, are shown graphically in curves B. The irregular serrated appearance of these curves is very marked. The advance and retreat of the seasons is accomplished by a succession of waves of rising and falling temperature, of unequal period, but averaging about three to four days. These changes accompany the areas of high and low atmospheric pressure which pass in continual succession from west to east within the temperate zones of the northern and southern hemispheres, and which have become familiar to us in the daily weather charts now isssued by nearly all national governments. A study of the curves representing the daily temperatures for the year, shows a greater variability in the winter months than in the summer months. This is more readily recognized in the curves of extreme tem- peratures (Plate IV, curves A and C), than in those representing the average temperature for a long period (curve B). In the past thirty years the temperature has been lowest, on the average in Baltimore, on the 5th of February (30.9°). The day having the highest average tem- perature of the year is the 16th of July (79.6°). Hence the temperature rises during 161 days and falls during a period of 204 days. From April 20th to October 23rd the temperature remains above the average for the year; from October 23rd to April 20th it is below. The temper- ature rises most rapidly in March and falls most rapidly in November. As stated above, the temperature of the air at the earth's surface lags behind the temperature of direct solar radiation nearly a month, the latter attaining a maximum value on June 22nd, the former about July 17th, at Baltimore. This lagging effect is particularly noticeable in the tem- perature of late summer and the autumn. On the 22nd of March and of September the direct rays of the sun which fall upon Baltimore are presumably of approximately equal intensity, as the sun is at these times directly over the equator. The temperature of the air, however, screened from the direct rays of the sun, is 65° on September 22nd, wliile it is only 42° on the 22nd of March. This marked difference between the MARYLAXD "WEATHER SERVICE 79 temperatures of corresponding days of the ascending and descending branches of the annual curve holds good throughout the year. TABLE XV.-MEAX DAILY CHANGE OF TEMPERATURE. (Expressed in degrees and tenths of a degree.) Jan. —0.4 Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 0- 1 —0.5 1.7 0.4 —0.5 -0.2 0.0 -0.2 -0.6 -0.2 1.0 0.1 1-3 0.8 -2.4 -0.2 2.2 1.2 -0.2 0.3 —0.2 -0.6 -0.9' 0.9 2.0 2- 3 -3.1 2.4 0.7 —0.4 —1.6 2.2 2.4 0.2 0.9 0.6 -2.3 0.2 ^4 1.0 0.0 -0.2 -0.1 —2.9 1.8 -0.9 —1.0 2.2 -0.4 0.6 0.2 0.3 1.7 -0.2 1.1 -0.3 0.4 0.6 -1.2 0.0 0.0 0.1 0.3 -0.7 0.4 1.2 0.8 -2.4 —2.1 0.8 0.2 0.5 0.5 4-5 —1.1 5-6 0.4 6-7 1.4 -0.7 -0.7 -0.8 1.6 1.0 -0.6 —0.3 1.2 -0.7 0.8 2.8 0.0 —0.2 -1.2 0.8 1.0 1.7 -0.4 3.2 0.5 0.0 1.3 -0.3 0.2 0.0 0.2 0.0 -0.2 0.2 0.4 0.1 -1.8 -0.4 -0.8 -0.3 —1.7 -0.2 0.4 0.8 —0.4 0.3 0.0 -0.6 1.0 7 8 — O.T 8- 9 0.1 9-10 -0.9 10-11 -0.6 2.4 1.6 0.6 -1.7 1.4 0.4 -0.4 0.3 0.6 -1.3 -0.2 -0.2 -0.1 -0.7 0.0 -1.7 1.1 -0.3 —1.6 -1.3 -0.6 •> 11-12 -0.1 12-13 -0.8 -0.6 -1.0 1.2 -1.8 0.0 0.9 -0.6 -0.1 0.6 -0.8 -0.2 13-14 1.6 -0.8 —1.2 3.8 -0.8 0.9 —0.3 -0.6 —2.0 0.1 -1.4 —1.4 14-15 -1.7 0.6 -1.6 —2.0 1.9 0.5 0.7 0.0 -0.5 -1.2 0.4 -1.4 15-16 1.5 -0.2 0.9 -0.4 0.2 —1.0 0.7 -0.3 1.8 0.9 0.9 -0.7 16-17 -0.8 0.8 —0.7 —0.1 0.6 1.1 -0.7 -0.4 --. ^^ l.\ treme range of temperature. E. Average daily range of temperature. MARYLAND WEATHER SERVICE 83 Average Daily Kaxge. The average maximum and minimum temperatures for each day of the 3'ear for a period of thirty years are shown graphically in curves A and C on Plate III. These curves show the characteristics already described in considering the mea7i daily temperatures, which was to be expected as the latter were derived from the daily maximum and minimum. The average daily range of temperature, or the average difference between the highest and lowest readings for each day, is shown in Table XVI. The daily range is also shown on Plate III by the difference in value of correspond- ing points in curves A and C and directly in curve E on Plate IV. Dur- ing the winter months the range is least; it increases in the spring months, reaching a maximum in May and June, then decreases steadily to a minimum in January. As the daily range is largely dependent upon the amount of cloudiness and atmospheric movement, as shown in pre- ceding paragraphs, a considerable variation in the range from year to year in the same month may be expected. In January, for instance, with a range of 13.3° as an average for thirty-two years, it has varied from 11.2° in 1891 to 17.0° in 1876. The March range has varied from 11.8° in 1891 to 22.0° in 1873. The smallest average range for any month occurred in Xovember 1884, namely 10° ; the greatest average was that of March, 1873, with 22.0°. When the daily range is averaged up for an entire year the variability is reduced to comparatively narrow limits. The annual average was smallest in the year 1882 (14.1°) and greatest in 1900 (16.9°). The ten-year averages have varied only between the limits 15.2° and 15.9°. AVERAGE DAILY RANGE OF TEMPERATURE. Period. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year 1871-1880 13.8 14.2 15.9 1881-1890 13.5 14.4 14.6 18!tl-HK)0 13.5 13.8 15.0 1871-1902 13.2 14.1 15.3 16.1 17.1 17.2 16.7 17.6 16.6 18.0 17.4 16.7 17.0 18.0 17.4 16.3 14.8 15.7 16.2 15.7 ; 15.1 17.3 17.3 I 17.3 16.7 16.0 16.1 16.8 16.0 16.7 14.0 13.6 13.5 13.3 14.6 13.4 16.3 14.0 13.4 15.5 15.2 15.9 Diurnal Variability of Temperature. A climatic factor of the highest importance, especially to those who are not in the best of healtli, is the variability of tomporature conditions from 84 THE CLIMATE OF BALTIMORE day to day. The magnitude of the diurnal change may be represented in various ways : either by comparing the extremes of temperature of one day with those of the following, or the average daily temperatures, or readings o' )° 2° 3° \° 5' 6' 7" B" 9° 10 1 2 14^ 16 18° 20" 2 2 " 24 • 2 ' ; 1 I , I , 1 ; 1 ! - "V - 1 1 j - / l_j ! ' \ 1 /; f / 1 -4^ - V 1 i \ / ' ' I 1 \\ \ |. \ 1 1 V 1 > , I Year , \ 1 > : V I \ I V > ^ - - 1 \ ■ 1 1 V \ ^ \ j V 1 y- i \ ^ \ \\ / s \ /l ^ k N. / .- ^ 1 * . ^ y \ V b "■•*•. s V \ 1 '''< ] ' **. L. \ c / ^ ' ^ ■^T^ s d J / V V s / s '^ ^ e s ^ ■^ s^ v '^ V s k V >s 'v -^ _ \ *>^ - — _ _ _ _ _ I- _ _ ^ -- ^.~ -iaT?^ =:) .:^ S:?-^ = 1 1 - J _ _ _ _ _ ~ 1 - n — - - - - - -n- H ^ f '-^ ■^ " r^^ T " " -1 Fig. 19.— Total Seasonal and Annual Frequencj' of Stated Diurnal Chang-es of Temperature. (a) Total Annual frequency. [d) Total Spring frequency. (b) " Summer " (e) " Winter " (c) " Autumn " rig. 19 shows the total number of stated changes In the mean daily temperature during each season and during the year. The upper horizontal line of figures indicates the degree of change, and the marginal figures to the right of the diagram show the frequency of stated changes. See also Table XVII. made at the same hour of the day. The frequencj of changes of a given amount will depend somewhat upon the method chosen for determining the daily change. "We have seen above that the normal change in temper- MARYLAND WEATHER SERVICE 85 ature from day to day varies from 0.5° or 0.6° in the summer months to 1.0° or 1.2° in the winter months. But this average change for a long series of years is of less significance than the frequency of changes of a given amount. Large, and especially sudden, changes in temperature within short periods have never been considered particularly desirable from any point of view. Such changes may have advantages, but the un- comfortable, if not actually harmful effects, of rapid changes are sure to outweigh these. As a general rule proximity to the ocean will insure an equable temperature, free from sudden and large changes. Especially Jan. Feb. MCH. Apr. May UNE July Aug. Sept Oct Mo /. De -^-^^ i 1 \ _ 1 -| — \— r+4- 1 1 I 1 - _^_ -i — - "IP ?= -■ ■= -- ^s -*- r - -j- -^ ^\ _L t^ 1 k N i 1 1 ! ] ^■*^ "v. 1 ' 1 1 1 . 1 1 1 ^ ^ 1 k ,, -6 1 "SJ" Jf^ ^^~ r^i . i 1 r ^ ~" T 1 1 1 -, ^ ^""^^ ; , ^N j ^ -4- rf' ^-i ""i P -U r -- — — H J^ ^< ^- k_. --4-J r -- -- -- - -,^- -- -- 1 1 1 1 "^^ y ^ 1 t s / k.* ' -I-l 4- ==m:;- ^^; f- h- t^ ^ = _L ^ ^'_ = = c ►_. -t^ 1 1 — +^ 44^-"«' 1 "" ^-^ ■^ - ~: -- n T s* -j — -i. w^ ^ V'- 1 1 ..II 1 ; , ^^^. iit^ 1 -I -- L_ -HI 1 1 ' i : ! ' ' ! 1 i 1 ' ' I ' Fig. 20.— Diurnal Changes of Temperature of less than 6°, 6° + , S°+ and 10° + each month. Fitr. 30 shows the f reaucncy of changes of 0° to 5°, of 6° and above, 8° and above, and of 10° and above, in the mean temperature of the day for each month of the year. The degree of change is indicated by the curved lines marked —6°, 6°+, 8°+, and 10°+ respectively, while the frequency of the stated changes is shown by the marginal figures to the right of the diagram. For example, a rise or fall of .5° or less in the mean temperature of the day occurs on the average 1" times in January, 21 times in May, and 25 times in July, etc. ; a rise or fall of l(J°or more in the mean daily tempemture occurs 6 times in January, 2 times in May, etc. See Table XVII. is this true on small islands, or along the western coasts of the continents where ocean winds prevail. The diurnal variability increases rapidly as the interior portions of the continental areas are approached. The changes in the daily average temperature at Baltimore, covering a period of thirty years, have been computed and arranged according to fro> o 3 3 1-5 3 < *3 C ® CO O > o 2; c5 ii a Summer. Autumn. 1 0° 2.0 1.9 1.7 3.5 2.5 2.9 3.5 3.7 3.6 1.7 1.9 l.S 6.710.2 8.3 5.7 .30.8 1 3.0 3.4 3.1 3.2 3.9 4.4 5.5 5.4 4.4 3.6 3.1 3.1 10.215.311.1 8.545.0 2 3.2 3.6 3.6 3.4 4.3 4.7 5.6 5.5 4.3 3.8 3.6 3.4 11.115.711.7 9.347.8 3 3.3 3.7 3.3 3.4 3.9 4.2 6.0 4.9 4.2 4.2 3.9 3.5 10. 6,14.012.3 9.446.3 4 3.7 3.7 2.6 3.3 3.3 3.3 3.3 3.3 2.9 3.5 3.6 3.0 9.1 9.610.0 8.437.1 5 3.6 3.7 2.6 3.3 3.0 3.7 3.6 2.5 3.6 3.2 3.3 3.0 8.9' 7.8 9.0 8..3'.34.0 6 3.3 3.3 2.2 3.6 3.3 3.0 1.8 1.5 1.9 3.3 3.4 2.5 7.0 5.3 6.5 6.925.7 7 2.1 2.0 2.0 2 2 1.9 1.7 1.3 1.1 1.8 2.1 3.0 2.1 6.1 4.2 5.8 6.223.2 8 1.8 1.7 1.6 lA 1.4 1.2 0.7 0.7 1.2 1.5 1.4 1.5 4.5 2.6 4.2 5.016.3 9 1.7 1.5 1.8 1.3 1.3 0.9 0.5 0.5 1.0 1.3 1.3 1.3 4.1 1.91 3.5 4.614.1 10 1.3 1.1 1.6 0.8 0.9 0.5 0.3 0.3 0.6 0.7 0.9 1.0 3.3 1.1 2.5 3.510.2 11 1.1 0.9 1.3 0.7 0.7 0.3 0.3 0.1 0.6 0.6 0.8 1.0 2.6 0.7 2.0 3.0 8 3 13 1.0 0.6 0.7 0.5 0.4 0.2 0.1 0.1 0.3 0.4 0.6 0.7 1.7 0.4 1.3 2.3; 5.6 13 0.8 0.6 0.6 0.5 0.3 0.1 0.1 0.1 0.3 0.3 0.4 0.6 1.3 0.3 1.0 2.01 4.7 14 0.6 0.4 0.5 0.3 0.1 0.1 0.1 0.3 0.3 0.5 l.Oj 0.1; 0.5 1.5j 3.0 15 0.3 0.5 0.4 0.2 0.1 0.1 0.1 0.1 0.2 0.3 0.4 0.7 0.2' 0.5 1.2 2.6 16 0.3 0.3 0.3 0.1 0.1 0.1 0.1 0.3 0.4 0.1 0.3 0.8 1.6 17 0.1 0.3 0.2 0.1 0.1 0.1 0.2 0.3 .. 0.3 0.6 1.3 18 0.1 0.2 0.1 0.1 0.1 0.1 0.2 0.2; .. 0.2 0.5i 0.8 19 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.3 0.4 0.8 20 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.3 0.6 21 0.1 0.1 0.1 0.2 0.3 29 0.1 0.1 0.2 ^ 0.1 0.1 0.2 24 0.1 oil 0.1 0.2 25 0.1 0.1 0.1 26 0.1 0.1 The first column indicates the degree of change from day to day in the average daily temperature. The figures in the remaining columns show the average monthly, seasonal and annual frequency of occurrence of indi- cated changes, based upon observations during 30 years, from 1871 to 1900. The larger daily changes decrease in frequency on the approach of the summer month.*. A cliange of 10° in the average temperature of two consecutive days has occurred about 3.5 times in 30 years during each winter month and 9 times in each of the months of Julv and August. MARYLAND WEATHER SERVICE VOLUME 2, PLATE I A. Daily maximum temperature. B. Daily mean temperature. BASED UPON D.«lV0'^^^'ONSPo„3„^.^ C. Daily ."ini"'""*"""^' D. £ reme range of temperature. E. Average daily range of temperature. MARYLAND WEATHER SERVICE 87 The details of these changes are shown in the table above and in Figs. 18, 19, 20 and 21. These figures and the diagrams reveal the interesting fact that the average departure and the most frequent departure are not identical, or that the arithmetical mean of all departures for any given month is not the most probable value. In the winter months the average departure is about 1°, in the summer months it is about 0.6°. The most probable departure is in all months larger than the average departure, as is clearly brought out in the following comparison of the average change from day to day and the most probable change. DIURNAL VARIABILITY. Jan. Feb. Mar. Apr. May June Julj' Aug. Sept. Oct. Nov. Dec. Year Average change. .± Most probable change ± 1.0° 3° 1.0° 4° 1.2° 2° 0.8" 3^ 1.0° 2° 0.6° 2° 0.6° 2° 0.4° 0.9° 1° 0.7° 3° 1.0° 3° 0.8° 3° 0.8° 2° The true measure of diurnal variability is the change in the average temperature from day to day. It is more convenient, however, to express this cliange by means of the difference between the highest or the lowest temperature of successive days, or between the temperature at any given hour, as 8 a. m. of one day to 8 a. m. of the following day, or from 8 p. m. to 8 p. m. The results will differ somewhat according to the method adopted. Assuming as correct the variability as measured by means of the daily average temperature, the variability based on differences of the minimum temperature from day to day will give too many changes under ()°, while those based on the 8 a. m., 8 p. ni. and the maximum tempera- tures yield too few small changes, the departure from the true frequency being in the order named. On the other hand when we consider the larger changes of G°, 8°, or 10° and above, the minimum temperature yields too many. Take as an illustration the frequency of changes of 10° and over. During the course of a year of average temperature conditions there are -11 diurnal changes of 10° or more, basing the count on the daily minimum, 59 on the 8 a. m., 58 on tlie 8 p. m., and 83 on the maximum temper- atures. That is to say, the variability computed from differences in the uiaxiimiiii temperature from day to day may show more than double tlie MARYLAND WEATHER SERVICE. jtN Apr. July Oct Jan. Apb, Jul 1817 1818 Oct. Jan Apr. July Oct Jan Afr. Jl : . [ , [ ^ -| ill.'' 1 ' ; i I ' 1 ' ' 1 1 I 1 ' ' . I 1 ; ' I ■ / L ' ' ' iv'M' 'aN. i A -^ — ^^- ' ' - ~^' '" "t 1 ■ 1 y V 1 yj^ i ■ J *" "^ /a fc/ \ /^ /^ y^L- f~ ' ' / '■^ ^"^— ^ ^ \ ' A\ -• \ f \i \y^ i\/ \ ' S/ 1 1 y r j! ' T 1 1 '" iL i -I- "■ """" -U- _u-'|- -(--- p.- .. - ! 1 1 1 1 1 i ! M i 1 1' ' 1820T: 1821 X 1822^ IL 1 1823 i ; i 1 1824 1 ! II 1" ■ r 1 'J ■■■' T j 1- - 1 1 [| Mill!,.! i 1 ^ 1 ; j ! ■ 1 1 L i 4 1 ■ A ' ! 1 ■ I ' \ ■■ 'A /\ I 1 1 _l A ' » /V i ■ /^ ''+\ __^^ /\. ^\ y\i i ij\ 11 i/\ / ^^ I 1 ' Jhii 1 f^ /V / ^ I _/ ' *' 1 ' 1 I / ^ •^''~~>* J ^"-^ Xi Ki- ,1 —^ /" ! \ /^ lA ^ / V / ^ ' \ , \ /\ . ! i \ \ 1 Y/, ^^^ , ^ — ..^ / ^~T^ ^r ' V ' ' -^ \f 1 P '''^ 1 1 I ' y i 1^ ' ■ ■' ; , ! 1 ■ ' 1 : ! ' . ' i 1 1 1 i ~r~^ — 1^ - M 1 i~ ■ 1 1 ■ ! 1 1 .1 1 1 1825! 1826^ ,1827 ,8?a 1 1829, _u^ f 1 i 111 itr "■ XSl _L 1! . U ' 1 1 ! ' 1 1 li ± '^"-J^ \ -. . _^i LL - i^^^-isi23^i±i,^ , —-- ^ "\ -TN /^^ r\ - - , ^ J- ' ^ ^^ ,_._^li/L... .>'^_^.v'._j_'!sAi^Z^ 1 , . 1 1 i ' i [ ' ' [ - -i- 4--L -A . ' --4 ' -Li -!_ -1- 11 \ ■• ill 1 1830 1831 1832 1833 j 1 1 IH..?* , j i 1 i ' i ] ' ' ! hi ' 1 , A /V M / \ ' /^ / \ y v^S. A 1 : . / N ■ ..J ' . rjT^s z""^- ;/:/ v^f/C/:^/^^'^" -t^ v \/ iS-, K y\' \ y^^t- - ■^. SjT" '^'^ .-^ ^. r v*s/V^ i ^^-^ , 1 >< .\-/^ — ' 1 ' ' j 1 ! '1 1838 t- - ■ -■ --j 1839 I 1 :. ....:!... ■.... 1 j I" 1^ "^ ; ., 1 I 1 1 i '1 11 ' i 1 i 1 > ^ <^ ' ■ ■ / \ ' ^.j/\ . . /K ,^^, ■ ; i/v ■■ vjA/^'^N.V Vr A. vA i_ /\/> ^"^^HW^S:t?^^Sfc „V^/^ V/ M . >/v/\ — / s/ -6-.-^ \t- ^^ -\^ 1 f itlt i 1 f < n^i^ irt40 la^i ift4.? 1 843 . ,1 844 ^ ill 1 1 1 I ' ^ j 1 ' — ""^ J 1 /^ ' ' ' ' ' J / /\ y''^ y ^^ 'J— ) /' 'S!^-^ xk . — , -. !• ""^ \ / "^ V ■."^-^"S f \ h" - /^ -"""'''''^ V ^~" \ ^ / \/ \V \ / 1 \j ^ \ / \ i\ f "^ , 1 ^ — /^ 1 1 ^"v'^ ' 1 \'/\/' '' li 1 ' 1 v V ; 1 ' L -u 1 t > 1 1 1845^ 1846 1847 ,_ it 1848 It: it 1849, [ 1 V .y . ' i-^^^i^i'xz .^±,zi . ^\.j\^^^L:lt. Ji \>-'-^:^-\;y /-''■'■s--''^/^--^ ^^ ' -T,.' - -».^ ■•■>/>/ y/ Lr/>V '•" *■ i 1 ' >/ \/ ^.' NF- 1 I 1 ■ 1 I 1 1 T . I850+- 1851a_ 1852 1853 1854' "T 1 1 I 1 A iH /-s y ^-N>v.y-sr i--L ^ .\/^A,r-A, iA i>-^\ ;> —' ^'^ V iv^/-s>.' - +---v^s^ ^ / aTS j i 1 1 1 1 1 1 'SoSjI 1856 1857' 1858 1859 |_ - ' 3^ =E "J i i \ 'A ^l - \ A - \ J i V ^» 2l V 7\ .--/ ^1 V/ '^ . -^ /V . £ i -V2\ ^\rJ--^'^^-~ -i^- Vy H" ^ I .. y\/.. \^..-S^.. 1.1- -,,,(- 4^ ^^ ^-■'^c ^:=-'^A^ T^ 1 / w ^A ,J-' \' 111 i 1 1 1 VOLUME 2, PLATE VI. 1819 Oct. Jam Apr. Juiv Oct .lAr^j DEPARTURES OF MEAN MONTHLY TEMPERATURE FROM NORMAL FOR 87 YEARS. MARYLAND WEATHER SERVICE. VOLUME 2, PLATE VI. I860 1861 1862 1863 1864 JAN. Apb. jul^ Oct. jam. Apr July Oct. j^n Apb July Oct. Jan. Apr. July Oct. Jan. Apr. July Oct. ja 1865 88 THE CLIMATE OF BALTIMORE true frequency of the larger changes of 10° and above, while the smaller changes are below the true frequency; hence changes in the daily maxi- mum temperature are not a safe guide to the diurnal variability in the geographical horizon of Baltimore. In all cases the changes based upon observation of the maximum temjjerature from day to day differ most widely from those based on changes of the average daily temperature. In Fig. 21 and in the following table these results are shown graphically for changes under 6°, for 6°+, 8°+, 10°+ and 20°+, when the diurnal \ \ \ / \ \ \ \ \ \ / \ ^ \ / I \ \ \ \ > / 20°+ 10^ bV eV -0 Fig. 21.— Diurnal Changes of Temperature of —6°, 6° + , 8° + , 10° + , 20° + . Fig. 21 shows the frequency of stated changes in the temperature from daj- to day when based on the minimum temperature of two successive days, on the mean temperature, on the 8 a. m., on the 8 p. m., and on the maximum temperatui-e. respectively. See also Fig. 20, and Table XVII. The frequency is indicated bj' the line of figures above the diagram, the degree of change, Vjy the line of figures below the diagram. variability is based on observations of the maximum, the minimum, the 8 a. m. and 8 p. m. readings and on the true daily mean. Diurnal changes were computed for a period of 30 years from the daily average tempera- ture, and for a period of 10 years from the maximum, minimum, the 8 a. m., and the 8 p. m. observations. FREQCENCV OF DIURXAL TEMPERATURE CHAXCiES OF STATED AMOUNTS. (Expressed in terms of departures from the frequencj- based on changes in the daily mean temperature.) Temperature changes. Minimum. Departure. Below 6° +11 6°+ - 5 8'+ i - 10 10'+ - 9 20°+ ' -0.4 Mean. 241 119 71 41 1.7 Departure. — 16 + 19 + 21 + 18 +2.7 • p. m. Maximum. Departure. — 21 + 21 + 20 + 18 +2.3 Departure. + 51 + 48 + 42 +6.6 MARYLAND WEATHEE .SERVICE 89 The smaller changes, under G°, increase in frequency very rapidly from February to July, then decrease at a similar rate to February. Changes of 6° and over occur most frequently in the months of December,. Jan- uary, February and March ; the decrease is then uniform until a minimum frequence is reached in August : then there is a more rapid increase to TABLE XVIir.-FIVE-DAY MEANS OF TEMPERATURE. (For five-day periods ending on given daj-s.) January. February. March. April. May. June. 5th 33.3 4th 33.5 1st .37.0 5th 48.4 5th 59.9 4th 70.4 10 33.3 9 as. 6 6 38.1 10 50.1 10 63.9 9 71.6 15 as.s 14 35.7 11 40.9 15 52.5 15 6:^.7 14 72.6 20 33.9 19 36.6 16 40.8 30 53.9 20 64.7 19 73.8 25 34.3 24 36.9 21 41.0 25 56.5 25 66.1 24 75.5 30 33.5 26 41.8 31 44.6 30 58.1 30 67.9 29 76.5 Jul}-. August. September. October. November. December. 4th 76.7 3rd 76.8 2nd 72.7 2nd 63.0 1st .51.8 1st 3S.7 9 77.7 8 76.8 7 72.5 7 61.3 6 49.5 6 38.4 14 78.0 13 76.9 12 70.1 12 58.7 11 49.1 11 3S.9 19 78.8 18 75.4 17 68.4 17 .57.6 16 46.0 16 38.3 24 77.2 23 75.2 23 66.6 22 55.4 21 43.9 21 35.6 29 77.7 28 73.6 27 65.0 27 53.9 26 42.3 26 38.3 31 33.7 TEN-DAY MEANS OF TEMPERATURE. (For ten-da J- periods ending on given days. Derived from above table of five-day means.) January. February. March. April. May. June. 10th 33.3 30 33.6 30 33.9 9th 33.5 19 36.2 1st 37.0 11 39.5 21 40.9 31 43.2 10th 49.3 20 53.2 30 57.3 10th 61.4 20 64.3 30 67.0 9th 71.0 19 73.2 29 76.0 July. August, September. October. November. December. 9th 77.2 19 78.4 29 77.5 I 8th 76.8 18 76.2 28 74.4 7th 72.6 17 69.3 27 65.8 7th 62.1 17 58.2 37 54.6 6th 50.7 16 47.5 26 43.1 6th 33. 5 16 38.5 26 36.0 Jan. 5 33.5 Table XVIII shows the mean temperature for each successive period of five (lays beginning with January 1st, and also for each successive period of ten days. The 5-day and 10-day means were computed from the normal daily temperatures for the 30-year period 1871-1900, after reducing the latter to the true daily temperature based on hourly observations. December. A change of 20° in the average temperature of two successive days has occurred at Baltimore about 50 times in the 30 years from 1871 to 1900. Of these occurrences 15 were recorded in February, 10 in January, 8 in December, 5 in ^larch, 5 in November, 2 in each of the 7 90 THE CLIMATE OF BALTIMORE months of Aprils, May, and October, none in the months of June, July, August and September. The most frequent change and hence the most probable, is a change of 2° in the spring and summer and 3° in the autumn and winter months. The Probable Error of the Meax Daily Temperatures. No law has yet been discovered governing the departures from the normal temperature of a year, month, or day. Departures above and below the normal for a long series of observations agree very closely in their distribution with chance occurrences. Hence the formula applicable to the latter has been employed in the determination of the probable error of average temperatures for a given period. The equation used for finding the probable error of the daily, monthly, and annual means of temperature for Baltimore is a form suggested by Fechner * and is as follows : J. ^ J. 1.1955 ^271 — 1 in which E is the probable error, v the average departure from the normal temperature (in degrees Centigrade) not regarding the sign of the departure, and n the number of occurrences, in this case the number of years of observation. The value 1.1955 of the factor .'. ^ is as follows for the stated periods of observation : \/27l — 1 ^ 20 30 40 50 60 70 80 90 100 yrs. 0.191 0.156 0.134 0.120 0.110 0.102 0.095 0.089 0.085 In order to determine the probable error of the daily mean temperature at Baltimore for the 30-year period, from 1871-1900, the above formula was applied to a representative day in each season, namely, for the 15th day of January, April, July and October. In winter (represented by January 15), the average departure v of the mean daily temperature from the normal is 7°, in spring (April 15), 5°, in summer (July 15), 4°, in the autumn (October 15), 6°. In individual cases these departures vary greatly. On January 15, 1871, the mean daily temperature was 62°, *See: Hann's Lehrbuch der Meteorologie. Leipzig, 1901, p. 107. MARYLAND WEATHER SERVICE 91 or 28° above the normal value; on January 15, 1893, the mean tempera- ture of the day was 22° below the normal. Thus the loth day of Jan- uary shows a range in the average temperature of the day of 50°. The extremes on April 15th were 17° above and 11° below, a range of 28° ; on July 15th 6° above and 12° below the average, a range of 18° ; on October 15th the extreme departures were plus 15° and — 19°, a range of 34". These figures strikingly illustrate the variability of temperature conditions within short periods at Baltimore. THE FREQUENCY AND AVERAGE VALUE OF DEPARTURES FROM THE NORMAL DAILY TEMPERATURE. January 15th. April 15th. July 15th. October 15th. Departures. Departures. Departures. Departures. Fre- quency. +18 -15 Sums. + 116.6° -116.5° Aver- age. +6.5° -7.8° Fre- quency. +13 -20 Sums. +84.9° —84.0° Aver- age. +6.5° -4.2° Fre- quency. +20 -13 Sums. +61.0° —60.2° Aver- age. +3.0° -4.6° Fre- quency. +17 —16 Sums. +106.6° —105.2° Aver- age. +6.3° -6.6° 33 233.1° 7.1° 33 168.9° 5.1° 33 121.2° 3.7° 33 211.8° 6.4° Entering the values of the average departure v in the formula we obtain as the probable error of the mean temperature of a typical winter, spring, summer, and autumn day the following values: January April July October Average Seasonal 1.1° 0.8° 0.6° 0.9° 0.8° These figures represent the probable error of a daily mean temperature in the respective seasons for a series of observations at or near Baltimore covering a period of 30 years. The daily mean temperature will not be increased or decreased by an amount greater than these values by extend- ing the series of observations. Mean Monthly, Seasonal, and Annual Temperatures. There is an excellent series of local temperature observations extending,, with very few interruptions, from 1817 to date. For the series from 1817 to 1824 we are indebted to Captain Lewis Brantz, who kept a careful 93 THE CLIMATE OF BALTIMORE record of the weather, in what was in his time West Baltimore, and pre- sented his published results to the Mar5dand Academy of Sciences, of which he was a member. His observations were made at five stated hours of the day, at sunrise, 8 a. m., 2 p. m., sunset, and 10 p. m., and com- prised the elements of pressure, temperature, wind-direction and force, clouds, and rainfall. In 1831 systematic observations of the principal climatic elements were made at 7 a. m., 3 p. m., and 9 p. m. at Fort McHenry, under the auspices of the U. S. Army. This series was main- tained to the year 1892 with the exception of two or three years just pre- ceding and during the Civil War. From 1871 to the present time a first order station of the U. S. Weather Bureau has been maintained at Baltimore. To complete the record since 1817 it has been necessary to interpolate observations made at neighboring localities, applying, however, the proper corrections. This could readily be done as the different records over- lapped. The break in the record from 1825 to 1830 was filled in by re- ducing Washington, D. C, observations to the Fort McHenry series; for the years 1859 to 1863 the excellent record maintained for 20 years at Shellman's Hills, about 20 miles due west from Baltimore, was utilized. A year's record by Captain Brantz in 1836 afforded a means of reducing his earlier observations to the Fort McHenry series. From 1871 to 1892 the Fort McHenry and the U. S. Weather Bureau records overlapped. All of these observations were ultimately reduced to the U. S. Weather Bureau series by applying the necessary corrections and were thus con- verted into a comparable and continuous record of great interest and value in the discussion of the climatic conditions of Baltimore City. The monthly, seasonal, and annual means are presented in Table XIX. The departures from the normal values are shown in graphic form in Plates VI and VII. The table and diagrams afford excellent material for the study of the changes in temperature conditions exper- ienced by Baltimoreans during the preceding century and incidentally the results throw light upon the assertion of the " oldest inhabitant " that our winters are growing milder, an assertion which has been persistently repeated since the earliest settlers arrived on our shores. MARYLAND WEATHER SERVICE VOLUME 2, PLATE VIL 880 1885 leSO .'8?? 1900 .t; .34.936.544.41.57.666.971.1 .26. 740.546. 4L55.4I62.2'72. 3 ..32.9-33.6 11.5 48.6.56.4 70.7 .38.939.94'.t.].".5.4r,(i.:i:o.l . 40. 929. 931. 251. .-.M. 77:!.: . 31. 7;i3. 943. 057. 167. 270. 4 .39.335.945.2.55.861.372.9 ! ' i ..34.831.4 43.1.54.2 65.569.2 ..3:5.234.339.1.56.961.971.] . 40. 0:!S. 041., s5-<.l(l,s. 1176.0 . .34 . 4 :J2 . 7 45 . 5 53 . 2 61 . 9 76 . 3 .40. 741. 6143. 9.51. 861. 975. 8 . .39.841.348.1 55.965.572.5 30.537.744.149.263.971.2 , 34.8 :)8. 5 43. 9. 54. 4 65.0 75.1! , 36 . 1 :58 .345. 5 .50 . 2 65 . 7:5 . 1 . 39.1,;50.740.6.58.965.572.3 26 . 3 28 . 4 35 . 3 56 . 2 63 . 3 76 . 6 25.94:i.l tl.5 17.7 ,'6.,s77.471.5.'i4.04:i.o:i6.,s 8.5 75.1 66.9.52.542.1:54.8 '7. 2!76.4i67. 1.55. 9)46. 430.1 5.4 ,270. :!.-,:;. n 1^ :!r.i ;'6.477.2ii.-,.,-,,V. .2i:;.2 15..-, 1'7 . 3 76 . 5 6,s. :.' :Vi . 7 54 . 6 :5,s . ', 1-9.7175.0168.21.57.81.52.4142.0 r9 . 4 74 . 6 119 . 7 58 .517.6 :S3 . 9 r5. 9:7. 066. 2.57. 944. 042.1 1 7. 176. 2 70.1. 5:!. 7 49.7:58.2 r9. 4 76. 4 71. 6.57. 951. 6:55. 4 r9. 176. 9170.2)63. 9149. 6'39. 4 sn. ').0(1,> ■,.411,- 5 17.2:55.9 916. t !:;.-< ;;i.l ■.:,.si\:.s:,s.', c;.'.) 11. ii 76.2 75.168.0 52.4 48.4:52.8 76. 7;75.8|65. 11.56.2145.6,33. 8 74.872.769.1.59.144.737.8 74.675.96'.l.:!.",s.l 12.,'<:!6.7 77.1 77. H64.:!.".2. 7 46.2:51.:! 78. ;!i79. 569. 557. 248. 5:57. 7 79.476.174.1.53.147.8:59.8 79.1 71. 670. .K.5S. 151.5:55.6 77.9",6.:iii!t.6.5-<.(i49.'. :!4.4 ,s: i . 2 ; 9 . (1 71 1 . 55 . ,>< 4.>< . 4 34 . 5 78.176.6,70.15:3.143.240.5 83. .5j80. 671. 1159. 4148.4137. 5 <^ 55. 4| .55.8 .54.6.34.951.8 56. 7,-37. 81.52. 5 .55.037.153.4 g.0|-57. 11817 6.91.56. 511817-8 77. 8i57. 611818-9 5. 8i54. 411819-20 .55.. 1.35. 437, .340. .6:40, I .538, .4:57, .24:!, .2:^5. .6|.38, ].51.077.9.57.91.<«20-1 4."9.]78.7 61.61S21-3 ' 156.6 76.0 55.71822-3 7. 54. 4i75. 6.57.71823-4 4.56.6[77..5[58.4jl824-5_ 9.58.4 77.1.59.9182.5-6' 6 .■.7.7 77. 9. 5,-^.^1826-7 :>.":).,^7.s.4.5.s.Jl,S27-8 1 . 54. 2|75. 1.55. 21828-9 41.55. 8177. 9161. 11829-30 .55.1:54.2.57.0176.5.57.21830-1 56 . 9 :5:! . 5 .54 . ;5 75.7 59 . 7 1831-2 57 . 1 39 . 8 56 . S 75 . 7 .56 . 4 1832-3 .56.7:59.5.55.577.6.55.118.3.3-4 .54. 0,34. 4|.52. 3174. Ij56., 511834-5 .51. 8:53.0 50.2 71. 6. 5:5.1 183.5-6 54.l:!:i.s5].77:i.!i5(;.l ls:56-7 .^4.i:>5.t5].i;s.6."M.5i,s:s7-8 .55.6:34.7.56.;5i74.4 56.]il8;-58-9 53 .9 34 . 4 54 . 7j74 . 21.54 . 61839-40 .52.9.32.748.8 74.4 54.31840-1 55.l):is..-,,54.9^:!.'; .54. 11841-2 .'3. 9 :!5. (I 44.,-.! 76. 0.56. 2 1842-3 54 . 4 ::54 . 1 1 .55 . 8: 74 . 7 .53 . 81843-4 55. 2i.36 . 7154 . 1 175 . ,5.56 . 511844-5 i j ! i i .■4.7:52.1.54.373.357.5184.5-6 55. 6 :i4 . 6 .53. 6 75. 4 -''S- 1 1846-7 57 . 2 :!9 . 2 .56 . 2 ■; 6 . 5 .55 . :! 1847-8 .56. 2::57. 5.53. 5 76. 7.59. 8 1848-9 .57. ,5[40. 3152. 5i76. 81.59. 5 1849-50 .57.2 41.0.56.575.5.58.618.50-1 ,'4 . 9 :!1 . ( 1 .52 . 4 74 . 7 56. (1 18.51-3 .51 i . 4 : ;,-< . 5 ,",4 . 4 ■; 6 . :! .57 . M852-3 .56. 6:57. 5. 5:5. 676. ;i60. 4 18.53-4 .57.ip5.1|55.0j76. 1161.21854-5 ,'4.131.4.51.6 77.5.57.5185.5-6 ,55.ii:i5.ll.50.7';i.,'^,57. 1 18.56-7 .56.,s4(l.4 .5:>.4 77.5.56..S ly,-)7-8 .55.6:59.3 ,56.2 74.U.56.0 18.58-9 54 . 41.33 . 11.54 . 4|74 . 41.55 . 6 1859-60 i I ! I i 55 . 35. 2 .52 . 5 73 . 7 57 . 6 1,860-1 .•3.'; :i4.."51.1 7:5.256.71861-3 .':!.6:!5.1 .■rt)..s75.:).54.4 1862-3 .57.3:57.554.9 77.5.58.4186.3-4 .58.235-5159.277.6160.01864-5 i i ■ I 56 . 4 .37 . 1 55 . 4 74 . 7 60 . 1 1.86,V6 .55. 7:51. 6.5:!. 2711. 059.:! 186(3-7 .55.3:12. 1 .52.4 7,'<..s.5.s.] 1S67-8 .56. fi:5,S.,s.>5. 676. 4.55. .51868-9 .58. £ 40 . 4 54 . 4 81 . .59 .611869-70 94 THE CLIMATE OF BALTIMORE Table xix Con't.— MEAN TEMPERATURES AT BALTIMORE FOR 8S YEARS, 1817-1904. Tears. 1873 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 •O ' _. fe S h^ f s . . .35.■3'.39.847.758.96.^.474.0 ..!34. 436. 0.37. 055. 367. 275. 3 .. 34. 2!.35. 640. 4 52.062.3 73.9 ..39.3,37.443.,S47.063.175.9 . .130.329. 439. 649. 564. 173. 7 ..i41.7i37.9j39.9,.52.264.2i75.9 ..132. 340.614]. 553. 762. 773. .'34.9 40.4 49.2 58.6^3.570.2 . . .31.5.'J2.243.S52.S65.7'73.3 .42.9 41.2 42.855.170.875.3 ..130.334.641.8.51.767.870.8 ..,34.741.345.052.1.59.674.0 ..|32. 339. 1.39. 552. 264. 274. 6 . 132.042.244.1.52.465.073.2 ..|34. 0:28. 535. 4. 54. 363. 272. 6 ..;29. 131. 341. 7.54. 7162. 169. 9 ..132.4.38.237.9.51.367.672.3 .. 129.1.35.. 537. 4.52. 7 62. 8 73. 2 ..138.730. 644. 1.54. 8165. 771. 2 ..43. 7:43. 040. 9.53. 963. 775.1 I , ..37.341.639.0.56.0162.1171.7 ..31.936.237.3.51.963.175.5 ..24.334.040.4.52.561.672.5 . . 37.434.447.S.''i2.364.W73.1 ..31.4 26 . 2 40 . 2 52 . 8 62 . 5 74 . 3 ..133.335.237. 8.56., 568. 7 71. 3 ..130.8.36. 6 45. 2. 52. 66;^. 4 69. 9 }6.7;34.7;48.6.51.46:3.573.7 !2.7as.041.6r)3.664.875.3 35.333.138.6.54.864.9 72.9 34. 428. 742. 951. 062. 0'73. 5 31 . .529 . 8 46 . 2:.53 . 2,63. 9!71 . 9 .33. 536. 9149.7 54. 8165. 5167.5 27.538.641.350.5165.871.6 Ten-Tear Mean. 1821-1830 1831-1840 1841-18,50 18.51-1860 1861-1870 1871-1880 1881-1890 1891-1900 Gen'l Averages. 1817-1870 1871-1903 1817-1903 35.737. 043. 653.5j64. 9 33.6136.440.8.53.064.2 33.134.041.653.463.9 .36.337.746.155.765.575.0 34.135.843.7.54.264.272.6 36 . 7 35 . 1 42 . 3 .'i 4 . 3 62 . 72 . ( i 34. 7 36. 3 43. S 53. 4 64. 2 73. 3 35.2|35.742. 754.663.9 74.0 35.4 34.9 36.1 34.035.5 35.8 43.6 42.1 43.1 54.6 53.3 54.1 63.8 64.3 64.0 74.1 73.7 73.0 73.5 73.1 73.3 76.0177.563.758.745.233.8 81.7|79.869.5.56.943.7|32.4 ■9. 676.668. 3.55. 341. ,5;40 ■7.573.1 70. 21.57. 145. 8j39. 3 ■8.273. 766. 2:,56. 043. 438. 6 80 . 6 76 . 2 65 . 9'52 . 9 47 . 6 29 . 8.977.968.260.24S.443.,s 80.876.069.71.58.847.4 35.6 9.075.365.263.046.741.9 ■7.7 74.7|68.3|,56.O42.031.3 ■8.977.2'77.2:a3.,549.3;43.8 6.9 74.269.361.944.5:^.7 '7. 173. 265. 41.57. 94S. 430.0 5. 475. 572. 4,60. ti Hi. r,;;;. t ■9.974.9,67.2:55.8 t:,.,s;i;.ti 4.773.869.91.59.046.6131.3 0.673.665.01,56.345.3137.2 4.676.565.11.53.048.337.2 6.573.766.3.54.147.945.1 6.374.167.8.57.047.534.9 71. 6(74. .5'70.5|54. 4143. 9143.0 ■6. 8j76. 4:66. 7, 56. 044. 0133. 7 ■6.875.665.9157.1 43. 9'38.] ■7.473.769.9.57.443.337.3 3.1 77.271.71.53.4 47. 03S 7.7 77.068.0,54.3.50.936.2 7.074.968.9.58..5'46.238. 8.677.771.2.59.044.6.36. 77.1 76.066. 5,59. 046. 736. 8 9. 680. 373. 562. 0,49. ,5,36. 6 9.876.9168 77. 274.1167 77.373.767.7 75.7. .56.1141.8134.7 58.6 58.7 51.735.3 43.333.9 78.678.069.8.57.348.3140.8 7.775.467.1,55.345.135.3 7.475.96,s.(i.54.ii4(i.;):!7.5 ,s.0T;-).>fi,s.(l.-,7.14;.437.7 8. 676. 669. 8157. Oi47. 136. 9 79.0 77.1 76.6 78.1 77.6 77.9 76.1 74.7 76 369 76.4 75.6 76.1 67.5 68.6 3 57.5145.236.6 .57.947.038.0 57.146.0 56.0 57.5 .37.5 37.7 37.1 56.6146.637.4 47.0 46.0 ,56.237.5.57.375.8.55.91870-1 34.7,53.3 •55.034.1,51.6 ■55.8,39.1.51.3 •53.533.0,51.1 ■55.3.39.4,52.1 .56.8.34.0-52.6 •57.] 39.7.57.1 •55. 8a3. 1.-4.1 •56.442.0.56.2 •57.232.1 .'S. 8 •55.8.39.9.52.2 .55.236.0.52.0 56.337.7 r3.« .54.0:33.3.51.0 .^^3.6132. 7! .52. 8' ,':4.7,34.0,52.3 ,'■4.133.951.0 55.635.5.-4.9 .56.4 43.9,52.8 •55.437.9.52.4 ."4.037.0.50.8 .'3.4,30.7.51.5 .55.6.36.6.55.0 •54.031.6.51.8 •55.. 5 35. 7 •■4. 3 .55.234.5.53.7 .56.2.36.6.-4.5 •■4.832.3.-3.3 .56.635.1.52.8, .54.1*3.2.53.0 ■55. 033. 0.54. 4' .54.935.2)56. ■8.9.56.7:1871- 6.7 .55^5 1872-3 ■5.5.57.71873-4 ■5.2.55.21874-5 '7.6.55.5187.5-6 ■6.8.58.91876-7 ■5.7.58.61877-8 5.9.58.31878-9 ■5.9.55.41879-80 ■5.863.31,880-1 ■5.0,58.61881-3 5.0,57.31883-3 4.7-59.91883-4 5.8.56.31884-5 73.8|.58.5il88.5-6 ■5. 5155. 51886-7 ■4.8.55.51887-8 3.8.56.11888-9 5.3 57.41889-90 ■2.61.56.31890-1 •6.2.55.61891-2 '5.0-55.6 l,s92-3 ■4.7.56.91h9,3-4 ■4.9.57.41894-5 ■5.3.57.7.1895-6 3. 9 .57. 9: 1896-7 ■6.71.58.31897-8 ■6.1.57.41f*9,S-9 ■7.6:61.71899-00 ■6.7 55.51900-1 4.4.59.31901-2 2.5(56.61902-3 29.7i52.5l....l 1903-4 Ten- rear Means. 57. 4|38. 3.55. 8177. 2'58. 5 1821-1830 .55.0!35.] .^4.0|75.2.55.8 1S31-1840 .55.;!:i6.4.52.9 75.3 56.51,«41-18.50 .55.,s;iti.:.'.-3.,S:75.7.57.7).S,5l-1860 56.0,35.9,53.7 76.4.58.01861-1870 .55.836.4,53.7 55.336.0.53.7 .55. 134.9 i3.0 55.936.4 .55.335.5 55.636.0 54.0 .^3.2 53.7 76.456.7:1871-1880 74.8157. 81881-1890 75.3.57. 5J1891-1900 76.057.2 1817-1870 75.457.31871-1903 5. 8 57.31817-1903 Table XIX shows the mean monthly, seasonal and annual temperature for Baltimore for 88 years from 1817 to 1904. The table contains three distinct records: (a) observations from 1817 to 1824, by Capt. Lewis Brantz, in what •was then west Baltimore; (b) observations at Fort McHenry, along the Baltimore harbor, from 1831 to 1870; (c) observations under the auspices of the United States Weather Bureau from 1871 to 1904. MARYLAXD WEATHER SERVICE 95 The Lewis Brantz observations were reduced to the Fort McHenry series by applying corrections derived from an overlapping period in 1836 and 1837. The monthly means for the period from 1825 to 1830 were derived from Washington, D. C, observations, and reduced to the Fort McHenry series by adding the departures from the Washington, D. C, normal temperatures to the Fort McHenry normal. The means for the years 1859 to 1863 were re- duced to the Fort McHenry series in a similar manner by means of a 20-year record of overlapping observations made at Shellman's Hills, about 20 miles west of Baltimore. The record from 1817 to 1870 was then made conformable to the Weather Bureau record from 1871-1903 by means of departures derived from overlapping records covering a period of about 20 years. Thus the entire record from 1817 to 1^03 may be regarded as an approximately uni- form series of Baltimore City temperatures. THE XORMAL TEMPERATURE. The variations of the mean monthly, seasonal and annual temperature during the greater portion of the preceding century are discussed in succeeding pages, while the mean for each month, season and year since 1817 is published in Table XIX, together with the monthly, seasonal, and annual averages for each ten-year period, and for the entire 87 years. The variations in value of the ten-year averages are observed to be small, even in the case of the month of greatest variability. The maximum variability (5,3°), occurs in the month of March with a mean tempera- ture, for a ten-year period, as low as 40.8° and as high as 46.1°. The annual average for ten years has varied between the limits 55.0® and o7.-4° a range of 2.4°. Xo progressive increase or decrease is indi- cated for the entire period eitlior in tlic monthly, seasonal, or the annual means. From the third decade (1831-1840), there was a steady rise in temperature to the sixth (1861-1870), and since then a steady fall to the present time. The series of observations is not sufficiently long, however, to draw the conclusion that there is a periodic change of this length. In the absence of changes of long period in the fluctuations of tlie annual mean temperature the normal temperature derived from the 87 years will remain fixed at 55.0° for Baltimore. The probable error of this value is not greater than one-tenth of one degree Fahrenheit. Hence a longer series of observation will not cliange the result by an amount greater than onc-tentli of one degree. 96 the climate of baltimore The Vakiability of the Monthly and Annual Mean. In tabulating the following lists of exceptional months and seasons the sole basis of selection has been an average monthly temperature decidedly above or below the normal for the entire period of 87 years. Such lines of division must necessarily be arbitrary as there are no fixed standards of cold and warm. The degree of departure from the normal fixed upon for classification as cold or warm varied with the variability of the month and season. In the comparatively constant summer months a departure of 2° may be regarded as exceptional. In the variable winter months a departure of 6° may be assumed to be necessary to make the month an exceptionally cold or warm one. The seasons and the year being less fluctuating, the departures selected were smaller, varying from 2° for the year and the summer to 4° for the winter season. A close examination of this list of exceptional departures from the normal will doubtless cause surprise by the absence of periods which left an impression of great heat or cold. Attention has already been called to the fact that an average temperature for the period of a month or season is sometimes an inadequate measure of the temperature conditions of the period. A month with 10 consecutive days of excessively hot weather for example, will long remain in memory as a hot month, no matter what the average temperature of the entire month may be. Yet a moderately cool spell preceding and following the hot days will result in an average value for the month little if any above the normal and hence would not be found in a list of warm months. The month of Jul)', 1898, may be cited as a case in point. The highest temperature recorded in the official records of the Baltimore station oc- curred on July 3, 1898, namely 104°. The month contained 10 days with a temperature of 90° and over. Yet this month is not listed as a warm month because the average temperature for the entire month was less than 2° above the normal. The proper place to look for such excessively hot spells is in the list of warm days rather than warm months. As a rule, however, the average temperature is a safe guide for expressing the gen- eral temperature conditions of a given period. MARYLAND WEATHER SERVICE 97 Warm Months and Seasons. The following list includes the months and seasons since 1817 during which the average temperature rose decidedly above the normal in the vicinity of Baltimore. The degree of departure required for each month and season in order to find a place in the list is shown in the column headed " Departure." WARM MONTHS AND SEASONS. De- parture. January 6°+ 1824,1828,1843, February.... 6°+ 1820,1827,1828, March 6°+ 1825,1826,1812, April 5°+ 1817,1822,1823. May 4°+ 1822, 1820, 1833, June 4° + 1828, 185S, ISti;'), July 3°+ 1822, 18.30. 1S31, August 3°+ 1819, 1821, ls:.'U'. September.... 4°+ 1822, 182t;, istio. October 4°+ 1855, 187'J, issi, November.... 4°+ 1818,1822,1830, December 5°+ 1824, 1827, 1829, Winter 4° + Spring 3° + Summer 2° + Autumn 3° + Year 2° + 1870, 1876, 1880, 1890. 1834, 1857, 1884, 1890. 1859, 1865, 1878, 1903. 1827. 1865. 1848, 1864, 1865, 1880, 1896. L';70. 1.S.38, 1868, 1870, 1872. 1S27, 1S28, 1870, 1872, 1900. ISSl, ]!)00. lss:.>, 1SS4, 1900. 1849, 1850, 1866, 1896, 1902. 1848, 1857, 1877, 1881, 1889, 1891. 1823-4, 1824-5. 1827-8, 1849-50, 1850-1, 1857-8. 1869-70, 1879-80, 1889-90. 1822, 1826. 1827, ls;il, 1833. 1865, 1871, 1878, 1903. 1819, 1821, 1822, lS-7, 182.S, 1830, 1838, 1868, 1870, 1872. 1822, 18:30, 1854, 1.S55, 1881, 1900. 1822, 1825, 1826, 1827, 1828, 1830, 1865, 1870. The years 1822, 1827, 1828, 1857 and 1870, are conspicuous in the list for sustained warmth during several months of the year. During 1822 there were five months of the year with an excessive departure above the normal. During one month only, namely January, was the temperature below the normal. The year attained the highest mean annual temper- ature on record, namely 3.2° above the normal. COLD MONTHS AND SEASONS. De- parture. January 6° + February . ... 6*-t- March I>° + At»ril 5°-i- May 4° + June 4°-l- July 3° + Au(fU8t 3° + September 4° + OctobiT 4° + November 4'+ December 5° + Winter 4° + Spriniif 3"+ Summer 2° + Autumn 3° + Year 2"+ 1821, 1829, IKlf,, 1821, ]K2(l, iKii;, lS2'.t. 183t;, ]8:t.^., 1819, 1820, 1831. 1840, 1836, 1843, 18(1, 1S3S, IS-Ki, l.SKI, ISlll, 1S4(), 1820, 18:16, 1840, 1856, 1838, 18.56, 1857, 1841, lst;2, IStil. lS6li, isti:i, lKi4, 1S38, 1845, 1857, 18.56, 1872, 1S74. 1843, 190:t. 1S62, l'.t03. 1S71. IKje. 1839, 1872, 1858, 1867, 1893, 1904. 1875, 1885, 1895, 1899, 1901, 1902, 1904. 1885. 1861, 1882. 1886, 1888, 1891, 1895. 18,38, 1841, 1844, 18.59. 1842, 1844, 1873, 1880, 1901. 1876, 1880, 1886. 185.5-6, 1866-7, 1892-3, 1894-5, 1901-2, 1903-4. 1836, 1841, 184:3, 1857. ]83t!, 1K42, 1H46, 1861. 1862, 1886, 1889, 1891, 1903. 18:36, 1838, 1841, 1842, 1844. 1836, 1841, 1863, 1875, 1886, 1893. 98 THE CLIMATE OF BALTIMORE The year 1836 occurs most frequently in the list of cold months and seasons. The average temperature for the entire year was the lowest in the record of 87 years. It also contained lowest average August and Oc- tober temperatures^ and the lowest summer and autumn averages. Jan- uary alone was above the normal, and this but l.-l° above. The tempera- ature was decidedly below the normal during nine months of the year. The summers of 1903 and 1886 follow close behind the memorable sum- mer of 1836. The recent winter of 1903-04: with an average temperature of 29.7° was the coldest experienced in Baltimore. jSTo excessively low temperatures were recorded, but the entire season was characterized by an almost unbroken period of moderately cold weather. There was an almost total absence of the usual and sometimes frequent " thaws " of previous years. The ice crop was the heaviest in many years. Navigation on the Bay was impeded to an unprecedented extent. The Bay was frozen from shore to shore to a distance of over 80 miles south of Baltimore. WARMEST AND COLDEST MONTHS. (Expressed in terms of departures from the normaL) Normal — Warmest + Year Coldest — . . Year Range Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec 34.9' +9.0 1858 -10.6 1893 19.6 35.8° +10.5 1834 —9.6 1895 20.1 43.1 + 7.' 1865 -11.9 1843 19.1 54.1° +7.8 1817 -7.1 1874 14.9 64.0° +8.3 1826! -4.9 1843: 20.5 73.3° +5.5 1870 —5.8 1903 11.3 77.9° +5.61 1870 — 6.3| 1891 1 11.9! 76.1° +4.9 1831 -5.1 1836 10.0 +8.6 1881 -6.5 1835 1.5.1 56.6° 46.6 +7.3 +8.0 1855 1849 —6.7 1836 16.1 1842 14.7 37.4* +8.2 1829 -10.4 1831 18.6 WARMEST AND COLDEST SEASONS AND YEARS. Winter. Spring. Summer. Autumn. Year. 36.0° +7.9 1889-90 -6.3 19a3-4 14.2 53.7° +5.5 1865 -8.9 1843 14.4 75.8° +5.2 1870 -4.3 1836 9.4 57.3° +6.0 1881 -4.2 1836 10.2 55 6° +3.2 18'^2 Year Coldest— Year -3.8 1836 Range 7 Winter and spring have varied most from the normal, the difference be- tween the warmest winter (namely 43.9° in 1889-90) and the coldest winter (29.7° in 1903-04) is 14.2°. The spring limits are -f 5.5° (1865) and —8.9° (1843), a range of 14.4°. The summer limits are + 5.2° MARYLAXD WEATHER SERVICE 99 (1870) and —4.2° (1836), a range of 9.4°. The autumn limits are + 6.0° (1881) and — 4.2° (1836), a range of 10.2°. A warm February may have the average temperature of a normal March, or approach that of a cold April, or cold October. A warm Sep- tember may have the average temperature of a normal July or August. October has been nearly as cold as a warm February. May has been as warm as a cold July. A month may have the same mean temperature as the second preceding or following month. Hence a season may be one month later or earlier than the average time, or, there may be two months difference for example, between a very late spring and a very early spring. Frequency of Stated Departures fro:m the Monthly Seasonal AND AxxuAL Mean Temperatures. During a period of 87 years the mean annual temperature was below the arithmetical average in 45 per cent of the total number of years, above the normal in 49 per cent, and exactly normal (within one-tenth of a degree Fahrenheit) in 6 per cent. the distribution of seasonal depart cres. Above Normal. Below Normal. Normal. _ artures ( — '' 519 No Departures (0) 14 In most months the departure is likely to be about 1° above or below the normal; in April the most probable departure is 2°, in October 3°, and in February above 4°. Fifty-two per cent of the mean annual temperatures have fallen within 1° of the normal value in the past 87 years, and in 37 per cent of the remaining years the mean was within 2° of the normal. No annual mean has risen to 4° above the normal and none fallen 4° below. Hence the annual mean temperature has a comparatively small range of depart- ure from the normal. The extreme departures occurred in 1822 (3.2° above normal) and in 1836 (3.8° below), an extreme range of 7° between MARYLAXD WEATHER SERVICE 101 the coldest and warmest years on record at Baltimore. The frequency of departures of stated values for the seasons is shown in the following table, in percentages of the total occurrences in 87 years: Fig. 22. — Frequeucy of Stated Departures from the Monthly Normal Temperature. Fig. 22 shows the frequency of stated departures from the normal value of the monthly temperatures, based on records covering- 87 j-ears. The upper line of figures represents the degree of departure above (+) or below (— ) the normal monthly temperature. The mar- ginal letters represent the months of the year. The curved lines and shaded areas represent the frequency of the changes expressed as percentages of total number of months. In- crease in intensity of shading i-eprosents increase in the frequency of stated changes. For e.xample, changes of + 3 or —2 occurred in 10 per cent, of the total number of instances in March, 20 per cent, in May and August, 10 per cent, in December, etc. See also Fig. 23. FREQUENCY OF STATED SEASONAL DEPARTURES. Winter. . Spring... Summer. Autumn r 2° 3- 4° 5° 6° 7° 8° 9° % :% % % % % « % % 19 29 n 19 10 2 32 27 26 8 4 2 1 46 32 14 6 1 1 3:3 34 20 8 4 1 13 winter SprluK Summer Antumii Average The arithmetical average departure 3.4' The most frequent departure 2.0'^ 2.6' 1.0' 1.7' 1.0' 2.3' 2.0' 2.5° 1.5" The PROB.A.BLE Error of the ^Monthly and Annual Means. Employing the formula given in a preceding paragraph for the deter- mination of the probable error of the daily mean temperature in order to 102 THE CLIMATE OF BALTIMORE arrive at the probable error of the monthly and annual means for the series of 87 years we obtain the following values : AVERAGE MONTHLY AND ANNUAL DEPARTURES AND PROBABLE ERROR OF THE MONTHLY AND ANNUAL MEAN TEMPERATURES. Average de- parture (V) Probable error (E) Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 3.4° 3.9° 3.1" 2.5° 2.1° 1.8° 1.6° 1.6° 2.0° 2.4° 2.5° 2.9° .32 .36 .28 .23 .20 .16 .15 .15 .18 .21 .23 .26 Year 1.1° .10 / \ / \ / c^ \ / 'Vl ii. / .' " v^ y / \- \ 1 / ' I b\ \ o\ / \ \ k. \ / \S ^?r^ i^ / SEASONAL- A ANNUAL- B Fig. 23. — Frequency of Stated Departures from the Normal Seasonal (A) and Annual (B) Temperatures, ia) Summer, (&) Autumn, (c) Spring, (d) Winter. Fig. 23 shows the frequency of stated departures from the normal seasonal (A) and annual (B) temperatures. The upper horizontal row of figures indicates the degree of change, and the vertical rows to the right and left of the diagi-am indicate the frequency of change, expressed as percentages of the total number of seasonal and annual changes from 1817 to 1903. The probable error of the mean annual temperature is about one-half that of the mean monthly temperature ; the probable error of the latter is MARYLAND WEATHER SERVICE 103 about one-fourth that of the mean daily temperature. Hence the prob- able error of the daily mean is eight times as large as that of the annual mean. SUCCESSIOX OF THE SEASONS. It may be definitely stated, as the result of a study of the Baltimore temperature observations for 87 years, that there is no regular periodic recurrence of cold and warm periods. Some interesting questions in prob- abilities are, however, suggested by a classification of the decidedly cold and warm seasons, in connection with the character of succeeding seasons. This has been done in the following manner : A winter was regarded as cold or warm when the departure from the average winter condition equalled or exceeded 3° ; when less than 2° it was considered a normal season. The point of departure for the spring and autumn was 1.5°, for the summer and for the year 1.0°. This classification yielded from 20 to 25 abnormal seasons of each class during a period of 87 years. For each of the abnormal winters the character of the succeeding spring, sum- mer, autumn and winter was then noted. Abnormal summers and ab- normal years were tabulated in a similar manner, with the following result : SUCCESSION OF THE SEASONS. Cold Winters (-2.0°+) 23 Warm Winters (+2.0* + ) 22 Summer (±1.0°+). Autumn (±1.5°+). Winter (±2.0°+). ? be > §1 6 11 Cold Summers (-1.0°+) .. . Warm Summers (+1.0° + ) .. . Cold Years (—1.0°+). I 20 Warm (+1.0°+).; 24 Autumn (±1-5° + ). Winter (±2.0° + ). Spri ng (±1.5°+). Summer (±1.0° + ;. 1.5 Year. 104 THE CLIMATE OF BALTIMORE Of 23 cold winters, 10 were followed b}' cold springs, 13 by average springs, and not one by a warm spring. Nine were followed by a cold summer, 11 by an average summer and only 3 by a warm summer. The succeeding winter Avas cold in 6 cases, average in 9 and warm in 8. These figures show a decided probability in favor of a cold or cool spring and summer following a cold winter, and of a cool autumn; there is no decided tendency as to the character of the succeeding winter. A similar tendency is shown in favor of a warm spring and summer after a warm winter. There is also a decided probability that a cold summer will be followed by a cold or a cool autumn, winter, spring, and next succeeding summer, and that a warm summer will be followed by a warm or an average autumn, winter, spring, and next succeeding summer. ]\Iore cautious, and perhaps safer, is the negative statement that a cold winter is not likely to be followed by a warm spring or summer; that a warm winter is not likely to be followed by a cold spring and summer. In general it may be said that an extreme season is not likely to be followed by an opposite extreme. Such a conclusion may not be regarded as of much practical value for determining the probable character of a coming season, but a more definite statement does not seem to be war- ranted by the statistical record. Considering the average temperature for the entire year there were 20 cases of a departure of 1° or more below the normal. Of these 8 were followed by cold years, 11 by years of an average temperature, and 1 by a warm year. Of 24: M'arm years 3 were followed by cold years, 11 by average years, and 10 by warm years. Here again the same tendency is shown against the occurrence of a succession of years of opposite character. That is, a decidedly cold year is not likely to be followed by a decidedly warm year, or a warm year by a cold year. All such classifications are, however, arbitrary and inferences as to the succession of the seasons should be accepted with caution Avhen based upon phenomena as variable in their nature as the climatic factors of the middle latitudes. Daily Exteemes or Temperature. Thus far only average temperatures for a day, month, or year, have been considered, with departures from the normal conditions based on many MARYLAXD WEATHER SERVICE 105 years of observations. The variability of a given climate is best illus- trated, however, by extremes of temperature within given limits of time, and by the frequency of occurrence of stated changes from day to day. TABLE XX.— DAILY EXTREMES OF TEMPERATTRE. (Spring.) Date. March. B ® i- ** ' April. O its May. ® . 1 72 1895 14 1884 2 69 1882 15 1886 :i 68 1871 12 1873 4 74 1880 5 1873 5 76 1880 9 1872 6 72 1894 13 1901 7 72 1878 12 1890 8 a5 1878 12 1873 9 67 1871 21 1885 10 70 1897 17 1877 11 71 1879 20 1892 12 76 J890 12 1900 13 75 1890 12 1888 14 66 1903 14 1888 15 '71 1886 21 1900 16 68 1886 18 1893 17 72 1898 15 1900 18 68 1894 9 1877 19 78 1894 12 1876 20 69 1903 12 1885 21 72 1897 12 1885 22 82 1894 19 1885 23 71 1871 16 1888 24 66 1903 18 1896 25 70 1904 21 1878 26 66 1896 21 1878 27 69 1903 20 1894 28 77 18!K) 24 1894 29 77 1902 23 1887 30 f.9 1896 21 1887 31 74 1888 29 1873 78 1893 30 82 1882 30 80 1892 29 83 1892 29 77 1880 25 75 1892 26 75 1890 30 85 1871 32 82 1871 32 8t 1887 32 85 1887 30 76 1899 27 83 1890 29 85 1896 34 82 1891 32 86 1896 36 89 1896 27 94 1896 26 93 1896 24 87 1896 27 1887 1876 1899 1904 1881 1898 1898 1896 1885 1894 1882 1874 1874 1885 1904 1893 1875 1875 1875 1904 S3 1896 32 1875 89 1902 32 1875 88 1902 36 1875 88 1886 38 1888 86 1895 34 188:3 88 1872 37 1883 82 1891 39 1893 81 1888 34 1898 90 1888 33 1874 91 1903 34 1874 48 87 1890 34 1876 87 1894 37 1903 84 1878 38 1882 86 1892 40 1900 84 1896 42 1875 86 1880 40 88 1872 44 89 1900 42 93 1896 40 96 1896 42 94 1896 45 94 1881 44 95 1881 40 91 1900 46 94 1900 42 87 1900 43 93 1896 44 92 1896 46 92 1877 47 92 1903 48 1891 1882 18ii8 1898 1900 1877 18a5 1895 1895 1895 1904 1891 1895 1895 1899 88 1893 44 1895 88 1903 42 1895 90 Mt02 47 1892 89 1884 46 1893 90 1880 46 1877 92 1880 4u 1886 93 1880 48 1897 90 1899 46 1902 89 1895 43 1894 95 1895 46 1884 95 1895 50 1873 53 50 46 46 42 46 44 47 53 54 49 50 55 45 52 44 49 46 45 44 44 46 43 43 44 47 45 44 46 49 45 Table XX shows the highest and lowest temperatures recorded on each day of the year during 34 years from 1871 to June, 1904, with year of occur- rence, and the extreme range for the day. In Table XX, and in curves A, C, and D of Plate IV, the highest and lowest temperatures officially recorded upon each day of the year dur- ing a period of 33 years are shown, together witli the absolute daily range. In addition the table shows the year of occurrence of the extremes. A studv of the curves of Plate V will most clearly and quickly reveal the ex- tremely changeable character of the temperature from day to day, and the 8 106 THE CLIMATE OF BALTIMORE relative variability of the seasonal changes. The changes in the average temperatures from day to day throughout the year have already been described in preceding sections of this report. As the average tempera- tures were derived from the daily extremes^ there must of necessity be a general agreement, with a difference only in the amplitude of change. Table xx Con't.— DAILY EXTREMES OF TEMPERATURE. (Summer.) June. July. August. ^9 u S) be bi u « be u ^ a, bo & a i- c c8 fl es u c es c -^ .-• 7i 93 « V •- o t^ cS O o ■^ a Date. s >* g t>^ B S >< g >* S fcH S ^ J< u 1 97 1895 47 1894 50 103 1901 56 1885 47 95 1890 57 1895 38 95 1895 48 1897 47 103 1901 59 1891 44 92 1879 58 1875 34 3 97 1895 52 1888 45 104 1898 .59 1888 45 92 1881 59 1895 33 4 91 1890 53 1888 38 100 1898 59 1891 41 94 1888 58 1886 36 5 93 1899 52 1886 41 96 1881 58 1892 38 96 1896 59 1874 37 6 98 96 1899 1899 47 47 1894 1894 51 49 96 96 1901 1900 58 60 1891 1891 38 36 97 lOo 1900 1900 60 62 1894 1897 37 38 8 96 98 1874 1874 47 50 1891 1891 49 48 98 99 1890 1876 55 56 1891 1891 43 43 99 100 1900 1900 58 62 1903 1887 41 38 10 90 1879 52 1904 38 97 1880 56 1894 41 100 1900 56 1879 44 11 92 1893 50 1904 42 96 1876 57 1898 39 100 1900 58 1879 42 12 95 1880 52 1887 43 96 1876 .■)i 1895 39 99 1900 60 1890 39 13 94 1902 53 1903 41 99 1880 57 1888 42 98 1881 .56 1902 42 14 95 1885 53 1873 42 95 1887 58 1895 87 96 1872 Oi 1893 39 15 93 1891 53 1884 40 96 1900 57 1895 39 90 1900 59 1887 31 16 94 1891 52 1884 42 101 1887 59 1903 42 9h 1SS8 58 1889 38 17 93 1887 55 1899 38 100 19(K) 59 1892 41 9l' 1900 55 1902 37 18 94 1887 54 1879 40 102 1887 60 1892 42 yj 1900 60 1874 31 19 93 1893 56 1886 37 96 1878 61 1890 35 9:- 1872 59 1896 34 20 98 1893 00 1879 43 98 1885 57 1890 41 97 18i.9 54 1896 43 21 93 1896 56 1897 37 99 1885 55 1890 44 97 1899 55 1876 43 22 94 1888 54 1897 40 96 1899 56 1890 40 96 1872 56 1876 40 23 97 1894 55 1898 42 95 188:^ 59 1890 36 9:- 1898 00 1888 38 24 98 1894 54 li^02 44 95 1884 59 1876 ;w 94 1898 51 1890 43 25 98 1898 55 1902 43 97 1892 59 1876 38 97 1903 56 1879 41 26 97 1875 61 1893 36 99 1892 61 1891 38 96 1900 52 1874 44 27 95 1876 Oi 1893 38 97 1892 59 1876 ■M 92 1900 53 1885 39 28 94 1898 56 1897 38 97 1894 59 1893 38 9J 1895 53 1885 38 29 97 1874 55 1888 42 97 1892 62 1897 35 94 1877 52 1874 42 30 99 1901 57 1899 42 95 1903 6(1 1880 35 91 1898 54 1896 37 31 95 1890 55 1895 40 95 1898 55 1887 40 The greatest variability in extreme conditions occurs in the winter months, with a gradual decrease to more uniform conditions toward summer. The greatest change of temperature which has been recorded within a period of 24 hours during 33 years at Baltimore is 47°. This remarkable range between the highest and lowest temperature of a single day occurred on the 24th of February, 1900. When we consider extremes which liave MARYLAND WEATHER SERVICE 107 occurred upon a given date, without reference to the year of occurrence, the range is greatly increased. For instance, upon the 11th of February a maximum temperature of 72° was recorded in 1887, and a minimum of 6° below zero in 1899, a range of 78°. Even in the months of least vari- Table x.v Con't.-DAILY EXTREMES OF TEMPERATURE. (Autumn.) Date. September. October. u c X « c c3 ai e S tH S >H November. ] 95 1898 2 96 1898 3 97 1898 4 i 9L 1898 5 1 91; 1880 6 94 1900 7 101 1881 8 94 1873 9 i 94 1894 10 1 98 1884 I 11 97 1897 12 9:1 1895 13 93 1897 14 89 1903 15 92j 1901 16 89 1897 17 90 1886 18 91 1898 19 94 1896 20 5)0 1895 21 96 1895 22 96 1895 23 95 1895 24 87 1881 25 90 1881 26 93 1895 27 90 1881 28 91 1886 2« 87 1884 30 1 88 1881 31 ! ! 56 1887 50 1892 51 1893 50 1872 50 1872 51 1883 51 1883 55 1892 51 1891 46 1883 49 1875 51 1879 53 1902 48 1902 40 1873 1873 1887 1875 1875 1875 45 1897 44 1873 46 1896 43 1875 42 1887 40 1879 43 1879 44 1899 43 1903 39 1888 89 1881 39 88 1881 38 89 1879 36 87 1884 38 85 1884 42 89 1884 36 81 1884 40 82 1887 38 84 1893 37 85 1887 35 79 1898 40 8J 1889 33 80 1884 35 »-2 1883 34 83 1897 ! 33 90 1897 30 »-Z 1879 36 84 1881 36 76 1899 35 76 1884 36 77 18.><4 39 78 1901 34 81 1901 36 79 1900 34 80 1903 33 77 1891 30 75 1899 35 77 IS99 'M 78 1874 31 75 19113 30 77 1896 . 31 1899 1899 1899 1888 1901 1893 1893 1876 1896 1895 1881 1876 1876 1875 1876 1876 1893 1876 1880 1900 1900 1899 1889 1889 1879 1879 1903 1898 1873 1873 1893 74 1903 31 1873 76 1876 33 1873 75 1903 33 I 1875 75 1903 38 1879 73 1896 25 1879 74 1888 31 1892 68 1896 28 1903 73 1890 38 1886 77 1895 29 , 1886 74 1879 31 1874 73 1899 31 1901 78 1879 27 1894 76 1902 28 1873 68 1889 I 26 , 1873 75 1902 28 1883 75 , 1897 75 1896 71 1896 74 1900 68 1900 79 1900 71 1883 68 1900 70 1896 65 1890 71 ' 1896 74 1896 71 1896 61 1879 62 1899 23 1883 25 1883 23 1891 21 1891 33 1879 21 1879 15 1880 16 1880 17 1880 24 1881 21 1903 18 ltt03 30 1903 23 1875 16 1875 43 44 43 47 47 43 40 45 48 43 41 51 48 42 47 ability the range is still largo. Tlio smallest range, namely 31°, is cred- ited to August 14 and 18. The higliest temperature of the year occurred on July 3, 1898, and the lowest on February 10, 1899. Although we have no systematic and official records of daily cxtromos of tomporature prior tc 1872, when self-registering ma.ximum and minimum tlicniidmeters were added to tlic (■(iiii|iiii('iit df the V. S. AW'atlicr liiireau stations, we have good reason to believe that the period from 1872 to 1903 comprised 108 THE CLIMATE OF BALTIMORE within its limits the warmest and coldest days experienced ar Baltimore. In the summer of 1898 all existing records of high temperature were broken during the hot spell of July 1-4 within the State of Maryland. In the following winter during the first decade of February all existing Table xx Con't.-DAILY EXTREMES OF TEMPERATURE. (Winter.) Date. December. 1881 1901 1874 1873 1883 1879 189B 1892 1889 1897 1897 187.3 1889 1881 1893 1877 1877 1877 1900 1877 1885 1889 1891 1893 1893 18S9 1881 1889 1893 1898 1-84 1875 1886 1-86 1886 1871 16 1901 15 1885 10' 1882 4 1.S76 1 1876 13 1880 22 1895 14 1895 14i 1898 s ® i^ C 15 1900 13 1876 h 1876 12 1875 1(1 18S4 r 1871 5 1871 6 1872 17 1872 8 1872 8 1872 12 1903 11 1903 13 1,H72 4 1880 -3 1880 — 1 1880 January. 61 1885 1876 1876 1874 1890 1890 1874 1898 1876 1876 1891 1890 1890 1892 1871 1901 1885 1876 1876 1880 1901 1874 1874 1894 1879 1878 1890 1876 1876 1896 1880 1881 6 1899 1879 5 1879 1877 1904 1884 1878 6 1875 -2 1875 1875 1886 1886 1886 1893 1893 1893 1893 1904 1901 1893 1893 1883 1882 1897 1897 1888 1888 1873 1878 1873 13 o be n a 65 65 February. ii u cd cS 0) e § ;>H S ^ 63 1891 8 1900 58 1877 4 1881 61 i 18813 66 I 1903 1895 1886 71 , 1890 I —1 1886 68 < 1884 64 : 1904 57 : 1892 58 1,H78 1876 1887 1898 66 1903 63 1884 67 j 1886 67 1891 73 1S91 71 I 1891 60 I 1887 60 1887 72 I 1874 74 ' 1874 II 1895 61 1895 2 1S95 2 1S99 —7 1899 —6 1899 5 1899 6 1899 6i 1899 6 1899 1875 1896 1903 1903 1896 1H74 1872 1871 1890 18>h0 1903 1880 15 1885 1896 1889 1873 1900 1900 1900 1888 1884 C 4) o be S a records of great cold were lowered. The variability of temperature con- ditions during the past five or six years has been phenomenal. Eecords of extremes of temperature which have remained undisturbed for many years were broken to a remarkable extent. In addition to the instances of absolute extremes of temperature just referred to, may be mentioned the high monthly average temperature of March, 1903, the warmest March in 87 years or more; the summer of 1903, the coolest in 87 jears or more; and the following winter of 1903-04 which was the coldest in a hundred years. MARYLAND AVEATIIER SERVICE 109 ABSOLUTE EXTREMES OF TEMPERATURE 1871-1903. Absolute Absolute Absolute Max. Year. Day. Min. Year. Day. Range. 73" 1889 26 - 3° 1880 30 76° 73 1890 13 — 6 1881 1 '•.9 78 1874 23 — 7 1899 10 85 82 1894 23 5 1873 4 77 94 1896 18 24 1875 19 70 96 1896 10 34 1876 1 62 99 1901 30 47 1891 8 52 104 1898 3 55 1891 8 49 100 1900 10 51 1890 24 49 101 1881 7 39 1888 30 62 90 1897 16 30 1876 16 60 79 1900 21 15 1880 22 64 78 1874 Feb. 23 y 1899 Feb. 10 85 96 1896 May K 5 1873 Mar. i 91 104 1898 July 3 47 1891 .Tune 8 57 101 1881 Sept. 7 15 1880 Nov. 22 86 104 1898 July 3 — 7 1899 Feb. 10 111 December. January ... Febi'uary . . March. ApriL. May . . . June July August. .. September. October — November. Winter. . . Spring . . . Summer . Autumn . Year. Jan Feb Mcm Apr M«» June Jui» Aug Sept Oct Nov Dec j«n 45* 40' 35* 30' 25 20 I? 4 5° 4cr 35° 3(f 25* 20 10' 0* Fig. 24. — Greatest Daily Raiitje of Temperature. (See Table XXL) The Greatest Daily Eange of Temperature, In Table XXI the gi-eatest difference between the daily maximum and minimum temperatures, or the greatest daily range, is entered for each 110 THE CLIMATE OF BALTIMORE month and year from 1871 to 1903, together with the daily average range for each ten-year period. There is a marked uniformity in the size of the daily ranges throughout the year, the average monthly values ranging be- TABLE XXI.-GREATEST DAILY RANGE OF TEMPERATURE. 1871. 1873. 1873. 1874. 1875. 1876. 1877. 1878. 1879. 1880. J881. 1882. 1883. 1884. 1885. J886. 1887. 1888. 18S9. 1890. 1891. 1892. 1893. 1894. 189,5. I89f). 1897. 189S. 1899. 1900. 1901. 1902. 1903. 1904. 20 Means. 1871-1880 1881-1S90 1891-1900 1871-1902 27.5 36.3 23.9 35.8 19 28 32 37 33 29 33 23 28 24 24 37 29 26 36 40 28 .30 35 34 27 31 36 28 39 31 22 37 35 47 33 19 28 20 37.5 38-9 39.3 28.1 29 25 33 28 30 29 29 24 28 32 34 38 36 26 37 39 38 34 34 33 34 38 35 39 30 32 38 29 32 31 34 3:3 40 30 20 29 30 31 31 29 25 37 a3 30 25 29 33 35 37 30 24 24 30 29 30 30 30 30 31 39 31 30 28 35 30 33 29 33 17 30 31 38 26 28 36 23 30 26 23 25 oo 38 37 33 29 38 35 36 37 26 29 37 24 35 30 27 31 29 29 26 37 29.1 37. 6i 38.5 35.5 37.4! .30.91 37.5! 25.8 31.1 31.8! 31.4 27.2 29.3 30.3! 29.31 26.3 33.4 24.8 26.3 25.0 27 36 36 38 30 22 36 26 29 22.8 23.5 26.2 24 36 24 38.6 36.5 23 24 25 29 38 27 26 25 25 23 24 23 32 38 23 28 28 25 29 21 33 33 33 30 35 27 30 28 26 31 35 27 30 25.5 35.1 30.4 37.2 24 26 24 28 24 28 27 27 37 27 81 25 26 26 27 30 31 26 24 39 31 28 28 26.1 36.3 37.6 36.8 31 36 23 42 26 29 39 23 28 25.4 26.1 26.5 29 35 33 31 33 38 33 35 33 33 34 31 33 30 42 40 38 34 34 33 .34 36 35 35 35 33.6 34.8 37.3 35.3 Table XXI shows the greatest daily range of temperature (the greatest ■difference between the maximum and minimum of any day) for each month and year from 1871 to 1904. Also the average of the greatest daily ranges for the entire period of 32 years and for each 10-year period. tween 30.3° for April and 24.2° for August. The extreme daily ranges for each month and for the year are shown in the following tabular state- ment and in Fiff. 24. MAEYLAXD WEATHER SERVICE 111 EXTREME DAILY RANGE OF TEMPERATURE. January — Februarj'.. March April May June July August September , October November.. December. . Year. 1871-1903. Day. Year. Range. Max. Min. 17 1885 42=' 65 23 34 1900 47'= 55 8 7 1873 , 36° 50 14 4 1903 40° 71 31 9 1896 39° 93 54 11 1900 31° 92 61 18 1887 1 32° 102 70 6 1900 ' 30° 97 67 16 1873 38° 79 41 19 1901 35° 75 40 13 1876 31° 68 37 31 1898 42° 59 17 Feb. 1900 1 ''' 55 8 J % Fe e M H A n M Af JU NE JU LV Al &. Se PT :t Nf )V D C. Ja 100° 90° / \ / \ 8(f ^ / y ^ "\ \, ^ r^ / ^^- ^ "^ s. > 7d / y N. \ \, / / ^^ ^ \ s V 60' -7 '-/ Y ^ \ N d 50' 40° b ^ z 7^ ^ ^ \ -A \ <^ V^ ^ 30' -r- ^ 7- :^ \ V 20° -f ^ 10° 0* / J- \ • y \ lo' Fig. '.^.5. — (a) Extreme .Moutlily Ma.ximum Temperature. (h) Mean '• " -i fe g 63 56 58 69 52 71 54 57 64 65 45 £9 50 62 65 57 65 .50 60 73 60 58 52 57 60 61 62 78 59 65 63 63 58 67 64 59 64 68 50 67 72 60 48 74 73 .57 61 19 62 61 56 65 60 65 Means. 1871-1880 60.9 64.2 1881-1S90 57.6 62.6 1891-1900 58.7 61.9 1871-1903 i 58.9| 62.3; 69.2 6' 70.0 91 78.7 82.8 83.8 82.1 Table XXII contains the highest recorded temperature during each month and year from 1871 to 1904, together with the average of the monthly ex- tremes for the entire period of 32 years and for each 10-year period. The observations were obtained by means of a self-registering mercurial ther- mometer, excepting for the time from January, 1871, to July, 1872, during which period the 3 p. m. observations were employed. Monthly axd Axxual Extremes. The highest and lowest temperatures recorded during each month and year from 1871 to 1903 are shown in Tables XXII and XXIII, together with the average values for each ten-year period and for the entire period. MARYLAND WEATHER SERVICE 113 The absolute monthly extremes of temperature, the average maximum, and minimum, and the monthly normals are shown in Fig. 25. Fig. 26 shows the absolute annual extremes and the mean annual temperature for each year from 1871 to 1903. TABLE XXIII.-MONTHLT MINIMUM TEMPERATURES. 1871. 1873. 1873. 1874. 1875. 1876. 1877. 1878. 1879. 1880. 1881. 1882. 1883. 1884. 1885. 1886. 1887. lang. 1889. 1890. 1891. 1892. 189;^. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901. 1902. 1903. 1904. Means. 14 11 —4 13 —3 17 1 6 17 10 15 15 i 13 18 20 12 15 4 23 23 10 3 — 1 21 11 3 33 16 14 11 8 1 6 18 10 —7 8 1871-1S80 7.3 KWl-lSflO 8.H 1S91-19UJ 11.1 1871-1902 9.6] 36 9 5 33 19 13 9 31 34 23 2T 26 16 14 12 15 31 12 38 12 16 30 16 20 21 16 28 27 36 12 14 13 13 30 5 , 29 5 I 20 42 38 38 37 24 30 33 43 39 30 25 29 30 34 32 34 30 33 34 31 30 32 36 30 34 31 29 26 29 30 37 35 27 27 51 48 44 41 43 34 41 43 43 38 46 38 45 45 44 45 51 41 43 43 40 46 45 43 40 47 44 40 47 40 47 43 37 43 13. S 18.0 33.2 42.5 11.9 18.3 31.2 44.1 8.4 20.2 30.7 43.2 ll.Oi 1».7| 38.0; 43.4! 53.8 r3.4 52.5 t3.2 61 55 58 55 45 50 40 53 43 45 48 47 40 50 J9 48 46 49 46 50 42 39 46 46 51 49 44 45 46 46 45 PS 42 60 46 48 43 62.4 58.6 60.1 56.5 57.9 57.7 67. 6[ 40 38 30 35 34 30 41 35 30 35 39 44 40 35 38 36 32 36 34 36 33 34 31 36 34 36 38 34 34 36 37 34 35 28 17 23 24 16 25 35 33 30 15 34 36 33 26 33 26 25 25 38 26 18 31 22 24 15 13 —3 34 10 17 9 15 15 16 16 23 18 17 14 18 6 —4 13 1 1 6 -3 —6 11 8 3 — 1 7 9 3 12 16 12 1 7 1 ■'1 46.1 34.8' 22.5' 47.1 37.0 26.1 47.0! 34.6 25.0 46.81 35.51 24.81 11.4 16.3 13.8 13.9 2.3 5.3 6.1 5.0 Table XXIII contains the lowest recorded temperature during each month and year from 1871 to 1904, together with the average of the monthly extremes for the entire period of 32 years and for each 10-year period. The observations were obtained by means of a self-registering alcohol min- imum thermometer, excepting for the time from January, 1871, to July, 1872, during which period the 7 a. m. observations were employed. 114 THE CLIMATE OF BALTIMORE The absolute range of temperature at Baltimore, the diflference between the highest (104°), and lowest (7° below zero) recorded readings of the thermometer, is 111°. The former occurred on July 3, 1898, and the latter on February 10, 1899. On both of these days records for extreme heat and cold were broken in many parts of the countr}'. 871 1875 t880 1885 189C 1895 1900 i9o: [ 4- ^ 1 : 1 1 j ] 1 i 1 . j 1 1 1 -4- 1 ] -- 1 1 ! i V 7/ ^ ^ -• i ' /«! 1 ^ 1 A 1 f ' V ; ' . J s / ^^ / s \ y s o^ / V , N / vi / ** h l" >.» N. sl / "L r j y t^ -- ^ ft 1 ■ 1 1 I , , 1 -- 1 1 1 ] ! ' ! 1 I ■ 1 1 ! i 1 1 1 ' 1 1 1 1 ' ' 1 ' 1 : 1 1 _• ' 1 ^ , ' ^ 1 i 1 ! ' j I 1 ' i 1 1 1 j ' 1 B . 1 1 J 1 1 __ ^ ^ =*K _^ ^ ^^ 4^ ^ ^tr J_. ^^ -k. _L^ -*- .:^ -^ ^ J^ Sr -_ __, [- -^ — ~ -^ -^--L- -^ ^ -*^ ^ — ~ -*- ■ r~ - , — -• 1 ^ I , \ ' ' ■ 1 , 1 ' 1 [ 1 [ ' ' 1 1 1 1 1 1 1 1 1 ' } ■ 1 1 ' 1 ' j ' 1 1 ! 1 1 1 1 ' i i 1 ' 1 1 ' t ' . I 1 i ' ! 1 1 1 1 j j ' . 1 j j 1 1 1 1 1 1 j 1 I 1 i 1 t 1 ] ! 1 1 1 1 j 1 1 ' j ' ' ' 1 1 t 1 1 \ t 1 — -- - - I ; 1 ! 1 1 ! ' - ' 1 1 N. 1 1 , 1 iS 1 1 1 /-^ > 1 'T V c 1 ' ,r ^ Jr- / ' r^ I / ' 1 1 / \ \ f ' , \ / \ / ■ t i\ 1 \ / s >, '^ ( N J / V 1 J ' . \ / \ J ' ,' • * \ / \ y - I J 1 \ - l—O^ / > ■ ' J ! V A ^ , \ / ' '■ ' 1 ' j I 1 ', 1^ I ' ' 1 1 ^ ] 1 1 A I t 1 1 t 1 -- ^^ - /' 1 ( ' [ t '"I" 1 ; ^ J- -L _L \ 1 _^ . _u Fig. 26. — (.4) Absolute Annual Maximum Temperature. (J5) Average " Temperature. (C) Absolute " Minimum Temperature. (See Tables XXII, XXIII and XXIV.) The Greatest Monthly Eange, The difference between the highest and lowest temperatures recorded during each month of the year is entered in Table XXIV for every year from 1871 to 1903. The extreme difference for each month during the entire period is shown in the table which follows, together witli the ex- MARYLAXD WEATHER SERVICE 115 tremes of temperature and the j^ear of occurrence. The extreme range is also shown in Fig. 27. JFMAMJJASONDJ 70' 1 bU 50° 40° lU 0° Fig. 27. — Greatest Monthly Raui^e of Temperature. (See Table XXIV.) EXTREME MONTHLY RANGE OF TEMPERATURE. 1871-1903. January.. February. Marcb April May. June July August September . October November. . December . . Year. Range. Max. Min. 1873 62° 58 — 4 ]8St) B8° 67 - I 18!tO 65° 77 12 19()3 64° 91 27 189.=) .5.5° 95 40 ISitl 51° 98 47 1W8 47° 104 57 1874 45° 97 52 1873 .53° 93 40 ]87!» 59° 89 30 1879 58° 78 20 1880 59° 56 — 3 Frequency of Days tvith Frost. As the frequency of occurrence of days with a temperature of freezing, and the distribution of such days especially in the autumn and spring seasons, is a matter of greatest practical importance in agricultural and commercial affairs, the subject is here given more than ordinary attention and space. For purposes of convenience all days during which a minimum temperature of 32° was recorded are classed as frost days. It is a well recog- 116 THE CLIMATE OF BALTIMORE nizecl fact of observation, however, that frost usually occurs before the temperature falls to the freezing point (32°) as ofiBcially recorded. The ap- parent inconsistency is of course explained by the method of exposure of TABLE XXIV.— ABSOLUTE MONTHLY RANGE OF TEMPERATURE. 1871. lS'i-2. 1873. 1874. 1875. 1876. 1877. 1878. 1879. 1880. 1881. 1882. 188:^. 1884. 1885. 1886. 1887.. 1888. . 1889. . 1890., 1891. 1892. 1893. 1894. 189.5. 1896 1897 1898 1899 1900 1901. 1902. 1903. 1904. Means. 49 4.5 62 56 54 54 £3 53 61 37 51 53 39 44 55 55 58 40 40 53 39 46 51 .39 51 1871-1880 52.3 1881-1890 48 1891-1900 1 47.61 1871-1903 1 49.01 ■S ' — f^ s 56 46 E3 63 53 53 45 43 46 63 60 36 60 58 47 46 50 51 61 56 38 55 67 57 35 46 66 59 35 43 56 50 63 37 49 41 44 50 57 45 56 48 51 37 47 54 54 50 33 59 43 53 n 44 51.0 53.51 53.8 51.71 46 44 56 45 51 46 45 63 49 51 63 53 63 44 65 .50 55 48 51 55 54 61 49 57 54 43 64 50 53 51.2 45.5 39 41 45 48 46 54 51 43 51 55 49 45 41 44 38 43 36 46 50 44 48 41 44 44 55 49 40 53 43 54 47 48. 9j 49.8! 50.6 51.6 53.1 50.3 47.2 43.6: 47. 0| 45. 7i 37 43 43 39 88 47 40 41 51 44 47 43 43 38 47 41 35 45 41.4 39.4 43.3 41.5 32 29 34 34 34 40 39 33 39 37 31 39 34 35 43 33 35 38 33 43 34 41 38 41 40 35 33 47 37 43 39 37 37 40 34.1 36. 3i 38 38 41 37 45 30 35 31 33 36 30 38 33 33 35 41 34 36 41 33 44 40 35 33 36 39 36 44 53 39 49 43 40 40 45 41 42 40 35 44 40 41 46 45 38 41 39 39 44 49 50 38 41 43 47 44 43 43 47 43 50 47 51 39 43 45 38 59 58 46 47 50 47 34 47 38 54 48 45 44 48 30 53 ;56 45 39 63 39 45 .33 46 36 44 39 46 34 34.6 43.0 36.7 41.2 37.1 46.3 36.0 1 43.6 38 53 38 48 43 53 49 63 49 40 41 53 60 44 49 44 46 48 43 47 44 49 43 47 46 49 40 46 51 46 48 43 41 51 41 44 57 44.fi 45.4 50 44.1 45.8 45.0 47. 9| 46.1 £0 45.5 45.6 48 47 41 43 57 49 37 43 43 50 41 50 50 49 53 47 63 50 53 68 47 54 41 41 ei g a u 68 62 50 65 57 51 53 63 61 57.8 69.0 58.1 58.3 Table XXIV shows the greatest monthly range of temperature (the difference between the highest and lowest temperature recorded within the month) for each month and year from 1871 to 1904, also the average value of these monthly ranges for a period of 32 years, and for each 10-year period. the thermometer. Usually thermometers are placed in a " shelter " which shields the instrument from undue radiation from the ground ; the shelter MARTLAXD WEATHER SERVICE 117 is also usually mounted at a considerable distance above the ground, varying from four or five feet to a hundred feet or more. In a quiet atmos- phere with a clear sky, radiation from the ground is very rapid during the night and early morning hours. The temperature at the surface of the earth may fall considerably below that of the air but a few feet above TABLE XXV.-XUMBER OF DAYS WITH A MINIMUM TEMPERATURE OF »2° OR BELOW. Oct. Nov. I Dee. Jan. Feb. Mar. Apr. Season 1871-3 . 1872-3 . 1873-4 . 1874-5 . 1875-6 . 1876-7 . . 1877-8 . . 1878-9 . , 1879-80 . 1880-1 . , 1881- 2 . 1882-3. 1883-4. 1884-5. 188.5-6. 1886-7.. 1887-8 . . 1888-9 . , 1889-90 . 1890-1 . , 1891-2 . . 1892-3 . 1893-4 . . 1894-5 . . 1895-6 . . 1896-7 . 1897-8 . 1898-9 . 1899-1900 . 1900-1.... 1901-2 . 1902-:} . 190:i-4 . 1872-1881 . 1882-1,><91 . 1892-1 901 . 1872-1903 Means. 0.4 0.1 0.1 0.2 18 26 12 20 15 30 9 30 14 24 9 15 15 11 17 25 19 16 8 21 10 17 20 15 14 20 13 90 19 19 19 23 25 7.0 I 18.8 4.8 i 15.6 5.5 i 16.9 6.1 I 17.5 19 31 14 29 17 29 19 27 9 36 19 35 34 20 24 24 29 16 10 18 33 28 17 26 22 32 15 21 20 23 21.0 20.9 21.7 21.6 17 18 19 24 19 19 14 25 17 30 12 18 11 34 19.3 16.6 19.9 18.8 13.1 13.8 11.2 13.3 2.3 1.3 1.8 1.9 73 94 75 104 77 103 48 85 64 61 80 78 81 87 64 114 81.8 73.1 77.1 78.4 under such conditions. Thus we may have frost, especially in the low places, when the thermometer records a minimum temperature of 35° or 40°, according to the position of the instrument. In view of these facts the figures entered in the following tables to represent the fre- quency of frost days must be regarded as the lower limit of frequency ; the 118 THE CLIMATE OF BALTIMORB figures would be increased to a small extent by placing the thermometer nearer the ground. To enumerate frost days upon the basis of the actual observation of frost on the ground, introduces additional diflficulties as the production of frost depends not only on a temperature of 32° or TABLE XXVL- -LONGEST PERIOD OF CONSECUTIVE DAYS WITH A MINIMUM TEMPERATURE OF 33° OR BELOW. D^'ysf '^*™'' °* Occurrence. 1871-3. 1873-3. 1873-4. 1874-5. 1875-6. 1876-7.. 1877-8.. 1878-9 . 1879-80. 1880-1.. 1881-3. . 1883-3.. 1883-4.. 1884-5. . 1885-6. . 1886-7.. 1887-8.. 1888-9.. 1889-90. 1890-1.. 1891-3. 1893-3. 1893-4 1894-5. 1895-6. 1896-7.... 1897-8 1898-9.... .1899-1900. 1900-1.... 1901-3 . 1903-3. 1903-4. Means. 1873-1881. 1883-1891. 1893-1901. 1873-1904. 15 33 14 30 37 9 10 9 13 24 9 34 15 15 15 34 13 38 Jan. 33-Feb. 5 Dec. 9-Dec. 31 Jan. 30- Feb. 13 Dec. 30- Jan. 38 ( Dec. 13-Dec. 31 (Jan. 30-Feb. 7 Dec. Jan. Dec. Feb. Jan. Dec. Jan. Jan. Jan. Jan. Dec, Jan. Feb. Mar. Feb. Mar. Jan. Dec. Jan. Feb. Jan. Jan. Jan. Dec. Jan. Jan. Feb. Dec. 1.5-Jan. 18 38-Feb. 7 16-Jan. 24 1-Feb. 11 23-Feb. 8 30-Jan. 7 2-Jan. 17 3-Jan. 10 10- Jan. 27 6-Jan. 26 3.5-Jan. 30 9-Feb. 4 19-Feb. 37 l-Mar. 10 37-Mar. 7 11-Mar. 32 3- J an. 36 1-Dec. 9 23-Feb. 25 9-Feb. 23 33-Feb. 6 27- Feb. 10 35-Feb. 17 3.5-Jan. 5 23-Mar. 1 26-Feb. 31 16-Feb. 37 36-Jan. 31 under at the place of formation, but also upon the relative amount of moisture in the atmosphere. The actual observation of frost has, however, been employed in the table in which first and last frosts of the autumn and spring respectively are recorded and a minimum temperature of 32° resorted to only in case of conditions unfavorable to the occurrence of frosts on account of a dry atmosphere. MARYLAND WEATHER SERVICE 11& In making a comparison of the relative severity of winter seasons the frequency of occurrence of a minimum temperature of 32° or under is in some respects a more satisfactory test of the general character of the season than the usual one of the average temperature. In Table XXV, the frequency of occurrence of such days is shown for each month from October to May, and the total number for each year from 1871 to 1904. The season having the greatest number of frost days from 1871 to 1904, is that of 1903-04 when 114 were recorded. There were 104 in 1874-75, and 102 in 1876-77. In 1877-78 there were but 48 ; in 1889- 90, 49. The average number for the entire period of 33 3-ears is 78. LLJ I I I I I I I I — LI — I— Fig. :i8. — Longest Period of Cousecutive Days with a >[inimura Temperature of 32° or Below. (See Table XXVI.) In Table XXVI the longest period of consecutive days with a minimum temperature of 32° is recorded for each year, with the time of occurrence; the duration of these periods is also presented graphically in Fig. 28. The cold periods of this class begin most frequently in the month of January, though many begin in December. In one instance the longest uninterrupted cold spell fell entirely within the month of March, namely in 1890, from March 1-10. The season credited with the longest period of consecutive days witli a minimum of 32° is that of 1878-79 when the minimum was 32° or below daily without interruption from December IG to January 24, or 40 days. In the winters of 1875-7G, 1881-82, 1883-84, 1888-89, 1890-91 and 1893- 94. the longest period was 9 days. The average lengtli of uiiinti'rru])t('d periods of freezing weather is 19 days. 130 THE CLIMATE OF BALTIMORE The cold days occurring in Baltimore from 1871 to 1904 were also tabulated on the basis of a daily mean temperature below 32° and below 14°. The results are shown in Fig. 29 for the entire season. The aver- age winter season contains 33 days with a mean temperature below freez- ing point. The winter season of 1903-04 contained 66, the highest num- ber recorded during any year of the period; the seasons of 1884-85 and 1901-02 follow with 50 each. In 1877-78 there were but 17; in 1879-80 1875-6 1880-1 1 885-6 1890-1 1895-6 1900-1 DAYS A 60 40 |0 B .J L _l L_ _l U _l II _l L J 1 J L -J L. Fig. 29. — (A) Number of Days with Meau Temperature below 32° (j5) a a .1 u .. 140 and 1881-82, 13; and in the winter of 1889-90, but 10, the lowest number on record. The average number for each 10-year period from 1871-1904 is shown in the following table: AVERAGE NUMBER OF DAYS WITH A MEAN TEMPERATURE BELOW 32° 1871-1880. . 1881-1890. . 1891-1900. . Decade. Nov. 0.8 1.3 1.3 Dec. 8.4 6.9 7.6 Jan. 9.8 12.. 5 11.8 Feb. 8.8 7.8 9.1 March. 2.9 4.8 3.7 April. 0.1 0.0 0.0 Season. 30.8 33.3 33.5 1871-1904. . 1.3 8.0 11.9 9.5 3.6 0.0 34.3 Greatest number (1903-4) . . . . Least number (1889-90) 6 15 1 23 2 20 1 2 6 66 10 MARYLAND WEATHER SERVICE 121 The frequency of days during which the highest temperature of the day fell below the freezing point is shown in Fig. 30. The average monthly and annual frequency for the entire period of 33 years and for each decade is given below. AVERAGE FREQUENCY OF DAYS WITH A MAXIMUM TEMPERATURE BELOW 32* Decade. Nov. Dec. Jan. 1871-1880 0.3 3.; 4.7 1881-1890 0.3 2.3 6.7 1891-1900 0.1 3.5 6.3 1871-1904 0.2 3.6 6.0 Greatest number (1892-3) 10 19 Least number 11877-8) 3 Feb. March. Season. 5.9 0.8 12.0 1.6 13.6 0.9 16.7 4.3 1.0 15.1 87>« 880-1 88 D-D 890-1 89>6 900-1 J J Fig. 30. — Annual Frequency of Days with a Maximum Temperature below 32°. With an average annual frequency of 15 days, the number varied from 3 in 1877-78 to 36 in 1892-93. The winter of 1903-0-1 contained 21. In Fig. 31, the annual frequency of cold days is shown on a basis of the occurrence of a daily minimum temperature of 20° in the months of De- cember, January and February and a minimum of 28° in November, March, and April. In the table below the average monthly and seasonal frequency of occurrence of such days is indicated for each ten-year period from 1871 to 1904, and for the entire period of 33 years. 123 THE CLIMATE OF BALTIMORE 187 « 188 OJ 188 5-6 1890-1 1895-6 1900-1 DAYS 50 -J 40 -J 30 1 - 10 - _ Fig. 31. — Annual Frequency of Cold Days. 20° or less in December, January and February. 28° " " November, March and April. (See Tables XXVII and XXVIII.) 187 5-6 18801 1885-6 1890-1 1895-6 1900-1 1 1 1 1 -Dec. 1 1 1 Feb i 1 1 1 1 1 1 1 1 — - Apr 1 1 1 1 l_ —i I— _l 1_ Fig. 32. — Monthly Frequency of Cold Days. (a) 20° or less in December, Jauuary and February. (&) 28° " " November, March and April. (See Tables XXVII and XXVIII.) MARYLAND WEATHER SERVICE 133 FREQUENCY OF DAY^S WITH A MINIMUM TEMPERATURE OF 20° IN WINTER AND 28° IN SPRING AND AUTUMN. Decade. Nov. Dec. Jan. Feb. March. April. Season. 1871-1880 2.8 5.0 0.0 5.2 6.1 0.5 24.8 1881-1890 3.0 4.1 8.2 4.7 6.9 0.1 27.0 1891-1900 2.3 4.2 7.1 6.6 6.6 0.1 26.9 1871-1904 2.8 4.6 7.3 6.2 6.0 0.2 27.1 Greatest number (1903-4) 11 7 14 18 4 1 55 Least number (1881-2) 3 7 1 11 TABLE XXVIT.-LIST OF COLD DAY'S.-January. (Minimum of 20° or below.) DAY OF MONTH. Year. 1 o 2 o 3 o i o 6 6 7 8 9 19 1011 1 12 13 14 15 20 ii 16 12 is 9 16 is 20 26 17 i3 ie 26 18 ii 26 i2 19 26 202 ii.' i8i 20. 26 i 122 23 14 24 _ 17 25 i9 362 15. 7282 « 30 31 g Is71 17 >i 4 18 6—4 11 8 4 3 14 1 4 18. *) ,5 15 18 6 2 2 17 17 19 20 i7 20 6 H fi h'.'. !i9 41 20 18 19 is 20 20 ii iti 19 1 '9 is 2 12 6 io 26 201 ■ 19 7 3 9 4 17 8 8 2 5 .. 6l.^l. .. ..1 ■17. 18 20 14 •49 1 1 1 Table XXVII contains a complete list of all days from 1871 to 1904 during which the temperature fell to 28° or lower in November, March and April, and to 20" in December, January and February; together with the minimum temperature recorded on the corresponding days, and the number of such days in each month. Fig. 31 shows the annual frequency of cold days, and Fig. 32 the monthly frequency. 124 THE CLIMATE OF BALTOIORE Cold days of the class described in the above table occur most fre- quently in the month of January, as is the case with the other classes tabulated. A peculiarity in the seasonal distribution is however revealed TABLE XXVir.-LI8T OK COLD DA YS.-Februaiiy. (Minimum of 20° or below. i DAY OF THE MONTH. 1 3 3 4 5 16 is 16 6 10 26 18 7 i2 8 io 9 ig 15 6 10 is '4 11 12 is 13 14 is .. 20 is 26 is IS 17 15 io .. 26 18 16 '6 26 i2 io • • ii IS ■9 26 ig 8 i3 13 19 is ii ii- ig ii is 5 20 26 io ig ii 8 is is 12 16 21 ig "s 20 12 11 ig 19 22 ii 18 23 i2 13 is i2 16 24 '2 12 14 '3 is 20 '8 16 26 30 " ii "s *8 19 ig 26 ig 26 is is 37 17 ig 'g .. 28 is ii ie 18 i.3 is 2g io 3 H 1871 3 n 16 17 18 20 16 13 ig 4 3 H 4 5 20 17 6 6 1 9 .... .......... [.........[ 20 20 .. -• 3 5 1880 15 4 '' 12 8 19 14 -1 18 7 15 26 19 17 is 12 '3 • • u 13 14 26 4 1 o 3'.'.'.'.'..'.'.'.'.'..'.'.'.'.'.'.'.'. ".'.'.'.'. i 5 15 8 2 1^ 6 18 11 8 ii>'.'. 6 9 18 is i 17 18 '4 8 1890 1 20 18 u 26 is 9 4 3 16 14 9 [on •S 4 19 20 15 5 ig 26 i2 6 4 17 r 16 T) 6 18 14 8 8 14 io 14 8 19 i2 is ii is 19 13 IS 14 16 8 ii 19 ig -6 ig *5 19 6 16 18 'e 14 16 6 1 8 Ii 9 11 1900 11 1 11 1' 3 5 4 12 11 16 17 19 ..18 14 20 IS 14 18 17 18 211 in the comparatively high frequency of occurrence of such days in March, after making due allowance for the fact that the March minimum is 8° higher than the minimum for the winter months. The distribution of cold days of this class by months and years is shown in Fig. 32; a complete list with temperatures recorded is con- tained in Table XXVII. MARYLAND WEATHER SERVICE 125 TABLE XXVIL— LIST OF COLD DAYS.— March. (Minimum of 28° or below.) DAY OF MONTH. Year. o 27 2 o 24 27 8 i 5 6 7 o 23 14 8 9 o 10 11 o 12 o 38 13 o 23 14 35 ,. 3« 15 o 24 16 o 37 27 IT o 25 20 13 23 is 9 19 12 15 20 o ig 24 12 21 o 2i 26 26 27 32 o i9 25 23 24 23 .. 36 23 i9 38 37 18 25 28 2i 22 36 25 26 2i 28 .. 37 27 o 23 30 37 28 28 24 29 o 23 24 38 30 3i 28 •• 31 o o H 1871 o 24 12 o o is 9 10 9 1 8 12 3 2 4 4 1 11 3 4 51013 5 fi 20 20 27 37 26 28 36 i? 21 g 9 l^^iO 24 36.. 28 28 ii 25 20 24 25 12 16 26 ii 3^ 24 18 24 21 25 28 25 25 23 33 23 18 28 ii: 37 38 24 24 27 25 27 28 ih 28 io 26 24 24 25 27 14 23 20 23 ■■ 20 12 26 22 i2 3i 3i 33 36 38 24 i9 ie 23 ■»7 26 is 24 26 1 ^ 27 27 1 ..:28 28 i9 1823 28 27 22 ie 35 2i 37 2.5 20 4 14 24 23 S 13 3 9 13 1 10 9 11 i 5 6 S. ...'.'..'..... '".'.'.'.'.'.'.'. 9 1890 1 •> s'.'.'.'.'.'.'.'.'.'.".'.'.'.'.'.'.'.'..'.'.'. i .5 19 26 20 15 24 16 23 19 24 28 27 24 34 25 20 24 16 21 9i is 32 34 20 i2 27 24 1>3 15 \ 9 27 __ ..'.. 2i; 26 28 22 is 36 21 32 1 4 1900 1 q 28 ..19 13 36 14 <| 3 3 4 '■ '.'ki j6 4 307 TAI5LE XXVII.-LL>\ 11 94 and February, and to a minimum of 28° in ]March and November. The designated minimum must be reached not later than 12 hours after the expiration of the 24-hour period. Thus three events are essential for the technical verification of a cold wave, namely: (a) a fall of 20°; (b) the fall must occur within a period of 24 hours; (c) a designated minimum temperature must be reached witliin 36 hours. Cold waves MARYLAND WEATHER SERVICE 127 fulfilling all these conditions are not of frequent occurrence within the geographical horizon of Baltimore. It has been shown in the para- graphs dealing with diurnal variability of temperature to what extent the frequency of occurrence of given changes in temperature from day TARLE XXVII.-LIST OF COLD DAYS.— December. (Minimum of 20"^ or below.) DAY OF THE MONTH. Year. 1 o 2 o 3 4 o 5 12 6 7 8 ' 10 11 12 « 14 15 16 17 18 19 « 21 22 23 24 25 26 27 2829 30 31 ■5 1 11*71 o 17 is in i is ii 20 18 is o o o o ii o 13 ie 8 o i2 15 ie 17 o ie 10 io 7 i2 12 9 20 5 14 ie 20 20 14 o 16 6 11 o i7 is is is 20 's 20 ii is 14 16 's 18 is ■■ • ■ ■■ i4 18 16 ig o ii •• ie IS 20 20 ii 20 20 is 12 o ii is io 20 ii is 20 11 o is 18 20 ii is ii 15 is is 19 ie • • 4 ie is is '7 20 ie i9 ,', -3 20 17 ig is 16 ii 17 -i 26 26 17 ii 9 5 10 1 5 le 6 9 9 .... 17 ie 18 20 .. 3 4 5 fi T 8 9 1880 1 1=) ii'C. 3 3 4 3 6 6 l*)!"! IS .. 19 15 16 <> 6 17 16 18 18 20 11 8 18 4 9 1890 20 5 1 1 o 10 3 18 •• i4 is 20 is i4 i9 is is 19 19 17 ig 19 18 20 20 17 1R 4 3 6 19 7 8 2 1 9 1900 1 20 16 18 6 4 11 §!!!!!!!. .............. 'w" •■ •■ i 154 to day depends upon the method of determining the change. Basing the 20° fall upon the minimum temperatures, the 8 a. m. or 8 p. m. temperatures from day to day, the Baltimore records from 1871 to 190-1 show a frequency of cold waves indicated in tlie table below. The table shows the extent of tlie fall, and the minimum temperature attained within ."50 hours. 128 THE CLIMATE OF BALTIMORE TABLE XXAail.-FREQUENCY OF COLD WAVES. 1870-1 . 1871-2. 1872-3.. 1873-4.. 1874-5. . 1875-6.. 1876-7.. 1877-8.. 1878-9.. 1879-80. 1880-1. 1881-2. 1882-3. 1883-4.. 1884-5. . 1885-6.. 1886-7 . 1887-8.. 1888-9.. 1889-90. 1890-1. 1891-2. 1892-3. 1893-4. 1894-5. 1898-9. 1899-1900. 1900-1. 1901-2. 1902-3 Nov. Fall I Min. 36 26 35 22 1895-6 21 1896-7 1897-8 1903-4 21 Total number 8 Average number. . . 0.2 Dec. Fall Min 19 0.6 13 7 20 8 19 15 9 13 18 12 10 IS 15 16 Jan. Fall Min 20 23 0.7 Feb. March. Fall Min. Fall Min. Q ^ ^ 21 21 15 20 26 15 21 9 19 27 24 20 16 21 25 7 26 20 25 17 16 33 23 27 25 23 28 33 28 16 12 5 10 20 23 26 22 28 18 25 23 24 "o 11 3 ~0 18 22 18 21 20 25 27 23 22 12 20 13 8 23 18 34 20 5 1^ 25 19 24 2 25 10 28 13 26 13 22 27 30 20 20 18 8 19 25 0.7 17 0.5 Season. No. Gr. fall 92 2.7 21 20 15 20 18 12 5 14 18 13 13 16 17 13 18 Table XXVIII shows the frequency of occurrence of a fall of 20° or more in the temperature of two successive days, combined with the attainment of a minimum temperature of 20° or less in December, January and February, or a minimum of 28° or less in March and November. MARYLAND WEATHER SERVICE 129 As a special chapter is to be devoted to an analj'sis of cold waves in Part II of this Eeport, reference is here made only to the frequency of their occurrence. In the period comprising 34 winters the total num- ber of occurrences fully satisfying the technical conditions imposed is 92, or an approximate average of three per year. Since 1871 there were three seasons without a cold wave, namely, 1873-74, 1885-86, and 1889-90. The greatest number occurring in any one season was 6, in 1871-72, 1884-85, and in 1903-4. Of the total of 92 in 34 years, 8 oc- curred in JSTovember, 19 in December, 23 in Januar}^, 25 in February, and 17 in March. The greatest fall in temperature recorded within the prescribed time of 24 hours was 38°, which occurred in December, 1901. Three cold waves occurred in each of the following months: December, 1871-72, January, 1898-99, February, 1903-04, and March, 1882-83. Killing Frosts. A factor of the highest importance, especially to the agricultural and trucking interests of a community, is the average date of occurrence of the first " killing " or " black " frost in autumn, and the last in spring, and their variations in time of occurrence from year to year. Frosts are usually designated as " light," " heavy," or " killing." The term " light " is applied to frosts which are destructive only to tender plants ; " heavy " to copious deposits of frost, but which do not destroy the staple products ; " killing " to such as are blighting to the staple products of the locality in which the frost occurs. First and last killing frosts are tabulated below for each year from 1871 to 1904 for the vicinity of Baltimore. In the absence of a killing frost before a minimum tempera- ture of 32° was observed, the date of the first record of a freezing tem- perature was entered in the table. The interval in days between the last frost in spring and the first in autumn is likewise given in order to show the length of the period of safe plant growth. The average date of occurrence of the last killing frost in spring, based on observations of 34 years, is April 4. It has occurred as early as February 26, namely in 1903, and as late as May 3, as in 1882. The first killing frost in autumn has occurred, on the average on November 3. 130 THE CLIMATE OF BALTIMORE The earliest appearance is that of October 6, 1892, and the latest that of December 6, 1878. In the ordinary course of events, accordingly, the period of safe plant growth in the neighborhood of Baltimore, based upon the occurrence of killing frosts, is from April 4 to November 3, TABLE XXIX.-KILLING FROSTS. 18T1 lS-2.... 1873 1374 1875 1876. 1877. 1878. 1879. 1380. 1881. 1883. 138-3. 1884. 1885. 1886.. 1887.. 1888.. 1S89.. 1890.. 1891 . 1892. 1893. 1894. 1895. 1896 1897 *Mar. 1S98 1899 1900 1901. 1902. 1903. 1904. Average date 1871-1903. Earliest date Latest date Last in Spr mg-. First in Autumn. Interval in days. Min. Min. *Feb. 23 30° *Nov. 28 31° *Mar. 25 33 * •> 16 30 " 31 29 Oct. 29 31 OJ.7 Apr. 13 29 Nov. 10 31 211 " 22 3.' '* 3 33 195 .* » 30 Oct. 15 33 196 '• 3 33 Nov. 4 37 215 Mar. 2fi 21 Dec. 6 32 255 Apr. n 32 Oct. 26 30 204 ■' 12 30 Nov. 8 35 210 .. 21 39 " 27 34 220 May 3 38 " 19 30 200 Apr. 25 34 " 13 32 203 Mar. 30 31 " 7 30 •;•» " IB 31 '^ 1 33 2:30 .. 24 29 Oct. 17 36 307 Apr. 6 30 *• 31 32 308 Mar. 19 30 " 22 33 217 * n 30 28 Nov. 6 35 *Apr. 2 31 Oct. 31 36 » 9 36 " ;:9 33 303 " 15 34 " 6 36 1:4 " 16 36 " 17 36 184 " 11 32 Nov. 12 ''V 31. ** 11 34 Oct. 29 34 301 " 8 33 Nov. 14 32 320 *Mar. 29 34 Oct. 31 39 Apr. 6 26 " 28 34 205 Mar. 25 ;!0 Nov. 4 38 224 ** 22 26 " 16 38 239 * — 8 + 5 — 3 — 5 — 4 + 2 —18 -i-11 + 5 + 7 -14 -16 -11 -11 + 2 — 6 -14 + 5 — 3 — 9 -i- 1 ^2'.» -11 The average departure from the normal period of 226 days is about 12 days. In the year 18T1 the last freezing temperature occurred 53 134 THE CLIMATE OF BALTIMORE days before the average date; the nearest approach to this excessive deviation was a departure of 31 days in the opposite direction in 1875. The figures in the above table indicate that a departure exceeding one month is extremely improbable, having occurred but twice in 3-1 years. In 28 out of a total of 34 years the limit of variability was under 30 days; in 25 it was under 15 days; in 15 under 10 days; and in 10 under 5 days. I , • 1 ^^^^^ - ^ ; ^^* ! 1 1 , ^ 1 i ■ ' 2*" ' 1 1 ^^™ ' ' : ! ** ^ , : 1 , . , . 1 _ H ■B^MM ^^-^ -t- - - - —- ^M "'~~^~^"r \~ - 1 j - J_ ' ! 1 j [ 1 1 ■ ^^ 1886 ._ _ _ -\ 4S ^■^^M I ^ _ — -^ -- ^ \ — - - -- 1 ■ ^^1 1 ^; 1 1 ^1 nss * ! ' '. i 1 i! ■: : 1 MJJ ■¥W Til ^^ ^Z 1 iffi L. j_ 1 T -L 1^1 ^^^■1 ^^ ■^■21 "□"•" , ' . , ■ H^^E ■ , , 'II 1 , 1 i 1 1 1 M 1 1 j 1 I.I 1 . . i. ^. ±1± MEAN 1 1 ! ! 1 ■r*T L . il McH. Apb- May June J"' v Aug. Sept. Oct. Nov. Dec. Fig. 34. — Interval Between Last and First Occurrence of ^Minimum Temperature of 40° (Light Frosti. Fig-. 34 shows the time of last occurrence of a minimum temperature of 40° in spring, and that of the first occuri-ence in fall, together with the length of the intervening period in months and daj's. The lowest line marked " mean" shows the date of average occurrence and tbe average length of the intervening period. (See Table XXXI. i The probability of the occurrence of an injurious freeze some time within the month of April may be expressed by 65 on the basis of a possible 100, a temperature of 32° or less having occurred at some time within the month 22 times in 34 years. The probability of occurrence from the 1st to the 10th of the month is 41 ; from the 11th to the 20th, 21; from the 21st to the 31st, 3. It has occurred 15 times in 34 3'ears on some day between the 1st and the l(^th of April; 7 times from the 11th to the 20th; and 1 time after the 20th. These figures represent the chances of an injurious freeze in the first, second and third decades, respectivel}', of the month of April. MAKYLAXD WEATHER SERVICE 135 Light Frosts. A light frost, injurious onh' to tender plants, may, and frequently does, occur with a recorded minimum temperature of 40°. Hence the period of safe growth for the frailer varieties of vegetation is reduced TABLE XXXI.— LAST AND FIRST OCCURRENCE OF A MINIMUM TEMPERATURE OF 40°. Last in Spring^. Year. Mar. Apr. May ISTl 1ST3 1873 1874 1875 1876 1877. 1878 1879 1880 138L 1S82 1883 1884 1885. 1886 1887 1888. 1889 1890 1391 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1903 1903 1904 Average date, 1871-1880 " I8fl-1890 " 1891-1900 " 1871-1903 13 13 First in FalL _ o6 Sept. Oct. Nov. Sa i 5"^ t 31 30(1 13 179 is 145 15 168 13 140 8 157 4 203 39 244 36 161 35 177 15 308 184 itJ lf,9 33 194 23 l-ifi ■.'10 3 160 9 185 •2S 191 18 3 165 160 17 173 15 185 9 149 9 183 18 180 17 171 1 167 17 160 18 189 29 196 19 170 17 178 23 186 12 169 18 179 a": +37 -»4 —11 -39 +34 +6.-. —18 +39 + 5 —10 +15 +31 — 3 —19 + 6 + 13 -14 -19 — ti + 6 -30 + 3 + 1 — s —13 -13 +10 +17 — 1 + 7 -10 still more. A minimum of 40° has been recorded at Baltimore as late as May 26, namely in 1875; but this is an exceptional case. The last spring mini iiui 111 of 40° has occurred as early as March 29. The dates of the last in spring and first in autumn, witli the length of the 136 THE CLIMATE OF BALTIMORE intervening interval in days, is shown in Table XXXI. This interval is also shown graphically in Fig. 34. The Period of Effective Temperatures for Plant Growth. It is generally conceded that every plant requires a certain tempera- ture in order to develop successively the leaf, bud and fruit ; that there is a minimum temperature below which physiological activity in the plant ceases, and hence that only temperatures above this limit are effective in carrying the plant forward from the sprouting of the seed to the maturity of the fruit. This "critical" point in the history of plant growth is assumed to be a mean daily temperature of 43° F. In order 1875 1880 1885 1890 189b 1900 DAYS - L - 200 - - - - - _ DAYS 300 Fig. 35. — Annual Number of Days with Mean Temperature above ■42°. Fig. 35 shows the total annual number of dajs having a mean temperature of 43° or above, the degi-ee of heat which marl£S the beginning- of physiological activit)- in plants. to determine the number of days in the year during which this " effec- tive " temperature prevails in the vicinity of Baltimore, the days with a mean daily temperature of 43° or above from 1871 to 1903 have been tabulated. The normal number of such days for each month and for the year is shown in the following table, while the variation in the total annual number is represented in Fig. 35. PERIOD OF effective TEMPERATURES FOR PLANT GROWTH. (Average number of days.) Means. Jan. Feb. Mar. Apr. May .Tune July Aug. Sept. Oct. Nov. Dec. Year 1871 1902 5.3 0.5 14.1 27.0 31. n 30.0 ; 31.0 31.0 30.0 30.0 19.4 8.2 264 MARYLAXD WEATHER SERVICE 137 The first appearance of a daily mean temperature of 43° in spring normally falls upon the 25th day of March, and the last upon the 27th day of November, an interval of 245 days, or about eight months. How- ever, such days occur throughout the year and are probably effective in directly or indirectly promoting physiological activity in the plant. Hence to the period above mentioned must be added the days of the winter and early spring months before the permanent appearance of the daily mean of 43°. This will materially increase the annual period of " effective " temperatures, as a considerable proportion of the winter days fall within the prescribed limit. Calculating on this basis the average period comprises 264 days. The longest period, namely, that 1675 1880 1885 1890 1895 1900 DAYS 10 Fig. 36. — Annual Number of Days with Maximum Temperature of 90° aud over. (See Tables XXXII and XXXIII.) of the year 1878, contained 293 days, and the shortest, 244 days in 1886. The ten year average values of the three decades from 1871-1900 varied only from 261 days to 268 days, showing a remarkably constant average length for this period. The Frequency of Warm Days ix Summer. It is a well recognized fact of observation that the temperatures above the normal heat of a locality fluctuate to a less degree than the tempera- tures below the normal. In other words the extent of variability in the temperature for a given locality is generally determined by the cold days and not by the warm. The extreme maximum temperature in the United States has a range of about 40°, or from 80° to 120° ; the 10 138 THE CLIMATE OF BALTIMORE extreme miniiiuim varies from (15° below zero to about 40° above, a range of 105°. While variability in the temperature of warm da3-s is of less importance in agricultural and mercantile life than that of cold days, it is a faetoi- of much concern in personal comf(jrt, especially to the dwellers in large cities. TABLE XXXII.-LIST OF WARM DA YS.-Aprii.. (Temperature of 90° or above.) Year. ' 2 3 4 o 5 o 6 o 7 o 8 o 9 o 10 11 o 12 o 13 o 14 o 15 1617 181920 o o o 9493;; 21 22 2334253637 00000 3829 ..90 30 9i 1 1888 1898 o o o 1 1903 1 4 May. 1 3 4 5 6 7 8 9 10 11 13 13 14 15 16 17 18 19 20 •21 22 33 90 34 35 90 •■ 26 ■■ 92 37 93 38 90 29 30 95 31 94 95 1877 92 90 92 93 9i •• 92 1879 18S0 •• 91 9i 94 94 !? •• 90 93 1 7 18S1 1 3 1889 90 93 93 96 1895 1«!)6 __ :: 3 5 189S 1 1899 1900 .. 9i 9i 94 1 3 19U-,' 1 1903 . 92 1 39 Table XXXII contains a complete list of all days from 1871 to 1903, during which the temperature rose to 90° or above; together with the maximum temperature recorded on such days and the monthly frequency of occurrence. Fig. 36 shows the annual frequency of days upon which the maximum tem- perature equalled or exceeded 90°. The frequency of occurrence of days with an excessively high tempera- ture hence plays an important part in the composition of local climates. We can provide against extreme cold in Avinter; from the hot and ener- vating summer weather there is no escape for the great numbers who are compelled by circumstances to remain in the large cities beyond the reach of mountains or seashore. A temperature below 90° is not especially uncomfortable or unwhole- some unless accompanied Ijy a high degree of humidity and a stagnant MAEYLAND WEATHER SERVICE 139 atmosphere. Defining a hot day as one with a temperature of 90° or above, the Baltimore statistics show a frequency and distribution indi- cated in Table XXXII. TABLE XXXII.-LIST OF WARM DATS.-June. (Temperature of 90° or above.) Year. 1 3 4 5 o 6 o 7 o 8 o 9 o 90 10 o 11 13 90 13 o 14 o 15 16 o 17 o 18 o 19 93 20 o 21 9i 22 23 95 96 24 93 96 90 95 9i 93 98 (V) 94 25 95 90 92 93 97 92 93 92 9S 36 .° 9i 97 94 95 90 92 93 91 95 92 27 94 95 92 92 91 90 9;i 92 38 91 92 93 93 93 92 94 92 94 9i 93 29 97 90 93 93 91 94 .. 95 90 99 1871 1 3 ^ 4 90 96 98 90 9 5 7 6 5 90 90 95 92 90 93 91 9i 9i 90 yi 90 92 .. 92 91 90 90 93 93 93 93 97 91 4 H . 9 94 1 5 1880 9 1 3 o 90 90 n ;j 90 1 4 4 5 91 95 90 94 9S 94 4 6 n 92 94 90 s 9 1890 91 90 93 94 93 94 90 91 9i 90 9i 9:i 4 1 90 n R 3 4 90 92 90 90 5 5 9t 95 97 5 6 H 7 8 90 93 9i 92 91 94 90 93 •• 93 2 7 9 93 98 96 95 91 8 1900 n 1 7 93 4 3 147 Such days do not generally make their appearance until the month of June and disappear in the first week of September. The average num- ber for the entire season is 21, and the monthly distribution is as follows: June 4, July 10, August 5, and September 2. They occasion- ally occur in May (about one in two years on the average), and have occurred on two occasions in 33 years as early as April and once as late as October. There is considerable difference in their total frequency during 140 THE CLIMATE OF BALTIMOEE a period of 10 years: From 1871 to 1880 there were 204; from 1881 to 1890, 161 ; from 1891 to 1900, 243. The annual frequency has varied from 8 in 1871 to 43 in the memorable summer of 1900, as shown in Fig, 36. It is remarkable that the summer containing the highest total TABLE XXXII.— LIST OF WARM DAYS.— July. (Temperature of 90° or above.) Year. 1 2 3 4 5 93 91 92 92 92 6 o 94 90 92 7 o 90 90 8 o 92 97 93 9 o 91 92 99 91 91 93 90 10 o 91 93 96 97 94 93 97 93 11 o 90 96 oi 95 94 'M 13 ° 96 90 13 o 91 90 95 'M Itfi 9i 96 9i 94 97 93 93 14 o 92 94 92 91 91 90 it;.' 95 90 90 91 9i 15 90 9i 94 91 92 94 Itl 9i 90 9i 96 92 16 o 92 91 90 9i 90 91 99 94 91 93 loi 92 92 92 100 90 17 9i 93 90 9i 91 90 95 99 97 90 90 100 9fi 18 o 95 96 96 93 98 96 102 92 93 98 90 99 19 o 93 96 93 94 93 20 9i 97 93 98 90 97 93 90 9i 90 95 21 95 99 95 92 94 96 92 33 o 90 96 93 ■■ 9i 96 93 33 90 93 95 93 92 90 • • 94 94 o 92 90 91 95 92 90 93 90 25 o 94 92 90 90 96 97 92 91 9i 94 9i 26 94 93 94 93 91 90 93 94 99 98 93 ■• 94 37 93 90 9i 97 92 96 92 38 o 92 93 95 97 92 90 29 90 93 97 90 95 94 94 96 92 30 o 9i 92 91 92 92 9i 92 94 95 95 31 90 95 90 • • 90 o 1871 96 o 97 93 96 96 93 93 92 s 1" 3 15 4 7 1 6 94 93 95 92 98 IS 93 ^ H 16 9 95 10 1880 10 1 91 93 96 96 91 11 8 3 91 92 94 94 93 7 4 s 5 91 93 94 93 90 93 98 94 93 90 93 90 90 92 90 If) g i 7 8 90 9i 10 5 9 ») 1890 91 8 1 10 3 91 90 93 95 90 96 90 94 95 95 92 9i 93 q 4 91 11 5 5 6 90 104 90 92 97 97 95 l66 90 97 96 94 94 95 90 96 96 95 96 90 9i 93 92 90 10 91 97 <) 8 100 10 9 8 1900 15 1 103 103 19 3 10 3 93 95 1'^ 307 number of excessively hot days in a period of 33 years did not have an average temperature sufficiently high to class the season as excessively warm. The " hot-spell " occurred late in July and in August and was followed by an exceptionally warm autumn. This period will receive further attention in Part II, in the discussion of summer weather. Inspection of Fig. 36 reveals a strikingly uniform increase in the num- MARYLAND WEATHER SERVICE 141 ber of days with a maximum temperature of 90° and above since the year 1889. Starting with a frequency of 9 in the latter year, the number rose steadily to 43 in 1900 with but one marked interruption and two or three minor ones. Since 1900 there has been a steady fall, repre- TABLE XXXII.-LIST OF WARM DAYS.-AuGUST. (Temperature of 90° or above.) Year. 1 o 2 o 3 o 4 o 5 o e o 7 8 9 o 10 o 11 12 13 14 o 15 o 90 16 o 9J 17 18 90 9i 19 93 92 92 92 20 o 93 97 92 94 97 21 o gi 96 " •• 91 90 97 90 22 o 96 90 90 23 o 90 90 93 24 94 94 94 93 25 90 92 90 97 26 92 90 96 27 92 90 92 28 90 90 9i 29 o 94 9i 92 30 o 9i 31 91 96 93 "5 o 1871 o 92 92 94 95 96 90 1*? 3 4 94 91 90 4 3 5 6 90 92 92 9i 93 98 90 g 8 91 90 90 94 90 94 90 93 90 (f 9 qoqo r, 1880 1 92 2 8 1 3 p 4 3 5 90 92 90 oi 90 90 96 90 9i 90 90 9i 92 92 9i 3 6 90 91 94 92 9i 91 90 9i 97 4 90 90 93 94 90 92 95 94 95 s 8 90 94 93 10 9 1890 95 1 9 1 5 3 3 9 4 93 91 94 90 94 97 96 90 95 96 o 6 10 6 91 96 94 98 10 1 8 93 91 93 97 93 100 92 99 l66 100 93 i66 90 99 gi 92 9 9 1900 90 90 8 17 1 8 •J 3 90 91 90 •• 4 2 m sented by the figures 43, 30, 20, 16. The numbers for the rising branch of the curve of frequency are: 9, 14, 11, 19, 16, 23, 28, 29, 12, 35, 27, 43. The real discomfort of a hot summer depends not so much on the actual number of hot days as the length of the periods of uninterrupted hot weather. A month, for example, with 10 scattered days having a temperature exceeding 90° would be far more comfortable than one with an equal number of consecutive days with the same degree of heat. In 14S THE CLIMATE OF BALTIMORE fact, liot spells of tlie latter description are of comparatively rare occur- rence in Baltimore, not having occurred more than five times in 33 years, TABLE XXXIL-LIST OF WARM DAYS.-September. (Temperature of 90° or above.) Year, 1 2 3 4 5 f) 7 8 o 9 o 10 11 o 12 13 o 14 o 15 16 17 18 o 90 9i 19 o ■■ 94 20 o 90 21 o 96 o 96 23 o 95 o 25 o 90 26 o 90 93 27 o 90 28 o 91 29 30 o -2 1871 o o o o o o 2 94 90 93 92 92 98 90 ' 1 o 3 P3 90 o 4 o 5 92 90 9i 9i 90 93 90 101 90 >) 6 7 8 9 1880 1^ 1 R 2 3 i 5 6 90 •• t 7 8 9 1890 1 1 2 3 4 94 94 90 97 95 93 93 92 [ ' ■■ o 5 7 6 91 93 94 94 93 9i 90 9i 91 92 3 7 4 8 95 96 97 91 90 8 9 o 1900 90 4 1 1 9 92 " 1 3 n m October. 1897. 1011 192021222324252627 293031 while the former condition has occurred about 25 times. A list of the longest periods of consecutive days with a maximum temperature of 90° or above for each year from 1871 to 1903 is published in Table XXXIII. The table likewise shows the dates of beginning and ending of the periods MARYLAND WEATHER SERVICE 143 and the maximum temperatures attained. Their annual average length is a little less than six days, with limits of variations hetween 2, as in 1871, 1886 and 1889, and 14 in 1900. The season with the longest hot spell on record likewise contained some of tlie highest temperatures ever TABLE XXXIII.-LONGEST PERIOD OF CONSECUTIVE DAYS WITH A MAXIMUM TEMPERATURE OP 90° OR ABOVE. 1871 1872 187.3 1874 1875 1876 1877 1878 1879 1880 1881 1883 188;^ 1884 1885 1886 . .. 1887 1888 1889 1890 1891 1893 1893 1894 1895 1896 1897 1898 1899 1900 1901 19f)2 1903 Average Began. July 9 June 30 July 14 June 7 June 23 July 8 July 25 July 4 Aug. 2 July 9 July 3 July 34 July 2 June 19 July 16 July 7 July 11 Aug. 3 July 8 July 30 Aug. 9 July 34 June 19 July 2.i May 30 Aug. 4 Sept. 9 Aug. 30 .lune 5 Aug. 6 June 26 Aug. 2 July 8 Enrled. July Aug. 1 12 31 23 29 •lune 3 13 11 Sept. 1 8 19 July 7 4 11 Length. Days. Max. temp. 8 4 11 * 2 4 * 2 * 3 10 3 9 * 4 14 12 3 4 5.8 91° 93 94 92 99 96 93 94 93 99 92 96 94 93 95 94 99 98 97 97 97 97 98 100 103 91 96 ♦Two periods of equal length ; the period with the highest maximum temperature selected. recorded in Baltimore. On six successive days the maximum tempera- ture ranged between 99° and 100°; the month contained 17 hot days, and was preceded by a iiiiuitli with I't. This was doubtless the most try- ing period in tlic history of Pialtimore summers. A more extended ac- count of tliis roniarkal)l(' hot s])ell will l)e given in V^vi IF of this Report. 144 THE CLIMATE OF BALTIMORE -EEE:::::!;i|^ 1 5 i I il: ■; 1 ' 1 - ^ Fig. 37. — Time of Occurrence of the Lowest aud Highest Temperature of the Year. Fig. 37 shows the time of occurronce of the lowest temperature of each winter season, trom 18a to 1904, and of the highest temperature of each succeeding' summer season, from 18/1 to 1903; also the length of the intervening period in months and days. The lowest line marked ' mean " shows the average time of occurrence and the average length of the inter- vening period. The line for 1904 shows only the time of occurrence of the lowest tempera- ture. (See Table XXXIV.) ^ / // ,- \ // // ■^^ \ / 1 1 // 1 1 11 1 \ 1 I // 1 \ / II 1 1 '/ \ / // / b /' c \ /: s V ■"^ "*> Fig. 38.— (a) Air Temperature at 2 p. m. (Harbor). (6) Temperature of Surface Water, 2 p. m. (Harbor), (c) << " Water at depth of 10 ft. (Harbor). Fig. 38 shows the mean monthly temperature of the water in the harbor, and of the air at 2 p. m., at the foot of youth street, (a) Air temperature ; (70 temperature of the water at the surface ; (c) temperature of the water at the bottom (10 ft). See Table XXXV. MARYLAND WEATHER SERVICE 145 Time of Occurrence of Annual Minimum and Maximum Temperatures. The dates of occurrence of the winter minimum and the summer maxi- mum temperatures, and the length of the intervening period, are ex- table xxxiv.-time of occurrence of annual minimum and maximum temperatures. Min. Date of minimum. Date of maximum. Max. Summer. Dec. Jan. Feb. Mar. June July Aug. Sept. 1870-1 18T1-L' \^-,o 3 10° 5 —4 13 -2 13 I 13 —6 7 11 8 3 —1 7 9 3 13 16 13 1 8 1 5 8 10 8 13 11 5 si 18 10 37 23 5 8.04 30 16* 10 's 3 1 24 23 6 ■3 23 i7 16 26 6 15 4.08 6 ii 5. 24 25 6 17 10 1 io 11 4.01 "7 3 6 3 13.07 '9 27 26 25 26 24 3 "e 8 B7.«2 16 18* 9 is 16 13 23 24 21 7 18 '9 8 26 '3 It 18 19 97.°9 ie 11 7 25 4 96.03 7 ii 3 99.00 930 97 96 98 97 99 95 98 99 99 101 97 96 95 99 93 102 96 93 98 94 99 98 98 97 98 97 104 98 100 103 99 97 1871 1872 1873 l}Jt73 4 1874 l»T4-5 1875-6 1875 1876 1S76-7 1877 1877-8 1878 1878-9 1879 1879-80 . . 1880-1 lt<8l •' 1880 1881 1883 1} 3 1883 1883-4 1884-5 1884 1885 1885-6 1886 18iS6-7 1887 1887-8 1888-9 1889-90 1885 1889 1890 Iy90-1 1891 1891 3 1893 189"^ 3 1893 1893-4 1894-5 1894 1895 1895-6 1896 1896-7 1897 1897-8 1898-9 1899-1900 1900-1 1901-" 1898 1899 1900 1901 1903 iyo-_'-3 1903 1903-4 1904 No. of occurrences •"•• * On other days also. ceedingly variable quantities. While the lowest winter temperature usually occurs in January, it has on several occasions appeared in December, frequently in February, and occasionally in March, in the past 34 years. The earliest occurrence was on the 10th of December in 187G, the latest on the 7th of March in 1890. 146 THE CLIMATE OF BALTIMORE TABLE XXXV.-MEAN TEMPERATURE OF AIR AND OF SURFACE WATER IN THE HARBOR AT 3 P. M. Date. n Air Water 10 II W 15;- 16^ 18- 20 28 Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec Monthly Average Air ... «« Water. Si 35 36.3 SA.3 35.3 Si.O 40.5 i^ii.e 35.2 Si.6 37.0 54.6 89.3 Si.l 45.0 Si.O 43.7 Si.O 43.1 Si.i 41.0 Si.S 35.7 SJ,.5 43.7 Si.S 46.7 35./, 45.6 36.1 45.6 56.6 43.3 36.3 44.9 56. i 41.9 36.5 41.0 56. :? 35.5 5,5.7 38.3 56.0 38.0 36.9 38.3 ,36.6 36.8 56.6 44.3 56.7 40.0 .1)6.6 43.3 36.3 43.3 57.5 4 40. .8 35. 42.1 37.1 46.3 57.7 43.6 .37.5 43.7 56. S 42.4 57.2 43.3 37.9 41.7 5S.0 35.7 5«.3 39.3 58. i 46.6 5«.5 43.7 55.7 48.1 3S.S 45.4 39.5 47.0 59.9 .50.9 40.5 46.8 40. i 45.3 40.2 46.9 40.6 49.9 40.9 41.9 40.9 45.3 40.5 41.2 40.6 39.3 /.1.2 47.3 4i.7 49.2 42.5 53.6 42.6 53.5 45.0 53.8 45.9 46.7 45. i 48.6 42.S 53.3 45.1 54.1 iS.O 58.4 45.7 62.9 46.0 51.0 46.0 55.1 46.6 55.3 47.0 66.8 47.6 55.5 47.9 61.8 47.6 53.5 iS.l 48.3 iS.7 .54.9 45.5 58.1 49.5 57.3 50.4 54.3 50.6 57.5 5i.4 59.3 5i."4 63.9 52.0 6.). 2 55.2 64.6 54. i 63.6 5/t.5 63.8 55.3 61.5 55.5 65.0 55.9 59.9 56.5 63.1 56. i 65.8 57.0 67.5 57.5 61.4 57.4 63.5 57.6 45.8 68.7 40. -Z 51. i 63.8 57.6 68.3 58.3 65.8 59.0 69.7 .59.5 69.0 59.2 65.9 59.2 61.6 59.4 63.9 59.6 73.3 59.6 69.0 60.8 64.6 60.6 63.0 6i.O 69.0 60.5 60.9 60.5 68.1 60.8 67.1 6?.0 66.2 6i.O 69.0 61.6 69.1 62.5 71.5 62.8 73.3 64.^ 73.1 64.4 73.3 66.1 71.9 65.6 70.0 65.0 71.6 65.2 72.9 65.5 73.8 65.7 69.7 66.2 70.4 66.8 73.2 67.0 62.7 78.1 67.4 74.4 67.7 76.0 67.9 76.8 68.3 79.9 69.2 77.9 70.1 79.9 70.6 79.2 7i.2 77.7 7i.O 81.8 71.8 77.6 's.k 73.6 79.9 75.7 77.0 79!3 75.9 80.5 74.5 79.3 75.3 80.6 75. i 80.0 76.4 81.5 76.2 80.3 76. i 79.8 75.7 83.1 76.0 83.0 76.1 77.9 76.5 77.5 76.3 79.3 76.7 80.7 76.9 76.7 79.1 75.4 78.3 75.5 78.1 75.9 82.8 76.5 80.9 76.5 81.3 77.0 83.6 77.5 85.3 77.8 84.7 77.5 81.5 77.5 85.6 77.6 83.3 77.7 82.6 77.9 80.3 77.7 79.4 77.9 81.1 75.0 83.2 75.2 83.1 75.2 84.0 79.4 84.6 79. i 83.0 79.1 83.5 79.0 87.6 79.4 87.9 79.5 86.7 79.5 84.0 85!o 79.5 83.5 79.5 83.1 79.7 82.9 79.9 83.3 79.5 82.6 79.7 81.4 79.7 80.3 79.5 79.0 78.9 80.6 75.6 80.5 79. i 78.6 75.7 76.4 75.0 79.3 75.2 78.8 75.2 79.0 75. i 81.6 75.5 81.3 78.1 84.7 75.5 83.5 78.8 75.2 78.4 75.0 83.5 75.2 81.6 75.1 83.8 75.4 83.4 75.5 84.5 75.4 81.3 75.5 80.8 75.4 83.0 75.2 83.0 75.2 80.1 75.0 77.5 77.6 77.5 77.5 76.0 76!9 76!9 77.1 80.3 75.5 77.1 76.8 75.3 76.7 78.3 76.6 79.0 76.6 78.8 76.5 77.2 76.7 77.7 76.4 78.5 76.5 77.7 76.4 77.3 76. i 73.3 75.7 73.3 75.2 73.3 74. i 76.7 7i.l 76.4 74. i 78.0 74.0 79.1 74.5 73.8 74.4 76.4 74.0 76.8 75.0 73.7 72.6 68.8 72.0 70.6 7i.6 73.0 71.1 73.4 70.9 69.6 70.6 73.0 70.5 77.8 71.0 73.5 70.5 73.7 70.0 75.5 74. ( 70.7 69.5 66.5 69.0 66.4 65.7 68.1 65.6 68.6 65.5 67.1 67.5 67.7 67.6 66.8 67.-? 67.9 67.5 69.1 67.5 71.5 67.2 70.9 66.5 72.5 66.5 69.4 66.3 63.7 66.0 60.7 65.2 64.3 64.6 67.8 65.4 66.6 65.0 68.0 64.6 61.5 64.2 61.4 63.0 57.8 62.9 57.8 62.5 60.9 62.6 .59.4 61.9 59.5 61.2 59.7 67.0 60.1 60.6 68.9 60.2 58.6 59.7 65.i 61.2 59.2 60.3 55.7 .^3.4 57.7 55.5 57.4 59.1 57.1 57.0 .56.7 53.0 56.3 .53.9 .^1.3 54.3 54.9 56.6 5i.7 59.7 .54.5 55.1 54.5 53.9 53.6 47.6 ,55.0 48.4 .'■>2.5 47.8 ,5-?. 5 63.6 51.5 48.7 51.3 47.9 50.5 48.9 ,'-)0.2 49.7 49.7 .51.6 49.5 66.5 49.5 46.8 49.2 43.1 45.4 43.8 47.4 44.7 47.0 45.6 46.5 43.4 46.5 43.0 46.5 64.8 51.3 40.1 52.4 43.0 /,6.2 40.7 45.5 38. 4 44.5 40.3 44.1 42.9 45.6 44.2 45.4 38.4 42.7 41.3 42.4 43.7 41.9 45.5 41.7 44.5 41.5 41.6 40.4 43.0 40.4 43.6 iO.2 37.3 .19.9 34.7 39.7 36.3 55.7 38.0 55.1 34.8 55.2 32.2 38.0 43.0 .S7.3 43.9 .S7.S 38.3 57.5 40.8 37.6 36.7 37.3 33.7 36.5 36.4 55.5 39.3 56.5 40.2 56.2 39.8 35.7 46.9 39.8 Annual average : Air 60.3; Water 57-1. Table XXXV shows the average daily temperature of the surface water in the harbor, and of the air, at 2 p. m., for the period of five years from 1882 to 1886. The roman figures show the air temperature, and the italic figures the water temperature. MARYLAND WEATHER SERVICE 147 The highest annual temperature occurred in July in the great major- ity of cases in the past 34 years; it has never appeared earlier than June 3; on two occasions it fell in the months of September, on the 7th in 1881, and on the 11th in 1887. The average interval between the occurrence of the lowest and highest temperatures of the year is 181 days, from January 25 to July 15. The longest was 250 days, from January 1 to September 7, 1880; the shortest was 116 days, from February 10 to June 6, 1899. The details of occurrence, together with the minimum and maxi- mum temperatures recorded, are shown for each year since 1871 in Table XXXIV. The length of the intervening period is represented graphically in Fig. 37. Temperature of the Water in the Harbor. From September 1, 1881 to March 31, 1887, observations were made daily at 2 p. m. of the temperature of the surface water in the harbor, from the wharf at the foot of South Street. At the same time the tem- perature at the bottom (a depth varying from 9 to 12 feet according to the tide) and the temperature of the air were also noted. The average values of the surface water temperature and the air temperature for the five years from 1882 to 1886 are presented in Table XXXV. Fig. 38 also shows graphically the mean temperature of the water at the surface, and at the bottom, and of the air at 2 p. m. The temperature of the surface water is approximately 5° to 6° cooler than the air temperature from February to July. The difference diminishes gradually from July to October. In October the temperatures are approximately equal, and i-emain so until December when there is again a gradual divergence. The difference between the temperature of the surface water and that at a depth of ten feet is very small at all seasons of the year, averaging about five-tenths of a degree. 11 148 THE CLIMATE OF BALTIMORE HUMIDITY. Introduction. — Two distinct gaseous envelopes surround the earth; one is the dry air, with small quantities of other relatively permanent gases; the other is the vapor of water which may be condensed into visible forms of dew, frost, cloud, rain or snow, under ordinary conditions of temperature and pressure. It is of the highest importance, in the consideration of climatic con- ditions, to understand the functions and the distribution of the element of water in the atmosphere, in its great variety of forms and proportions. Water is being changed into invisible vapor from the ice and snow of the frozen north no less constantly, though less abundantly, than from the warm ocean waters of the tropics. As a result, the atmosphere is never free from the vapor of water. It may be present in small quantities only, as in the dry desert regions of the earth, or in the cold zones of the north and south, or in the rare and cold atmosphere of the mountain tops. The vapor capacity of a given space increases rapidly with in- crease in temperature. A cubic foot of vapor at a temperature of 50° F. at normal sea-level pressure, and at saturation, weighs about 4 grains ; at 70° it weighs 8 grains; and at 100°, about 20 grains; hence with an increase in temperature from 50° to 70° the quantity of moisture at saturation is doubled. The invisible vapor of water in the atmosphere is generally referred to as the humidity of the atmosphere. When the amount of vapor is actually weighed in grains, ounces, or pounds, or when it is measured in terms of pressure, as so many inches of mercury, it is referred to as absolute humidity. When it is measured in terms of percentage of the total amount which can exist in a given portion of the atmosphere, it is referred to as the relative humidity. For example, as stated above, the total quantity of invisible vapor which may be contained at ordinary pressure, in a cubic foot at 70° temperature, is 8 grains. The atmosphere is then said to be saturated, and the relative humidity is 100 per cent. Suppose the amount of vapor to be reduced to 4 grains, or one-half the full capacity, the temperature remaining the same, the percentage of MAEYLAND WEATHER SERVICE 149 the relative hiimidity would then be 50. The point of saturation is also called the dew point. As the temperature of the atmosphere diminishes rapidly from the earth's surface upward, the capacity for water vapor diminishes, and at a more rapid rate. Calculations have been made of the amounts of aqueous vapor which the atmosphere can hold in suspension at different temperatures and below given altitudes. In the following table by Fer- rel, the figures given show the depth in inches of water which would result if all of the vapor which it is possible for the atmosphere to hold in suspension under the given conditions were condensed to water. AMOUNT OP AQUEOUS VAPOR IN SATURATED ATMOSPHERE OF STATED TEMPERATURE AND DEPTH. Elevation. 80° F. 70° F. 60° F. 50° F. 6,000 feet. 1.3 inch. 1.0 lE ich. 0.1 inch. 0.5 inch. 12,000 2.1 1.5 1.1 0.8 18,000 2.5 1.8 1.3 0.9 24,000 2.7 2.0 1.4 1.0 30,000 2.8 2.1 1.5 1.1 The conditions assumed probably never occur in nature, as the rapid decrease in temperature with elevation would preclude the possibility of an average temperature sufficiently high, or at unifonn saturation, to the given elevations. Probably under the most favorable conditions of a moist atmosphere in the tropics, the amount of vapor from the earth's surface to the upper limits of the atmosphere, if condensed, would measure less than two inches. The decrease in the amount of vapor in the atmosphere with decrease in temperature may be stated in another form. At the equator the amount of vapor may reach about 11 grains per cubic foot, or about 20 tons per cubic mile; at latitude of 40° it may reach about 5 grains or about 9 tons per cubic mile, assuming a temperature of 55° as an average for the year; at latitude 70° with an average temperature of 30°, it may attain about 2 grains per cubic foot, or about 3.5 tons per cubic mile. It has been estimated that one-half of the total quantity of vapor in the entire atmosphere lies below an elevation of about (>500 feet, or below the summit of !Mt. Washington ; hence the decrease in quantity is very rapid 150 THE CLIMATE OF BALTIMORE and we may easily realize how mountains of moderate elevation, and even the higher hills, may afi'ect the rainfall and cloudiness of a locality. As the absolute humidity of the atmosphere at any given place depends upon the local temperature and a local water supply, there is a steady decrease in the amount of water vapor from the equator to the poles, from the surface of the earth upward, and from the oceans toward the interior of the continents. This general law of distribution may, how- ever, be modified, and even completely reversed, by the direction and character of the wind movement over a given area. While the vapor of the atmosphere is taken up by evaporation from a variety of surfaces, such as lakes, rivers, moist fields, or the foliage of the forest, the great source of supply must always be the surface waters of the equatorial and tropical oceans. From these it is carried up by the winds and distrib- uted to all parts of the globe. This invisible moisture of the atmosphere has a most important func- tion to perform in tempering the heat and cold. Where it is found in abundance, extremes of temperature are unlikely. Its absorbing power is great compared with that of dry air, and the earth's surface is pro- tected against the heat of the sun by day, and the rapid cooling by radiation during the night. Its abundance in the tropics is largely responsible for maintaining the uniformly high temperature of those regions. Its absence in limited areas of the tropical and sub-tropical zones is marked by great diurnal fluctuations in temperature, as in the arid regions of the southwest. Most important of all, it is the great source of supply of the rainfall and snowfall of the world; without first passing into the form of vapor, the oceans and rivers would have but little effect in watering the fields and forests of the earth. While the benefits of the atmospheric vapor are numerous and appar- ent, it may also be a source of much personal discomfort. When the relative humidity is high, that is, when the vapor is near the saturation point, or dew point, evaporation from the body becomes sluggish, or ceases, the air begins to feel muggy, when the temperature is high, or raw when it is low. A temperature which would be considered mod- erate with a dry air becomes oppressively hot when the humidity ap- MARYLAND WEATHER SERVICE 151 Mdt 3 6 9 Noon 3 6 9 Mot Mdt. 3 6 9 Noon 3 6 9 Moi JAN. 66 ^ ' / / / APRIL kJ MCT, 3 9 Noon 3 9 Mdt. — ~ -N /- YEAR -^ Mot 16 9 Njon 3 6 9 Mdt Mdt 3 6 9 Noon 3 6 9 Mdt ^ \ / I j JULY J ^ J 1 ' OCT. J Fig. 39. — Mean Hourly Relative Humidity. Till- mean hourly relative humidities for the months of January, April, July, October and for the year are oxiircssed as pcn-entii^'os, coniplctc saturation boiu},'- reprcsouted by 100 per cent. The curves are based on the 'M months' record of a Richard hytjrograph. 153 THE CLIMATE OF BALTIMORE preaches or exceeds 80 per cent. On the other hand, a cold of 15° or 20° above zero with a humidity of 80 per cent, a condition which is common in the Atlantic coast states, will cause much suffering, while temperatures of 20° below zero in the northwest, with a humidity of 25 or 30 per cent, are described as comfortable and exhilarating. The atmosphere does not lose in transparency with increasing humidity either relative or absolute; on the contrary, a high humidity is often accompanied by greater clearness, and an unusual transparency has come to be regarded as a sign of rain. When, however, the vapor reaches the dew point, just beyond the point of saturation, we have a series of phenomena of the highest interest and importance to us, resulting from condensation into the visible forms of dew, fog, clouds, rain, snow, frost and hail. The particular form of condensation is primarily a function of temperature, modified by local conditions of topography, elevation above the earth's surface, and the movements of the atmosphere. Hourly Variations in Humidity. Continuous automatic records of the variations of relative humidity are available for about two years and a half at Baltimore. While this is table xxxvi.-mean hourly relative humidity. (1902-1904.) 1 A. M. 2 3 4 5 fi 8. 10 11 Noon... 1 P. M. 2 3 4 5 . 6 10 11 Midnight . Jan. Feb. Mar Averag-e. 62 61 59 58 58 60 62 64 66 er 70 71 71 72 73 74 76 75 71 70 66 64 62 60 58 59 60 62 64 66 67 68 70 70 78 .2 67.4 70.6 62.6 Apr. 76 May June July Aug. Sept. Oct. 76 80 80 77 70 66 60 55 51 50 47 46 46 48 60 54 60 64 6S 71 73 75 .4 67.4 70.4 72.1 72.4 Nov. Dec 72 72 72 74 74 74 74 72 66 61 56 63 50 50 51 62 66 60 62 65 70 70 70 70 70 70 70 70 66 62 69 56 64 54 55 56 69 61 63 64 66 5 64.3 64.0 67.4 An'l MARYLAND WEATHER SERVICE 153 a much shorter record than that utilized in the discussion of tempera- ture, pressure, wind direction and other factors, it is still of sufficient length to enable us to establish firmly the form of the diurnal curve. The diurnal variation in the relative humidity is represented by a simple curve with its maximum point at about 5 a. m. for the year, but varying between 4 a, m. and 7 a. m. according to the season. The time of maxi- mum follows closely the time of sunrise. The minimum point, or the drj^est time of day, occurs between 1.30 p. m. and 3.30 p. m. The 70 s H. 70 Fig. 40. Mean Hourly Relative Humidity. As in Fig. 39, the hourly humidities are expressed as percentages, 100 per cent repre- senting complete saturation. The light shades represent the lower humidities, or the dryer portions of the day and year ; the heavy shades, the time of higher humidities. The dotted lines, S. R. and S.S indicate the time of sunrise and sunset respectively. The diagram is based on the 30 months' record of a Richard hygrograph. details of the hourly variation are shown statistically in Table XXXVI, and graphically in Fig. 39 and Fig. 40. The seasonal distribution is revealed at a glance in Fig. 40, in which the light shades represent the lower humidities, or the dryer portions of the day, and increase in the density of the shades shows an increase in the humidity. It will be observed that the dotted line representing the time of sunrise- (S. K.) passes through the areas of heaviest shading, approximately in their central portions. The values in both tables and figures are shown in percentages, total saturation of the atmosphere being represented by 100. 154 THE CLIMATE OF BALTIMORE The actual values for relative humidity from hour to hour during the day fluctuate rapidly on days with unsettled weather. The changes are particularly marked during the course of a thunderstorm. There is frequently very little resemblance between the curve showing actual con- ditions and that representing average conditions for a month or more. Every change in the direction of the wind, or in the temperature, is reflected in the form of the actual curve. In Plate VIII some typical relative humidity curves made by the self-recording instrument are repro- duced and they may thus be compared with the curve in Fig. 39 repre- senting average values. Direct observations of relative humidity were made at varying hours of the day from time to time in past years. The continuous record enables us to determine the corrections to be applied to any of the com- binations of hours employed in the past in order to obtain a correct daily mean based on 24 hourly observations. A daily mean derived from any of the series of three observations per day, one in the morning between 7 and 9, one at 3 or 4 in the afternoon and the other between 9 and 11 at night, gives a value so nearly equal to the true mean based on 24 hourly observations that the corrections to be applied fall within the limits of probable error of observation, and hence may be neglected. For the series of observations made at 8 a. m. and 8 p. m. the departures from the true average are large enough to require the application of the necessary corrections shown in the following figures : CORRECTIONS TO OBTAIN TRUE DAILY MEAN HUMIDITY. (Expressed in percentages.) Hours of Jan. Feb. Mar. Apr. May June Julj- Aug-. Sept. Oct. Nov. Dec. Year observation. ( 8a.m. + 8 p.m. ) -3.3° -3.1° -1.9° -1.9° -1.6° -0.6° -1.6° -1.9° -2.6° -4.0° -4.2° -3.0° -2.6° 2 Phases of the Diurnal March of Eelative Humidity. The time of occurrence of the maximum and minimum values for the day, and the time, in hours and minutes, when the humidity is the same as the mean for the entire day, are sho^^^l in the following table and in Fig. 41 : MARYLAND WEATHER SERVICE 155 PHASES OF THE DIURNAL MARCH OF RELATIVE HUMIDITY. S s ^,S § >> 3 3 1-5 >-s OQ O O o a> Max. (a.m.) First mean (a. m.) — Min. (p.m.) Second mean (p. m.). 7.30 7.20 5.40 4.30 4.30 4.00 4.30 5.30 5.00 4.30 5.30 4.30 5.00 10.00 10.40 9.20 8.50 8.30 8.30 8.20 8.40 8.40 8.50 9.20 9.30 9.30 2 30 3.20 3..30 3. 00 2. 30 2. 00 3. 00 2. 30 1.30 2. 30 1.30 1.30 2. 80 7'.00 9.00 8.40 8.40 7.50 8.20 7.50 8.00 7.20 7.20 7.40 8.00 7.60 JFMAMJJASONDJ 6 A. M. / "^ \, ^ A % r "^ .^ , B / '^ s.^^ ^ "^ \^ y N^ y V y •^ — ' ^ ^ ^ V S ^ — 1 ■ — C , ■ , ^. I N N, i \ /^ -^^ ,^ ^V ' ^^ 0^ y '^ V / 6 P. M. Fig. 41. — Phases of the Diurnal Variations in Relative Humidity. B shows the variations in the hour of occurrence of the dryest time of the day from month to month ; D the time of occurrence of the dampest portion of the day ; A and C show, respectively, the afternoon and the morning hours when the mean humidity of the day is most likely to occur. 156 THE CLIMATE OF BALTIMORE Mean Monthly and Annual Eelative Humidity. The observed values of the relative humidity for the entire period from 1871 to 1903 were reduced to true mean monthly and annual values based upon hourly observations; these corrected figures are contained i F M fi ^ J A s o ^ D J 70 N N, /^ ^-^ / ■\ V ^ / 60 V y r — 50 7Q 50 Fig. 42. — The Mean Monthly Relative Humidity. The diagram is based on direct observations at two or three stated periods of the day during a period of 30 years; the average values were corrected for the diurnal variation, and expressed as percentages of total saturation. IB/S 1880 1885 lo »0 1895 1900 V»._^ /\ f\y\ /■\ '^Vy V Fig. 43. — Variations in the Mean Annual Relative Humidity. (Expressed as percentages of total saturation.) in Table XXXVII, together with the ten-year monthly averages and the normals for the entire period. The monthly normals and the variations in annual means are also graphically shown in Fig. 43 and Fig. 43 respectively. jVIAEYLand weather service 157 The normal amount of moisture in the atmosphere for the entire year is about two-thirds of the total capacity of the atmosphere for water vapor, namely, 66.5 per cent. The mean monthly amounts vary from season to season, being greatest in the month of September (70.4) and TABLE XXXVII.— MEAN MONTHLY RELATIVE HUMIDITY. Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. An'l 18T1 68 63 66 75 73 68 74 72 71 73 71 74 74 67 65 76 69 67 67 66 67 76 72 69 70 63 66 70 66 71 69 70 66 66 68 60 73 68 66 66 68 70 66 66 66 66 68 63 69 71 68 61 70 69 75 69 67 68 66 71 63 73 68 69 73 62 82 68 64 66 68 66 70 64 64 63 63 63 66 66 66 63 60 64 61 64 70 71 65 61 60 69 63 67 72 66 73 76 73 66 50 59 66 60 56 66 64 57 56 69 60 65 57 56 69 69 51 63 60 65 63 68 68 63 62 65 57 61 69 63 68 54 66 60 65 69 51 64 64 63 58 68 66 70 59 59 65 71 68 68 66 67 59 68 66 67 66 66 60 68 63 69 72 64 66 68 60 64 64 63 66 67 64 64 63 69 60 68 66 63 71 66 64 73 64 71 76 73 66 69 69 64 61 67. 71 68 66 75 65 60 63 63 65 63 68 66 61 64 61 66 67 65 63 73 71 66 73 62 77 71 65 64 64 68 70 66 68 63 74 73 65 69 65 78 62 76 69 62 73 70 71 60 74 63 69 69 71 70 68 70 70 78 73 65 68 64 63 68 73 70 68 76 69 76 65 66 76 72 66 73 71 73 70 67 68 74 70 63 67 70 70 74 75 74 78 69 67 73 67 66 65 65 69 83 74 75 70 66 62 73 66 67 66 71 66 68 67 67 76 73 64 76 66 64 65 64 68 71 63 71 67 63 63 70 69 71 80 66 74 63 65 67 69 63 66 70 68 67 63 66 67 65 64 61 69 61 60 67 71 63 70 71 68 60 70 65 66 67 67 72 66 73 66 63 58 71 66 74 71 66 68 69 69 70 65 67 68 62 71 68 62 67 62 67 71 68 68 68 63 72 66 64 73 77 70 57. 67.4 1872 58.8 1873 66.1 1874 66.4 1^75 . ... 66.3 1876 1877 66.3 67.7 1878 67.2 1879 66.4 1880 1881 1882 1883 65.0 65.6 67.7 65.9 1884 64.4 1885 64.3 1886 1887 1888 1889 69.0 66.3 65.3 67.5 1890 1891 1893 1893 66.8 69.3 70.5 68.1 1894 65.7 1895 64.2 1896 64.3 1897 65.8 1898 1899 .... 65.8 67.5 1900 69.3 1901 1902 69.7 70.1 65.2 Average, 1871-80 1881-90 1891-1900 70.2 69.5 ra.o 66.9 66.8 67.6 66.3 61.4 65.4 59.8 69.9 60.0 59.8 65.9 64.3 64.3 66.3 68.7 63.8 66.6 67.5 69.3 68.4 68.8 69.7 70.5 70.3 67.1 68.0 67.6 65.4 64.7 67.6 67.5 66.3 67.8 65.7 66.2 67.0 1871-1903 69.6 66.6 64.9 eo.i 63.6 66.6 66.4 69.3 70.4 67.6 65.8 67.2 66.6 least in the month of April (60.1). For individual years, the monthly averages vary considerably. For example, the average humidity for the month of March has been as high as 82 per cent and as low as 54 per cent, a range of 28. A similar range has been experienced in the month of October. The month possessing the smallest range in the monthly 158 THE CLIMATE OF BALTIMORE average value is January, with a maximum of 76 per cent and a minimum of 62 per cent, a range of 14. The range in actual conditions of humidity, as distinguished from average conditions for a considerable period, shows, of course, much greater fluctuations. As the upper limit, namely, 100 per cent, is reached at all seasons of the year during misty or rainy weather, the lower limit is the index of variability. The lowest values are most likely to occur during the clear, cold days of winter or early spring. During the two years and a half in which continuous automatic registration was main- tained at Baltimore, the humidity during the afternoon hours occasion- ally fell to 25 per cent, or one-fourth the moisture capacity. The occa- sions upon which the moisture content fell below this percentage were rare. A few of the exceptionally dry days during this period are here cited : EXCEPTIONALLY DRY DAYS. January 31, 1903, minimum humidity was 20 per cent. April 5, 1904, " 11 •• • May 4, 1904, " 15 " August 17, 1902, " 24 " ' October 30, 1903, " 14 " November 9, 1902, " 22 " The limits of variability in the annual average humidity during 33 years were 70.5 per cent in 1892, and 58.8 per cent in 1872, a range of 11.7 per cent. The ten-year averages have only varied from the normal for the entire period by the following small departures: 1871 to 1880 0.8 below normal ; 1881 to 1890 0.3 below normal ; 1891 to 1900 0.5 above normal. Absolute Humidity. Expressing the humidity in the terms of the actual weight of the water vapor in the atmosphere at different hours of the day throughout the year, the distribution is shown in the following table. These figures show the average amount of water present in the atmosphere at the hours specified during the five years from August, 1881, to July, 1886. As the amount of moisture in the atmosphere is primarily a function of the temperature of the air, we find a steady increase in the absolute SELECTED RELATIVE HUMIDITY CURVES. MARYLAND WEATHER SERVICE 159 humidity from January, the coldest month, to July, the warmest month of the year. MEAN ABSOLUTE HUMIDITY. (Weight of the vapor of water in grains per cubic foot.) Hours of Observation 1.41 1.47 1.60 1.6" 11 p.m 1.54 7 a. m 11 a. m . 3 p. m . 7 p.m Jan. Feb. 1.58 1.66 1.71 1.72 1.72 aiar. Average 1.619 1 1.78 1.78 1.84 1.91 1.82 Apr. 2 72 2171 2.78 3.01 2.93 1.827 2.828 3.997 5 May 3.89 3.80 4.02 4.20 4.08 June 5.56 5.50 5.51 5.79 5.82 July Aug. Sept. 6.50 6.12 6.11 6.68 6.71 5.98 6.06 6.16 6.38 6.38 5.29 5.53 5.48 5.79 5.61 Oct. Nov. Dec. 3.83 3.87 4.06 3.96 4.03 6.425 6.193 5.539 3.951 2.411 1.836 3.267 2.26 2.36 2. 47 2.48 1.74 1.82 1.91 1,88 1.83 Year 3.082 3.222 3^267 3.437 3.327 Mean Vapor Pressure. The average monthly values of the tension of the vapor of water in the atmosphere, based upon observations of temperature of the air and the temperature of the wet-bulb thermometer at 8 a. m. and 8 p. m. from 1892 to 1903, are shown in the following table expressed in frac- tions of an inch of mercury: Jan. Feb. Mar. ( April May June July Aug. Sept. Oct. Nov. Dec. .152 Year .136 .133 .193 .252 .394 .549 .631 .614 .512 .334 .222 .344 PRECIPITATION. Introduction, The German meteorologist. Dove, has aptly compared the atmosphere to a huge still, of which the sun is the furnace, and the sea the boiler, while the cold air of the higher elevations and of the temperate zones plays the part of the condenser; we, on a wet day, catch some of the liquid which distills over. The condition and method of formation of dew, fog, cloud, rain, snow and hail are admirably and concisely stated in the following extract from one of Dr. Hann's recent treatises : * ' Hann, J. Allgemeine Erdkunde. I. Abtheilung. 8° Wien, 1896. pp. 173 et seq. 160 THE CLIMATE OF BALTIMORE " Condensation of moisture gives rise to numerous phenomena which are collectively called liydrometeors. It takes place whenever the tem- perature falls below the dew point. Hence whatever favors a lowering of the temperature of the air favors the production of dew, fog, cloud, rain, snow and hail. " The ground cooling rapidly during a clear, calm night by radiation, lowers the temperature of the air resting upon it; we then have dew. With a temperature below freezing we will have frost. Warm, moist air mixing with colder air near the earth's surface will give us fog; thus we see fog formed over rivers in the early morning. Over the Banks of New Foundland we have a cold body of water over which pass warm, moist southerly winds, producing almost continual fog. With a tem- perature below freezing point the fog collects on trees and shrubs in the form of hoar frost. Mixing of air currents at different temperatures high above the earth, and rising and cooling moist air, produce clouds. Fog and clouds are composed of small drops of water. In winter and at high altitudes they are composed of ice crystals. High cirrus clouds all the year round are composed of ice crystals. At an elevation of 3500 meters (11,483 feet) in the middle latitudes the air temperature is below the freezing point throughout the year. The presence of ice crystals in the higher clouds is shown by the colored rings about the sun and moon. These ice crystals and drops of water float in the air because of the great amount of surface exposed to the resistance of the air in comparison with their weight. "As fog and cloud increase in density the drops coalesce and become larger, until they become too large and heavy to be supported by the resistance of the air; they then go over into rain. Should the rain pass through drier strata of atmosphere it may be reabsorbed. In winter condensed water crystallizes and falls as snow. In stormy weather and a temperature near freezing, the snow packs and balls and falls as snow pellets. As these become firmer and pass alternately through layers of air above and below freezing point they become coated with water in the warmer layer and with ice in the colder layer, and thus form hail. " With an increasing quantity of vapor in the atmosphere the entire MARYLAND WEATHER SERVICE 161 heavens become of a whitish dimness and lunar or solar halos appear. In most cases, however, the condensed vapor is not uniformly distributed in the atmosphere, but is collected in masses which float in the air, re- flecting the light and throwing shadows. We then call them clouds." The Causes of Precipitation. The theory that rainfall is primarily a result of the mixing of moist air currents at different temperatures has lost much in faVbr among meteorologists of to-day. Calculations have shown that the rain result- ing from such a cause must be comparatively light, and would not at all account for the abundant precipitation of the tropics, or the heavy falls connected with the movement of storms. Clouds may undoubtedly be accounted for upon the supposition of a mixture of air currents, with high humidity, especially the stratified forms, but even here it is neces- sary to find a more abundant source of condensation. This is found in the agency of a rising current of air. As already stated, the atmosphere contains at all times a considerable amount of moisture, especially in the lower layers of the warmer climates. Conditions are favorable for an upward movement of the air wherever there are opposing currents or where there is a considerable difference in temperature over adjacent areas. Such conditions are always present within storm areas, or, on a larger scale, in the equatorial region where we have the northeast and southeast trades meeting and causing a slow upward movement of the atmosphere at all times. Eising currents with the attendant decrease in temperature, combined with the presence of moisture, are capable of accounting for the heaviest rainfalls recorded. The Geographical Distribution of Rainfall. The rainfall of a locality being primarily dependent upon the -quantity of moisture in the atmosphere, that is, the absolute humidity, and this being in turn dependent upon the temperature, we find a steady decrease in the- amount of precipitation from the equator to the poles. In the equatorial regions the annual rainfall averages about 75 inches, while within the Arctic circle it is reduced to less than 10 inches. The decrease 163 THE CLIMATE OP BALTIMORE is not at all uniform, as other factors enter into the problem of distri- bution to such an extent as to overbalance the effect of the temperature. The prevailing distribution of atmospheric pressure, the prevailing wind direction, and topography, may separately or in combination be the determining factors in the distribution of rainfall over any given area. The Influence of Wind Direction. The effect of wind direction may readily be inferred from what has already been stated concerning the sources of atmospheric moisture. The oceans being the great source of supply, and the winds being the carriers and distributors of moisture, it follows that rainfall is most copious along the coasts, and decreases with distance from the coasts when the wind direction is from the oceans inland. In the United States there is a steady decrease from the Gulf and Atlantic coasts toward the central portions of the continent, being 50 inches to 60 inches in the southeast, and about 15 inches in the Eocky Mountain region. On the Pacific coast a similar decrease obtains but is here greatly modified by the presence of high mountain ranges near the coast. The same is true in a general way over all of the continental areas. The Influence of Topography. The distribution of rainfall along the Pacific coast from California northward affords an excellent illustration of the influence of mountain ranges crossing the path of rain-bearing winds. During the rainy season the moist and comparatively warm winds from the Pacific are first forced up over the Coast Eange and again over the Sierra Nevadas to altitudes varying from a few thousand to over ten thousand feet. The moisture is cooled below the saturation point and rains are copious. The winds descend upon the eastern slope of the Sierra Nevada Mountains com- jjaratively dry. While the annual rainfall on the Pacific coast decreases from about 50 inches to 10 inches in passing over a distance of about two hundred miles inland, a similar decrease from the Atlantic and Gulf coasts extends over 1500 to 2000 miles. In northern India the elevated range of the Himalaya Mountains lies MARYLAND WEATHER SERVICE 163 directly in the path of the southwest monsoon winds. The winds, blow- ing for days over the warm waters of the Indian Ocean, reach the moun- tains heavily laden with moisture, and are forced up the southern slope to elevations of 20,000 feet and more before they can proceed further on their way to the deep and persistent barometric depression which is cen- tered to the north of the mountains during the warm season. Among the foothills on the southern slope of the mountains, just north of Cal- cutta, and at an elevation of about 4500 feet, lies the station of Cherra Poongee, which has the heaviest annual rainfall in the world. The aver- age annual fall for 25 years approximates 475 inches; the annual amount has varied from about 300 inches to over 900 inches, nearly all of which falls in the six months from April to September. The Influence of Atmospheric Pressure. As previously pointed out, conditions which favor the production of rising air currents are favorable to the production of rain. Areas over which the barometer is relatively high are apt to be poor in rainfall, and areas with a low barometer in comparison with adjacent areas are apt to be comparatively rich in rainfall, other conditions being equal. This broad generalization may be verified by almost any daily weather chart which may be consulted, and is familiar to all who have occasion to study the weather charts issued by the United States Weather Bureau, or those of any other nation issuing such charts. When we see upon these charts an enclosed area of low barometer, or a "Low," as it is familiarly called, cloudiness and rain prevail within this area; where the barometer is high, relatively, the skies are prevailingly clear or partly clouded. " High area " weather is proverbially " fine " weather ; " low area " weather is generally " bad " weather. As the winds flow toward the center of an area of low barometer from all sides, the air at and near the center must necessarily rise, and rising, it is cooled. If it rises high enough, cloud formation and rain follow. On the other hand, where the barometer is relatively high, the air descends and the winds blow out from the central portions of the high area in all directions. We have seen that ascending air is cooled by expansion and radiation as it rises; 12 164 THE CLIMATE OF BALTIMORE conversely, descending air is warmed by compression. As it warms, its capacity for moisture increases, and not having an opportunity to take up more moisture in its descent, it becomes relatively drier; its relative humidity is decreased. Clouds which may be within this area at the beginning of its formation tend to dissolve, and near the center where the descent of air is most active, the skies are apt to be clear. These areas of low and high pressure (cyclones and anticyclones as they are technically termed) move from west to east in rapid succession in the middle latitudes and constitute the distinguishing feature of our weather. In equatorial regions there is a belt of varying width in which the pressure is constantly lower than it is to the north or south. Within this belt, situated between the northeast and southeast trade winds, the air has, in addition to its westward drift, an upward movement, produc- ing the "cloud belt" or doldrums with its almost daily copious showers. To the north of the northeast trades and south of the southeast trades there are broad belts, most regularly developed over the southern hemi- sphere where the surface conditions are more uniform, in which the pressure is relatively high. Here the air has a descending tendency and these areas are characteristically dry. Within them are the great desert regions of the globe — the Sahara, the Arabian desert, the arid regions of Australia, as well as those of our own Southwest. Over the oceans they are known as the " horse latitudes " with their light winds and scanty rainfall. The Seasonable Distribution of Eainfall. Climates are often classified according to the manner in which the rainfall is distributed through the year. We have regions of perennial rainfall, as in the United States east of the Mississippi Eiver, where there is a fairly uniform precipitation throughout the year. This is a condi- tion which prevails with limited exceptions between latitudes 35° and 60°. Within the tropics, and over limited areas elsewhere, as in the upper Missouri Valley, there are large areas in which most of the annual fall of rain occurs in the summer months, with light rain in winter and spring. In other regions, as in the Pacific coast states, nearly all of the precipitation occurs in the winter months with little or no rain in sum- MARYLAND WEATHER SERVICE 165 mer. Lastly there are the arid regions of the world which are nearly free from rain throughout the year. Hourly Amount of Eaixfall. A diurnal period in the relative amounts and frequency of rainfall is most distinctly revealed in tropical countries, but is still clearly shown in the summer months of the middle latitudes. The precipitation which occurs in connection with the movement ot a general barometric depres- sion has a fairly uniform distribution throughout the day; that occur- FiG. 44. — Average Hourly Precipitation. The average amount of rainfall or snowfall during each hour of the day, for every month of the year, is shown by the heavy black lines and the shaded areas. The light shades show the time of day and year when the precipitation is usually lightest. The figures attached to the curved lines show the amount of the precipitation In hundredths of an inch. The values are based on the ten years' record of a tipping-bucket rain- gage and on eye observations. Only days with an appreciable amount of precipitation were considered In determining the average amount of precipitation for the day. ring in connection with thunderstorms is restricted mostly to the after- noon hours, and is intimately associated with the diurnal variation in temperature and pressure. The most conspicuous feature of the diagram representing the hourly quantity of rainfall (see Fig. 44) is the uniform distribution of the pre- cipitation throughout the day during the winter and spring months. This is doubtless explained by the fact that the winter and spring snows and rains occur in connection with the more or less regular succession of 166 THE CLIMATE OF BALTIMORE the cyclonic disturbances of the middle latitudes whose eastward progress is but slightl}', if at all, affected by the diurnal variations of temperature and pressure. On the other hand, the intensity of summer rains has a distinct diurnal period, being light in the forenoon, increasing rapidly to noon, or 1 p. m., and then more slowly to a maximum at about TABLE XXXVIII.— TOTAL HOURLY RAINFALL PER MONTH AND YEAR. (In hundredths of an inch.) Hours. i 05 p. ^ 0) a 3 ^ bi 3 1 3 +3 > oil S 3 n .16 .16 < S >-i >-> .09 , § >. bfi 4i 4J > d a a ■^ 00 eg Hours. O ,2 8* 2.0 1-5 2.8 1.7 CO 2.4 3.0 3.8 3.0 3.1 <^^ Md't. to 1 A. M 37.1 1 " 3 " 3.0 3.7 3.7 3.5 3.7 1.9 3.4 1.6 1.6 3.0 3.8 3.0 2.9 34.9 2 " 3 " 3.0 3.5 3.6 3.6 3.4 1- 7 3.1 1.3 1.3 2.9 3.6 2.6 ''.7 33.5 3 " 4 " 3.3 3.8 3.9 3.4 3.6 1.8 1.9 1.3 1.3 2.8 3.4 2.9 2.S 33.4 4 '• 5 " 3.2 4.1 3.8 3.S< 4.1 1.7 l.H 0.9 1.4 2.5 3.4 3.3 2.8 34.0 6 '• 6 " 3.6 4.3 4.6 4.7 4.3 1.6 1.9 18 1.8 3.0 3.7 3.3 3.2 38.5 6 " 7 " 3.7 4.3 5.2 4.6 4.8 2.0 1.8 1.8 2 2 3.2 3.5 3.3 3.4 40.4 " 8 " 4.4 5.4 5.7 5.5 6.0 2.9 2.4 2. '^ 3.5 4.1 3.7 4.2 4.0 48.6 8 " 9 " .5.4 5.6 5.4 4.6 6.6 2.5 2.9 3.4 3.9 3.9 4.1 4.6 4.2 49.8 9 "10 " 5.4 5.3 5.2 4.r, 3.9 2.2 3.6 1.7 3.3 3.5 4.5 4.4 3.8 45.5 10 "11 " 4.8 4.4 5.1 4.5 3.8 2.3 3.8 1.9 3.7 3.5 4.3 4.2 3.7 44.3 11 " Noon 5.5 4.6 5.3 3.6 3.9 2.7 3.3 3.0 3.6 3.3 4.1 4.3 3.S 46.3 Noon " 1 P. M 5.7 4.6 .i.2 4.4 3.9 3.1 3.3 2.4 2.7 3.0 4.5 4.3 3.9 47.1 1 '* 3 " 5.0 4.2 6.1 4.3 4.0 3.1 3.5 2.6 3.0 2.8 4.9 4.2 3.9 46.8 2 " 3 " 4.T 3.9 6.4 5.0 3.8 3.1 3.5 3 .0 3.2 3.0 4.3 4.1 3.9 46.4 3 " 4 " 4.7 4.3 5.3 :>.2 4.3 3.5 3.0 3!i 3.5 3.8 3.9 4.3 4.0 47.8 4 " 5 " 5.7 4.7 6.1 4.6 5.0 4.8 4.4 2.6 4.1 3.7 4.0 3.9 4.5 .53.6 5 " 6 " 5.0 4.3 6.0 4.9 4.7 3.7 4.3 3.3 4.0 3.2 4.6 3.9 4.2 50.8 6 " 7 " 5.3 5.1 5.0 4.8 5.3 4.1 4.4 3.9 3.3 3.3 4.4 4.7 4.4 53.4 7 " 8 " ....".'..'.'.'....'...'.'...'.'. 5.4 5.1 5.2 4.S 5.4 4.1 5.0 3.7 3.3 3.6 3.9 4.7 4.4 53.3 8 •' 9 " 5.4 4.7 5.3 4.2 5.6 3.4 4.8 3^7 3.4 3.0 4.3 4.5 4.3 61.3 9 "10 " 5.1 4.8 5.3 4.6 4.5 3.0 3.3 3.3 3.8 3.0 4.3 4.2 4.0 48.0 10 "11 " 5.3 6.2 5.3 5.3 3.9 2.3 3.5 2.7 3.0 3.1 4..") 4.1 4.0 48.1 11 " Md't 4.9 6.1 4.9 4.8 3.8 2.3 3.3 2.0 3.7 3.0 4.0 3.5 3.7 44.3 Table XXXIX. Average monthly and annual frequency of precipitation for each hour of the day. The figures indicate the average number of times rain or snow fell per month and year during each hour of the day. The results are based on the record of a self-registering rain gage for ten years, from 1893 to 1902. The distribution of precipitation throughout the day is quite uniform. During the period of ten years there were but few months in which rain or snow was not recorded at all hours of the day at least once. Precipi- tation has most frequently occurred between 4 p. m. and 5 p. m. in the month of March, namely, 61 times in 10 years. The hour of least fre- quency is from 4 a. m. to 5 a. m. in the month of August, with a total number of 9 times in 10 years. On the average there is a fairly well MARYLAND WEATHEK SERVICE 169 defined diurnal period in the frequency of precipitation. Beginning with a minimum in the early morning hours there is a rise in the average annual frequency to a maximum at about 5 p, m., followed by a return to the minimum in the early morning. This periodic movement is most readily seen in the July curve (see Fig. 46). In this month the mini- mum frequency (1.8) occurs at about 6 a. m. and the maximum (5.0) at 8 p. m. There is a comparatively rapid increase in frequency between 7 a. m. and 9 a. m., especially in the winter months. The most uniform DT Mot . \J A A V Jan \ / \J V /' t /-^ ^ / _^ \J \ JULV \ ■v / ■\y Fig. 46. — Average Houi'ly Frequency of Precipitation. The curves show the average frequency, for each hour during the months of January and July, of the occurrence of precipitation to the extent of .01 inch or more. The values are based on the ten years' record of a tipplng-bucket ralngage, supplemented by eye observations. distribution occurs during the cold months when the rains and snows are associated with the regularly recurring cyclonic disturbances. In the summer months the diminished effect of cyclonic disturbances is particu- larly noticeable in the reduced frequency of early morning rains, while the increasing afternoon rains are due to the summer thunderstorms, which reach a maximum frequency between 3 p. m. and 5 p. m. The month of March exhibits the most uniform hourly distribution of 170 THE CLIMATE OF BALTIMORE precipitation frequency, especially from about 7 a. m. until midnight. Precipitation is more frequent at all hours of the day during the winter and spring months than during the summer months. In Fig. 46 the January curve of frequency is well above the July curve throughout the day. This seasonal and diurnal distribution of precipitation frequency is graphically shown in Fig. 47, in which an increase in the intensity of shading represents an increase in the frequency of rains or snows. The figures attached to the heavy black lines show the average frequency Fig. 47. — Average Hourly Frequency of Precipitation. The heavy shades show the time of most frequent occurrence of precipitation during the day for every month of the year. The small figures attached to the irregular curved lines show the average number of times precipitation was recorded per month at the times indicated. of occurrence per month for each hour of the day and month of the year. The hourly and seasonal distribution is shown more accurately in Table XXXIX, in which the average frequency is recorded to tenths for all hours and months of the entire yesLT. Duration of Precipitation. Precipitation is not as continuous as it is generally supposed to be. If an automatic record of a rainy day be carefully examined, it will be found to be made up of numerous showers, some of them perhaps extending over an hour or two, but most of them lasting less than half an hour. The MARYLAND WEATHER SERVICE 171 duration and continuity depend mostly upon the position of the locality with reference to the center of the cj^clonic disturbance which is the occa- sion of the precipitation. As the character of the storm and the position of its path depend largely upon the season of the year, the duration of the accompanying precipitation is found to vary with the season. An automatic tipping-bucket rain gage has been in use at the Balti- more office of the Weather Bureau since January, 1893. The ten years' record enables us to obtain accurate values for the duration of the pre- cipitation. These records were supplemented by direct observations during the daytime. In a later chapter of this report some attention will be devoted to an analysis of the character of the rainfall accompany- ing different t3'pes of storms. In the table below an effort has been made to arrive at average values only for storms of all kinds. It has been found desirable to compute the average duration for three different con- ditions. The first line of figures of the following table includes " traces " of rainfall, i. e., amounts perceptible but too small to measure accurately by the ordinary methods. In the second line, only rains of measurable amounts have been considered, or precipitations equalling or exceeding one-hundredth of an inch in depth. The third line relates only to pre- cipitations which amounted to less than one-hundredth of an inch, or to " traces." AVERAGE DURATION OF rRECIPITATION. (In hours and minutes.) Class. Jan. Feb. Mar. Apr. May June Jul y| Aug. Sept. Oct. Nov. Dec. Year A. Including " traces." 10.00 1 13.10 !l0.20 11.00 7.30 4.30 4.00 4.00 6.30 8.30 10.50 10.00 8.20 B. Excluding " traces." 11.10 13.10 8.45 11.00 7.00 3.30 4.00 3.30 5.15 8.05 9.25 9.20 7.50 C. Traces only. 3.16 3.00 1 3.20 1 4.05 3.00 2.10 1.40 1 1.40 1 2.30 3.45 1 1 3.45 2.66 3.00 Class A contains what may be regarded as the most trustworthy fig- ures for the average duration of precipitation, as these averages express the entire period of precipitation in connection with a passing storm. The 172 THE CLIMATE OF BALTIMORE rains of the winter, spring and autumn months show the influence of the more frequent cyclonic storms, while the rains of the summer months A S O N D J / \ / / / ^ 1 I. y / / \ / , i f \ \ / \ \ / \ B / ^ /' \ V y \ J / /^ \ ^^ \ c y / V y Fig. 48. — The Average Duration of Precipitation. The upper curve (B) shows the variation in the average duration of rain and snow storms during which the precipitation amounted to .01 inch or more. The duration is expressed in hours and tenths of an hour ; the values are based upon a ten years' record of a tipping-bucket raingage supplemented by eye observations. The lower curve (C) shows the duration of light sprinkling rains, or light flurries of snow, with amounts too small for accurate measurements. are mostly of the kind accompanying thunderstorms. The former have a duration averaging more than double those of the latter. r^IAETLAXD WEATHER SERVICE 173 An examination of the figures in the table above will show that the average duration of rainfall or snowfall is a little less than eight hours, when only such storms are considered as yield an appreciable amount of precipitation. When " traces " are included, the average duration is somewhat above eight hours. The summer rains are less than half the duration of those of the winter, spring and fall; they are obviously of the thundershower type, while the rains of the winter, spring and fall occur mostly in connection with the cyclonic depressions. The entire interval between the beginning and ending of precipitation of each storm was considered as the duration of rainfall or snowfall, regardless of inter- ruptions in the continuity of the fall. The duration of the actual period of precipitation is something quite different from the duration of the general storm, or local atmospheric disturbance in connection with which the precipitation occurs. Dur- ing the passage of a storm over any given locality, there are likely to be many beginnings and endings of precipitation with intervals of a few minutes or a few hours with no precipitation, or of so small an amount as to be negligible. There is a certain t}^e of storm which is of com- paratively frequent occurrence in the Middle Atlantic states — the " northeaster ; " this storm is generally accompanied by heavy and per- sistent rainfall. But even in this class the intensity of rainfall varies greatly from hour to hour and it is generally made up of severjil showers with intervals of several hours without appreciable rainfall. A list of storms accompanied by an uninterrupted rainfall exceeding 24 hours in the city of Baltimore during the past ten years would not be a very long one, the number probably not exceeding three or four per year. The following list comprises rains of exceptionally long duration, which have occurred during the ten-year period ending with 1903. In these storms the rainfall was practically continuous, although in most of them there were intervals of a few hours during which only light sprinkling, or misting, rains were recorded. 174 THE CLIMATE OF BALTIMORE RAIN AND SNOW STORMS OF LONG DURATION. Am't. 1893, Apr. 19-21 44 hotus of actual precipitation. 1.14 In. 1894, Apr. 10-12 52 " " " 1.95 " Dec. 10-12 49 " " " 2.01 " 1895, Jan. 8-10 55 " " " 1.55 " Apr. 27-May 1 102 " " " 3.69 " Nov. 24-25 42 " " " 0.13 " 1897, Dec. 3-5 44 " " " 1.18 " 1898, Feb. 18-21 47 " " " 1.18 " Dec. 3-4 44 " " " 1.27 " 1899, Feb. 11-13 54 " " " 1.50 " 1900, Feb. 16-17 41 " " " 0.40 " 1902, Feb. 20-22 51 " " " 2.53 " Nov. 24-26 44 " " " 1.60 " 1903, Apr. 13-15 43 " " " 1.68 " The long-continued rains generally occur in connection with our '' northeasters/' depressions originating over the Gulf of Mexico, or in the West Indies, and moving northeastward directl}^ over ]\Iaryland, or following the Atlantic coast line. The most notable case in the list of long-continued rain storms is that of the spring of 1895, when rain, though sometimes very light, fell for practically 103 consecutive hours, beginning at 8 a. m., April 27, and ending at 2 p. m., May 1. The total precipitation for the entire period (3.69 inches) was not very large, though the rate of fall was at times excessive. There was no well-defined storm area near Baltimore at any time during the period. The bar- ometer was high over the New England states, while there was a shallow and ill-defined depression over the Gulf of Mexico which moved slowly northward and eastward some distance off the south and middle Atlantic coast, causing a steady northeast wind at Baltimore. Frequency of Precipitation of Stated Amounts. A table of monthly and annual precipitation as usually compiled may lead to erroneous inferences as to its agricultural value. The beneficial effects of rainfall depend not only on the quantity, but often to an equal extent upon the time of occurrence and the rate of precipitation. A given amount falling rapidly is of less value, agriculturally, than an equal or even less amount falling more slowly, as a rule. The greater portion of an excessive rain is apt to find its way to the streams immediately, while the lighter rains will soak into the ground to be utilized later in MARYLAND WEATHER SERVICE 175 the processes of plant life. It is of great importance to know the exact seasonal distribution of rainfall in order to determine to what extent it TABLE XL.-NUMBER OF DAYS WITH PRECIPITATION OF OF AN INCH OR MORE. ONE HUNDREDTH Year. a >-> / / A n Fig. 50. — Monthly Frequency of Precipitation. The maximum monthly frequency, the mean frequency and the least monthly frequency of occurrence of days with an appreciable amount of precipitation are shown, respec- tively, by the upper, the middle and the lower curves. Daily Eaixfall Frequency. A matter of considerable importance to agricultural and commercial interests is the frequency of the occurrence of rain or snow in appreciable quantities. This has been determined with great accuracy for the vicinity of Baltimore, especially since 1871, when systematic observations were begun bv the Weather Bureau. Here asfain as in the determination MARYLAND WEATHER SERVICE 183 of the average daily quantity of precipitation, the average values for adjacent days vary remarkably. Thus the average frequency for Janu- ary 2 is 13, while that of January 3 is but 4. For June 24 and 25, the values are 6 and 16 respectively. The day upon which rain or snow fell the greatest number of times from 1871 to 1901 is December 26, namely, 20 times in 31 years. The lower limit, namely, 4 times, belongs to January 3, September 1 and 2 and October 10. (See Table XLIV and Plate IX.) The table throws an interesting side-light on the mooted question of the occurrence of " equinoctial storms." Are rains any more frequent on March 21 and September 21 than on the days immediately preceding and following? On March 21 rain fell 15 times in 31 years; on March 19 and 20, 16 times; on 22 and 23, 12 and 13 times respectively. The average for all days in March is 13 times. On September 21 rain fell 8 times in 31 years; the average for all days of the month is 9.4 times. That is, an appreciable amount of rain fell on only 26 per cent of the September equinoctial days of the 31 years, or 4 per cent below the average for all days of September. These figures show that rain is not as likely to occur on these days as on many other days of the month. The Probability of Eain. If we divide the actual number of occurrences of rain on any given day by the number expressing the total number of years under considera- tion (in this case 31), we obtain an expression which in a rough way represents the percentage of expectancy of rainfall, or the rainfall prob- ability, for that day. Rainfall in the middle latitudes is too erratic in its occurrence to place much reliance upon this percentage as a forecast for any particular day ; if, however, we have a long series of observations and take the average value for 5 successive days, we arrive at a figure which more accurately represents the most probable percentage of occur- rences of rainfall for any one of the five days. This has been done in Table XLV and the results graphically represented in Plate IX. The curve based on 5-day means shows some periods of the year to be de- cidedly freer from rain tlian others, although there is a fairly uniform 184 THE CLIMATE OF BALTIMORE distribution of precipitation throughout the A'ear. The period from the middle of September to the middle of October, for example, has shown in ol years from ISTl to 1901, an average rainfall frequency of about 28 per cent. The month of March, on the other hand, shows a record of about 42 per cent. The last week of July shows a probability of 52 per cent, the highest for any week in the year. The average daily proba- TABLE XLV.— RAINFALL PROBABILITY BY PENTADS AND DECADES. (In perceiitag-e of possible frequency.) Pentads. (Pentads ending on stated dates.) January. February. March. ApriL May. June. 5tli 37.4 4tli 41.8 1st 43.7 5th 32.0 5th 35.4 4th 40.3 10 38.6 9 43.8 6 40.6 10 39.8 10 41.0 9 32.6 15 40.6 14 44.6 11 41.8 15 36.0 15 38.0 14 36.6 20 43.8 19 38.0 16 45.2 30 39.8 20 37.8 19 36.0 25 86.8 24 32.2 21 43.4 25 33.6 25 45.8 24 26.0 30 35.4 26 40.8 31 40.0 30 45.8 30 36.0 29 39.3 July. August. September. October. November. December. 4tli 29.4 3rd 40.8 2nd 29.6 2nd 28.2 1st 40.8 1st 31.6 9 36.0 8 30.6 7 26.8 7 29.0 6 37.6 6 31.4 14 34.0 13 37.2 12 35.6 12 25.6 11 40.6 11 29.6 19 31.4 18 38.6 17 37.4 17 25.8 16 27.0 16 33.2 24 34.2 23 37.3 23 38.4 22 29.0 21 35.4 21 33.6 29 52.2 28 29.8 37 38.8 27 36.0 26 44.4 26 44.3 31 40.6 Decades. Jan. 38.0 Feb. March April May June 1 July lAug . Sept. Oct i 1 . Nov. Dec. 1st Decade — 42.2 40.2 35.9 38.2 36.4 32.3 36.4 26.0 25.1 33.2 30.2 2nd 42.2 41.3 i 43.7 37.9 37.9 34.0 34.4 37.3 35.5 26. C 31.8 28.2 3rd 36.9 39.1 41.1 37.4 41.0 33.3 42.5 35.9 29.8 39.6 36.5 42.1 Table XLV. The figures in this table represent approximately the proba- bility of rain or snow upon any one of each of the stated five- and ten-day periods throughout the year, expressed In terms of percentage of the total number of similar days in thirty-one years, or of the possible frequency. For example, the probability of the occurrence of rain upon the 15th of March (or any stated day from the 11th to the 20th) is expressed by 43.7%; for the 15th of October, by 26.0%. If rain had occurred upon every 15th of March and 15th of October in the thirty-one years from 1871 to 1901 the percentages of probability of rain upon these days would have been represented by 100. bility for the entire year is 36 per cent; the highest for any one day is 64 per cent, namely, for December 26, and the lowest is 13 per cent, for January 3, September 1 and October 10. The probability of rain is less VOLUME 2, PLATE I b PRECinTATinS I'DOBADILITY. A. Precipitation ti-onnoncy for each day of the year. fEipressp,] , B. rrecipitati.m freqiieiicy for each successive 5-day pcrloa oi i ™ as u^rcenlages of the possible freqiie C. Average amount of precipitation for each day of tl U. Average amount of precipitation for each successlv MARYLAXD WEATHER SERVICE 185 than 20 per cent on but few clays of the year, and does not often exceed 50 per cent. The probability of precipitation at Baltimore on the following days may be of special interest : Jan. 1, 48 per cent. Feb. 22, 35 Mch. 4, 42 Mch. 21, 48 Apr. 30, 16 May 1. 29 May 30, 35 per cent. July 4, 35 " " Sept. 1, 13 " Sept. 12, 42 " Dec. 25, 35 " Thanksgiving Day 35 per cent. S O N D J - / y ) f\ ^ \ / ^ V. J /■ ^ ■\ ^ \ - ■- -- N ^ y Fio. 51. — The Monthly Amount of Precipitation. The upper line indicates the variations in the maximum monthly rainfall from month to month ; the middle line shows the mean monthly rainfall based on 30 years of observations ; the lower line shows the least monthly precipitation recorded during each month in 30 years. The Monthly Precipitation. The usual method of representing the precipitation of any given locality is by means of monthly and annual amounts. - Owing to the great variability in the character of rainfall and snowfall, a great many years of continuous observations made under practically imchanged exposure of the gauge are required. The vicinity of Baltimore has to its 186 THE CLIMATE OF BALTIMORE TABLE XLVI.-TOTAL MONTHLY AND ANNUAL PRECIPITATION FOR 87 YEARS. Year. 1817.. 1818. 1819. . 1820.. 1821. 1822. 1823. 1824. 1825. 1826. 1827. 1828. 1829. 1830. 1831. 1832. 1833. 1834. 1835. 1836. 183". 1838. 18.39. 1840. 1841. 1842. 1843. 1844. 1845. 1846... 1847... 1848... 1849... 1850... 18.51... 18,52. . . 1853. . . 1854... 1855... 18.56.., 1857... 1858... 18.59... 1860. . . 1861... 1862.., 1863... 1864.., 1865.. 1866... 1867.. 1868.., 1869. . 1870. . ".25 0.90 0.70 2.80 3.30 1.80 5.60 2.30 0.62 0.81 i.03 1.1,0 5.50 i.n U.38 3.21, '2.81 1.11 1.96 3.94 3.10 2.10 3.. 50 2.30 6.10 1.80 1.60 3.65 3.40 2.83 3.92 1.58 1.03 3.58 1.70 3.60 1.30 4.40 2.50 2.11 3.50 1.83 7.06 2.29 3.70 0.60 3.33 0.80 1.19 2.50 0.90 2.56 3.4" 2!o6 2.80 2.00 1.90 2.20 5.40 4.80 0.70 5.90 2.87 1.85 3.13 2.1,1 4.40 1.79 2.1S 2.33 1.05 1.93 1.57 3.41 3.10 2.90 3.60 3.30 1.40 3.35 2.20 1.45 3.59 1.83 3.43 0.94 1.15 2.43 2.90 3.60 3.40 4.90 4.00 0.50 0.66 1.61 5.74 2./,2 1.79 i.n A. 15 0.14 1.21 4.90 3.80 2.21 2.85 1..50 •S • - 4.. 50 3.00 4.55 3.30 1.70 1.30 7.10 4.30 /,.53 5.70 1.13 3.25 9.10 I A. 02 j 2.87 I 1.80 2.12 I 1.92 3.73 1.64 6.30 4..=)0 4.00 3.70 5.95 2.40 3.80 3.00 1.70 3.54 2.38 3.70 3.63 5.90 5.70 3.90 2.70 4.70 2.47 2.30 1.31 6.36 1.32 S.tS 5.79 4.23 3.62 1.40 3.90 3.20 3.64 1.90 1.50 2.10 3.70 1.10 2.10 3.10 1.80 4.70 0.67 I S.il 2.U7 I 3.37 4.40 1.57 I U.61 2.61 0.56 2.1,7 3.82 4.23 2.10 2.80 9.10 4.30 4..'i0 4.. 30 2.90 1.60 1.49 2.38 0.41 0.81 0.87 3.85 4.70 7.80 3.10 7.20 0.39 1.48 1.84 4.33 6.96 5.13 5.67 6.25 3.. 54 2.83 2.50 1.50 1.65 0.50 3.03 2.60 9.00 6.45 1.15 4.10 1.30 4.40 4.60 5.10 1..50 2.10 2.95 1.59 1.80 1..50 1.60 5.03 3.08 0.21 3.86 2.29 ! 1.80 3.18 2.28 3.40 1 8.50 3.i2 I U.92 1.01 U.90 5.33 3.21 1.83 4.10 4.20 4.30 4.50 3.90 i 2.75 ! 4.00 3.55 4,00 2.36 5.77 1.19 2.96 4.18 3.08 4.60 1.70 4 .30 5.20 0.91 1.19 3.23 9.08 2.74 3.i8 4.76 2.12 .',.10 3.39 4.92 0.20 0.63 3.80 1.38 3.53 1.37 A.S6 3.32 5.15 9.20 4.90 4.70 4.10 5.10 4.35 2.65 0.90 1.70 2.93 1.78 3.36 4.34 \.IQ 1.66 1.20 2.70 0.60 4.80 0.92 7.45 4.90 1.16 2.1,1, 2.38 5.71 3.53 0.82 3.93 2.. 50 3.00 .1.03 1.76 3.37 3.50 10.00 4.10 3.20 2.20 7.50 4.35 3.60 3.37 1.7i 2.50 /,.08 4.64 3.55 S.6U 2.2U 3.62 3.80 5.78 2.35 4.30 1.90 5.60 1.85 1.35 3.70 5.40 3.90 1.26 2.. 51 4.42 2.06 3.10 4.20 5.70 3.30 2.60 2.62 1.82 2.47 3.23 6.20 0.77 7.06 2.11 5.29 0.41 1.74 2.15 2.03 5.05 0.30 0.35 3.30 2.00 3.20 4.30 , 3.00 8.00 1.50 0.30 lO.'iO 0.80 I 2.25 4.10 5.80 4.50 2.94 3.21 , 2.U7 2.1,0 A. 95 1.35 3.98 3.35 k.6U /,.90 2.9/, 0.59 1.81 6.70 5.10 9.10 2.20 3.35 1.92 0.83 i.28 1.93 2.16 U.92 1.38 3.56 3.33 2.1,9 3.15 3.80 4.50 1.90 3.80 4.00 1 2.30 4.40 ! 1.00 7.82 10.. 50 0..31 ; 4.47 1.51 7.20 3.97 3.34 2.55 4.70 3.30 4.60 4.70 3.00 2.50 4.88 4.43 3.37 3.76 7.25 3.31 0.85 1.30 2.36 1.90 1.92 3.. 52 1.33 0..50 1.68 3.88 j 5.. 55 ' 1.64 1.90 4.70 0.50 2.20 2.40 4.10 2.30 2.83 1.40 4.44 7.05 2.69 1.80 3.70 0.91 2.19 1.81 4.20 1.00 4.4S 3.15 1.76 1.80 3.10 0.70 7.80 3.40 3.. 50 3.80 1.77 0.88 U.5U U.60 0.99 1.72 3.32 2.60 1.92 2.51 0.85 4.00 3.10 ' 3.10 1.60 4.50 3.80 1.40 1.97 3.03 3.73 1.30 3.38 7.35 6.37 3.10 3.20 3.60 4.40 7.10 3.70 0.77 3.89 2.34 2.38 3./,9 3.96 3.69 1.8/, 1.33 3.68 1.55 2.60 0..50 5.08 3.00 3.70 2.00 1.10 2.70 5.60 5.10 3.10 2.27 l.~23 1.62 3.95 5.51 3.32 l,./,2 1.65 2.21 1.89 2.55 2.69 4.80 3.40 3.70 2.80 3.15 3.30 3.75 4.25 1.85 1.23 7.17 2.. 54 1.44 l.Ofi 4.30 5.60 7.90 3.. 50 7.:« 1.20 1.85 1.87 3.97 3.20 5.05 5.48 3.91 2.30 2.41 2.. 50 1.10 2./,9 3.. 50 1.86 0.28 3.60 2.60 2.20 1.90 3.30 1.20 6.25 2.25 3.. JO 1.16 2.95 0.25 1.37 /,.6S 1.09 U.69 5.12 2.11 2./,2 7.10 3.60 4.50 8.80 3.25 5.10 3.35 3.90 3.. 50 3.43 2.10 2.38 3.10 4.44 4.40 1..50 6.20 2.. 30 3.90 3.60 2.05 6.33 5.65 3.15 2.99 1.36 1.50 /,.1S 1.40 5.90 2.. 50 \.'m 3.90 1.04 48.55 32.60 28.75 43.. 50 50.20 29.20 44.. 55 42.28 36.35 30.68 33.69 33.01 53.26 38.97 37.40 34.27 41.38 29.51 34.10 .54.63 45.00 47.10 51.70 37.. 50 43.90 35.10 48.79 32.46 28.39 46.66 as. 01 34.42 30.63 44.80 38.10 .51.. 50 .36.00 .59.20 29.31 22.87 38.37 46.06 .55.64 37.54 43.. 55 35.48 43.97 33.03 33.32 37.48 33.90 33.63 27.34 22.43 :maryland weather service 187 TABLE XLVI CONT.— TOTAL MONTHLY AND ANNUAL PRECIPITATION FOR 87 YEARS. Year. 1871. 1872. 1873. 1874. 1875. 1876. 1877. 1878. 187P. 1880. 1881. 1882. 1883. 1884. 188.5. 1886.. 1887.. 1888.. 18S9. . 1890.. 1891 1893 1893 1894 189.5 189fi. 1897. 1898. 1S99. 1900. 1901.. 19ft?. . 190:i. . 1.55 0.88 4.27 2.22 sisi 1.67 3.80 4.. 51 2.. 59 2.26 4.84 5.38 3.16 4.81 3.07 4.48 2.. 57 3.35 4.22 1.80 4.89 6.42 1.78 1.46 4.67 2.62 2.05 2.99 3.. 50 2.11 2.45 3.05 3.81 Averages. l821-]H.30... 1831-1840... 1841-l8ro... 18.51-1,'W)... 1861-1870... 1.38 a.fts 1.46 3.06 4.74 3.02 3.18 1.41 2.91 4. 72 2.96 1.87 3.31 1 . 55 1.96 5.68 3.73 4.69 6.69 4.40 5.49 4.69 2.. '3 2^53 4.80 5.. 52 2.41 4.43 3.. '3 0.8:5 7.07 5.13 1.32 5.47 4.65 0.65 4.68 5.43 2.4B 3.33 2.81 2.43 2.75 2.18 2.98 2.97 2.00 2.67 6.37 3.60 4.74 1.65 4.82 7.59 3.43 3.68 6.37 1.60 4.85 3.49 4.62 5.71 4.07 7.94 7.20 1.38 1.19 2.94 4.70 2.40 2.. 58 4.93 3.17 3.58 3.41 4.40 4.21 3.16 3..';0 3.. 35 3.40 S -. 1.90 3.06 2.77 6.65 4.27 1.90 3.30 4.19 3.69 3.07 2.00 2.14 3.20 2.65 1.37 2.03 1.44 6.31 1.92 1.49 4.94 2.2.3 5!38 2.74 1.23 2. .30 3.42 1.^2 3!i7 4.50 7.07 2.06 2.44 2.. 57 2.11 4.22 8.70 6.82 3.94 5.98 2.48 3.15 3.. 52 3.80 7.42 1.44 3.19 1.84 1.89 2.06 5.53 2.90 3.29 2.66 3.66 2.. 31 4.11 3.07 3.11 6.35 3.78 7.26 3.04 1.61 6.88 3.86 3.29 1.00 3.67 1.62 3.33 2.. 57 3.73 3.. 38 3.64 3.28 2.82 4.16 0.94 1.11 2.85 4.09 3.53 4.09 3.92 5.48 7.81 2.30 8.08 2.. 51 6.31 5.64 4.44 3.22 6.17 2.42 5.45 4.87 2.26 3.29 2.83 3.94 2.57 1.06 2.16 4.;34 0.90 4.30 5.01 3.44 4.. 52 2.51 2.90 2.96 1817-1870 1 2.52 2.68 3..55 I 3.03 [ 3.40 3.32 1871-1880 2.63 2.53 3.64 3.48 2.97 3.30 1881-189 1884 1874 1870 11.29 7.69 7.62 5.50 3.05 1.H2.H 8 55 21 .07 1S64 Helow Year.. KanKe normal 15.70 1870 36.77 190 THE CLIMATE OF BALTIMORE The extreme monthly ranges are observed, from the above table, to be more than double the average monthly amounts, while the extreme annual range closely approaches the mean annual precipitation. Another inter- esting fact brought out by the above table is the ratio existing between excessive and deficient monthly precipitation. In every instance, except- ing the month of February, the excessive amounts are in round numbers about double the deficiencies. As, in the long run, the sum of the ex- cessive amounts must equal those of the deficient amounts in order to produce the normal value, it follows that a precipitation below the normal is the most frequent, and hence the most probable. It is the excessive rainfall which disturbs average values to the greatest extent. A single heavy rainfall will occasionally materially change the average value for a long series of years. A case in point may be cited. In 1897 the average precipitation for the month of July for the southern part of Anne Arundel County, Maryland, based on a series of observa- tions covering 7 years was 5.67 inches. On the 26th of July, 1897, dur- ing a local thunderstorm of great intensity, a rainfall of nearly 15 inches was recorded. This single fall changed the average July rainfall from 5.67 inches to 7.45 inches. The Seasonal and Annual Precipitation. The great variability in the total annual precipitation is best seen by an inspection of Fig. 53, in which the 3'early amounts are presented graphically from 1817 to 1903 after reducing all observations to the Weather Bureau series. The dotted horizontal line represents the average height for the entire 87 years. The fluctuations from year to year are so irregular and vary so greatly in amount that it is difficult to detect any periodic movement. There are suggestions here and there in the diagram which point to possible periodic swings. There are groups of years during which the precipitation remained constantly above the normal value, and others with a persistent deficiency. Take, for instance, the period from 1850 to 1861 (see Fig. 53), when there was a decided annual excess with the exception of two or three years in the middle of the period ; this was followed by 13 years of deficient pre- ^! — =:^ s— ^— I ^^^^^^^^^^^^^ ^^^^^^^^^^^^^^ — I — ^^^^^^^^^^^^ ^^t^a^m^^mamm ZZ Z^^ Z!ZZ ZZZSZZ ? J — ^^= ^— — — ^ ^^^" ^^^" ^^^" = ^ ?; c: ti > 2 .2 c S^ *< ^ i i •■« •- fc ? 5 o ?* r 192 THE CLIMATE OF BALTIMORE cipitation; from 1873 to 1889 the anuual values fluctuated a great deal, but on the whole there was a gradually increasing excess, culminating in 1889 in one of the heaviest annual falls recorded ; in the following years there was diminishing rainfall w ith an average Ijelow the normal to the present time. TABLE XLVII.-TOTAL SEASONAL PRECIPITATION FOR 87 YEARS. u C 1 h c t; a a bo c s a B O u bo C e ^ 3 ^ 5 be S S 1 p. a 3 3 < ; u a. 3 3 © CO h M 3 IK << ! 1817 1817-8 1818-9 1819-20 1820-1 1821-3 1822-3 1823-4 1824-5 1825-6 1826-7 1837-8 1828-9 1829-30 1830-1 1831-2 1832-3 1833-4 1834-5 6.50 11.55 5.20 11.35 7.20 ! 8.80 132.58 ' 7.25 7.80 . ,. 114.80 112.00 8.80 8.30 1845-6 8.08 11.69 15.87 12.35 1846-7 i 8.44 '■ 3.98 i 8.84 111.47 1847-8 ' 4.90 ' 6.47 11.90 10.43 1848-9 5.27 8.68 6.11 9.23 1849-50 10.45 12.83 9.46 12.10 10.60 8.90 9.60 19.70 9.90 4.90 6.65 ; 9.85 7.50 11.00 9. .30 11.70 14.45 11.95 12.90 6.98 5.74 6.79 ' 8.03 4.58 6.03 ; 9.33 9.46 8.08 6.33 I 5.89 9.31 9.3S 6.83 ' 9.80 8. .31 10.78 10.15 16.90 17.13 6.97 4.33 9.01 11.83 10.50 jll.19 6.66 ! 8.55 I 8.83 I 5.64 8.49 11.36 10.05 9.31 8.51 i fi.19 8.01 10.92 13.37 7.60 I 7.71 I 8.39 9.38 12.74 ' 6.03 1835-6 9.77 , 9.97 18.35 11.95 1836-7 !l3.30 13.60 14.30 10.30 1837-8 7.60 11.60 15.70 10.30 1838-9 11.60 17.60 11.90 ' 6.30 1839-40 13.40 10.90 9.30 ' 9.45 1840-1 10.75 13.20 9.70 8.40 1841-2 110.25 10.70 10.75 5.15 1843-3 7.15 ,10.25 14.12 16.72 1843-4 I 9.00 8.60 5.91 9.35 1844-5 9.49 5.55 6.96 6.46 1850-1 1851-3 1852-3 18.53.4 1854.5 1S55-6 1856-7 1857-8 1858-9 1859-60 1860-1 1361-3 1863-3 1863-4 1864-5 186.5-6 1866-7 1867-8 1888-9 1869-70 1870-1 1871-2 18T2-3 1873-4 1874-5 9.00 15.00 8.70 8.30 7.70 13.40 13.00 13.70 10.90 10.10 8.60 10.30 11.60 17.10 10.40 18.50 10.40 4.10 7.91 7.30 6.31 6.21 9.77 18.45 7.86 5.14 7.63 5.45 7.37 14.35 6.16 14.73 11.50 10.75 15.96 11.13 13.61 8.15 10.46 11.23 8.48 13.71 12.75 6.07 ' 9.24 ! 8.67 8.98 16.14 10.12 5.13 11.16 i 3.59 3.80 10.36 7.57 10.34 11.33 5.05 5.93 6.99 13.30 4.16 7.30 13.09 7.24 7.71 6.93 5.53 7.40 i 7.45 6. .57 6.85 7.55 I 6.09 10.01 8.48 3.56 10.09 5.40 I 5.04 3.97 I 6.96 12.38 , 8.57 4.24 7.56 10.33 12.31 11.23 11.60 13..S3 13.96 6.37 9.98 8.88 7.47 7.32 10.48 16.30 9.92 1875-6 1876-7 1877-8 1878-9 1879-80 1880-1 1881-3 1883-3 1883-4 1884-5 1885-6 1886-7 1887-8 1888-9 1889-90 1890-1 1891-3 1893-3 1893-4 1894-5 189.5-6 1896-7 1897-8 1898-9 1899-00 1900-1 1901-2 1902-3 1817-70 1871-03 7.77 13.21 11.49 6.99 ! 9.13 8.77 10.05 14.31 13.57 9.75 8.08 13.79 9.45 9.13 16.39 15.41 11.89 11.. 36 15.01 i 8.99 11.42 9.55 8.10 13.90 14.48 13.19 13.68 11.38 I 7.47 16.76 13.46 13.98 17.66 10.38 8.50 16.91 11.23 10.95 13.21 10.01 21.23 18.60 7.21 13.99 13.47 13. OS 13.53 17.48 13.07 16.70 10.77 8.49 8.68 : 5.95 7.38 12.25 6.43 9.62 13.40 8.66 13.53 7.75 13.19 7.55 13.47 14.21 7.71 8.28 10.66 13.31 10.11 ; 8.66 8.16 ■ 6.33 ' 8.76 5.17 13.78 '13.81 14.80 7.93 11.06 14.90 11.03 18.54' 8.50 10.01 10.35 9.97 10.89 12.64 ; 16.05 17.34 8.78 4.77 7.28 9.45. 10.89 7.69 4.60 11.85 7.38 5.88 10.93 15.16 11.23 9.55 6.47 9.02 10.53 10.07 8.59 10.33 9.87 11.45 7.75 6.28 17.76 5.27 9.37 Table XLVII. For the character of the record in the above table see foot- note to table XLVI. The greatest annual precipitation of the entire period of 87 years, making due allowance for the permanent differences between the Fort Mc- Henry record and that of the U. S. Weather Bureau, was that of 1854, with a fall of 64.63. The mean annual amount for Baltimore based on the Weather Bureau observations for 33 years is 43.34 inches. The rainfall and snowfall of 1889 followed close behind that of 1854 with a MARYLAND WEATHER SERVICE 193 Jan. Julv Jan. julv Jan. Julv J Fio. 54a. — Departures from Moan Monthly Precipitation (1817-1859). 191 THE CLIMATE OF BALTIMORE Jan. July ■>*" Jul/ J'" J"i- JULv Jan Julv Jan. 4^44-186 '+tt: lU X-U ^- -1 tJ-H h-t- --u ' i m i XX =*b- — , . . r x ., 3tlt-.^ iit y'^C i^ .-.-. . . V \ ^ i^^ii it jk, 3tx it /X* > ^ J ^\^r7V3rAy ii ^ a.>l/A i-Hry , ' V'' ' ' 'yi y"s/ i/.\i ^¥ ^ ;^ v\.AW^/ny . . ^^/^^^^^ -^' 7 rr"T"- -T ! " j- I'n " ' : ■ _...- ^ . _^- ^ t--4-i' It M ' ' -] — tHi r _- _i:xT I 1 ' ■ ^ ' I I ' 1 " i^ A All Ml i\ K i\ Iv4 A V L /^, __ t-^. j^ V A *^ I \ L / ^J V j^ J> J' ^ ^\> n \ /\ ; \ V L_ / "1/ * i? ^f \/\ /^ w ^/^ ^^ ^^ V \/ \ 1^ t / ^ ' V Y/ ' • ' V * ^ \ ' ! i , 1 . . 1 ' 1 nil 11 1 1 T^ TV,, . _ _. _.. _. it X ^iK lit V ^-^ -^ ni t\t\ 4\ iS ^. .L^2_X/-S-/ r- -jt-L ll iL :"iS?r "2: — ::x,'^ l7*s ~^Lix3u ^' v^t - \ ^^^f' ." v^^^ ^*' \ i^\ ^t^t^"^ ^ 5 2 jH/ 4- \v -_t-^ V u^ *=/ y - rr .^ i ± X T X 1 1 -1 !._..._ -)- 1875 } X ■ _i_ XC T~T^ """X -XX i 1- ilt i i I ^ • A A — t-h + ^ ] -,c 3:^^,111 ^^ - j^C^ Wa a At ^^s-r r J* /n^?1/l it i4,>i^/V\ K. y>. f\ li \ " 4'f\"7 cc s}^ / [/ a\> S/ L ' yy \i \ A^' yz " " ^ ' / \r- ^5^j^ v^ -^F-i^--'«^k X^X M \ 1 X-tt A> ^ 1 ' ' -T- "T~ ~^ V Ml Y ■ tx it ijit X r ' ' X ; "htifti X t 1 ir 1 '8« 0" -1881 ^* - • 1882 j,^ + ISSsI: 1884 ■^5 1 1 I 1 A , L 1 1 1 /) . 1 i/^r~~^j'»k i/\t ' 11 7"^^ I x 3k; t ^-X. XT XTt? S IIII ^ it y it i h tl - n i;s-j' 'r /I V /I A/ \l j/ 1 IALL /I J J^ \i /I jTtJ ° 3 l£ L^' w.\ /v ^::c^ X J ct y'^7 "^.n 7 S^ SL " \^ i: A^ ^/- i"- ' " 7 T 1 1 l\ xt i X X - i't' ik''' 'll! l/1 i I'A' \/i lityil 1 'iji il/ fl / Y ^^ ^ ^ ' ' / V A V^ i \ ' >k ' / A i \/\v w 1 iSi>^'i'* ' a'a / ^^ ^ L J A ^^ li u ' ^^/ 1 1 i ^ V 1 ] ' ' ; 1 1 ■18£ 1 " 1 yl 1 lI 1 1 i [ ■ ""^ T jc -jO"" A.^1- A -/V i jI A Al 1 it /t -,^ A 'Wu J . I / VA / y V7 \ . . i 1 , 1 X ^.- X^\ . -,' i^'lT 1 *^x '/ ;v y V\ / " / V \ t-^ J ,-k -it ^t 4w 4- -r->k-?- r*^ L' ' \/ yy t\/ ^ : X 7 \/r7 1^ T> ^wC/ ^ ^t^* n — h "V " *^+\/ ^^+'* — ^^ M^ r V/^ s^ \ -/ -'^u w ^ ^ It it X X^ It ' ^^ " v' X it X - it it X - - X - 1 189 XT" ' it ^ ■ X -i^ Th - 31 <, V A I it r- It - i\ /l J > i> * A \ 1-L ii ^L, /5 tl - ' tJ f\ fi /l«lAN\i\/^t A /\i^-t ^^, - \4^^l t\>*' \ y \AA '^^ V A/i \ ^i x^ vi* yIi"* V/ ^« / *"' \_f V\'\y Y ' V V" \>^r ^tf ^ 3l ' V T ; V -^^ F- -5'^ X X X it X ■■■- i 1 1 1 ---i-H- -^- 1 1 1 X 19( ~*~ 1901 X 1902' ~^ 1903" 1904 'X' X it ,«!t X ' 1 "^ ± lA J \^\. ^\ ^.^ - ^S 3^ I ^ -U\/< -,^\ JS A i A ' /\ i\ / A k / T YlWl/ \ A J l\ j^ rV A A -; 1 ii lu/i^vAy \7\ ,?r l^^'bt S2^ rV/\j / >' t7 y 1-'' 1 Ypr p^ v^^ af V ' ' ! 1 1 1 1 1 1 Fig. 54b. — Departures from Mean Monthly Precipitation (1860-1904). 1875 1880 1885 1390 1895 1900 XRTCRKS FROM THE NnRMAL PRECIPITATION ( 1817-1904'). MARYLAND WEATHER SERVICE 195 total of 63.35 inches. The smallest annual precipitation recorded was 27.86 inches, reduced value, in 1870. These figures show a total range in the annual precipitation of 36.77 inches. The average departure of the annual precipitation from the normal quantity is 6.68 inches, a figure which attests the great variability of this climatic element. If we compute from Table XLVI average annual values for each 10-year period from 1820 to 1900, we find a fairly close agreement, show- ing a strong tendency to return to a certain normal value in spite of great fluctuations in individual years. The most conspicuous departure is the great deficiency recorded from 1861 to 1870. DEPARTURES OF TEN-YEAR AVERAGES FROM THE NORMAL FOR 87 YEARS. Decades. Departures. 1821-1830 -0.11 inches. 1831-1840 +3.13 1841-1850 -0.30 1851-1860 + 3.34 1861-1870 -6.02 1871-1880 -1.98 1881-1890 + 3.88 1891-1900 -2.85 Monthly and Annual Departures. The chief characteristics of the monthly and seasonal departures for successive years are clearly shown in Fig. 54 and Plate X. The most con- spicuous feature of these curves is the irregular, short-period fluctuation,? above and below the line representing the average amounts for a long series of years. The extreme irregularity in the fluctuations makes it entirely impracticable to employ this method as a basis for making long- range forecasts. When the precipitation is charted by months in regular chronological sequence, as in Fig. 54, the same characteristic fluctuations are noted. They are irregular in amount and period, but with a general tendency to return to the normal level, and with occasional evidence of a long period of excessive or deficient precipitation. An inspection of Fig. 54 and Plate X will show at a glance that the exact average precipitation for the months, seasons or the year is an unusual occurrence. As sliown in the discussion of temperatures, the arithmetical mean amount is not the most probable amount. The pre- 14 196 THE CLIMATE OF BALTIMORE cipitation actually recorded is, in most cases, well above or below the normal value. The following figures show the average departure above or below the mean value, based on observations for 87 years, and disre- garding the sign of the departures : AVERAGE DEPARTURES FROM THJ: MEAN PRECIPITATION. (In inches and hundredths.) Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year Plusorrainus 1.07 1.25 l.m 1.33 1.40 1.54 1.64 1.72 1.57 1.33 1.27 1.32 The most frequent, and hence the most probable departure, differs from the figures above representing the average departure, as shown by the following table : FREQUENCY OF PLUS AND MINUS DEPARTURES FROM THE NORMAL MONTHLY PRECIPITATION. (In percentage of total frequency.) Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. 46 54 8 Dec. 45 54 9 Year Plus departure . . . Minus departure.. 40 60 20 48 63 4 46 54 8 40 60 20 46 53 47 53 5 45 54 9 46 54 8 39 61 22 45 54 9 46 54 8 In all months of the year the departures below the normal are in excess of those above the normal ; in January, April and September as much as 20 per cent of the total number of months in 87 years. The average monthly difference is nearly 11 per cent. Hence the monthly precipi- tation is most likely to be below the normal amount, but the minus de- partures are likely to be smaller than the departures above the norma). The most probable departures from the normal monthly amounts are indicated by the following figures: FREQUENCY OF STATED MONTHLY DEPARTURES. Departures (In inches). 0—1 1—2 2—3 3 — 4 4—5 5—6 6—7 7—8 Plus. Minus. 12.3 ' 19.5 •• 10.1 " 1.4 " 0.0 " 0.0 " 0.0 " 0.0 " 5.1 ' 3.8 ' 1.8 ' 1.0 ' 0.4 ' 0.2 ' 44.6 per cent. 55.1 per cent. maryland weather service 197 Excessive Eains. The heaviest rainfall recorded in Maryland occurred at Jewell, in the southern portion of Anne Arundel County, on July 27, 1897, when the local voluntary observer measured 14.75 inches, all of which fell in the course of 18 hours, and most of it in six hours, during a severe thunder- storm. On September 13, 1904, a storm of marked intensity developed over the Atlantic Ocean east of the coast of the South Atlantic states. It increased rapidly in intensity during the 14th and with rapid movement passed northeastward on the 14th and 15th with its center following the coast line. The precipitation in the path of the storm was of unusual intensity. The center of the cyclonic system passed just to the east of Baltimore during the night of September 14-15 accompanied by de- structive winds and torrential rains. The rain began about 8 a. m. and continued with but slight interruptions until about 3 a. m. of the 15tli. The total fall during these 19 hours was 5.06 inches; between 2 a.m. and 4 a. m. of the 14th, .02 inch fell, making a total fall in 24 hours of 5.08 inches. This is the greatest fall in 24 consecutive hours recorded at Baltimore since the establishment of the local office of the TJ. S. Weather Bureau in 1871. The greatest rate of fall during this storm occurred between 8.25 p. m. and 9.10 p. m., Avhen l.iU inches were re- corded in 45 minutes. Some of the more interesting records of tor- rential rains or " cloud bursts " mentioned by Professor Henry in his report ^ on the rainfall of the United States are cited here : "A cloud burst passed over the edge of the little town of Palmetto, Nevada, in August, 1890. A rain gauge that was not exposed to the full intensity of the storm caught 8.80 inches of water in an hour. At Tri- delphia. West Virginia, 6.90 inches fell in 55 minutes on July 19, 1888. At Campo, California, 11.50 inches fell within an hour, " and some of the fall was lost!" 'Henry, A. J. Rainfall of the United States. Bull. D.. U. S. Weather Bureau, 4to. Wa.sh., 1S97. 198 THE CLIMATE OF BALTIMORE a»Ba Clr-li-(i-l i-rOl i-l r-t Ci i-IC^ i-l gt qjaoK l^mv aot-«DiOO -^»QiOCCi-l i-O0CD»i^l™ OiOOeOi-ll— O530000t- ■ONtOl-'^ ei N -* 05 »i CO M ?i »i P5 co ci oj » -* M M »-< OS cj o i i CO O j « r-l C O g-1 33 « 3 a io»ccococo 00 1- to £J -* M I- U3 O O CO g^ f 1 -* cc O X CO 05 t- -* f— ( ct *-< -^ o c ei 1-1 03 = 0© _fcj-5 < X -/. 00 O I- O "-I ■>* r-c ■>* cr. » «3C0 toiiffj ■f f 3 O 31 »1 0> to "O CO « rl 51 ?5 CI X, 03 t~ OJ a^BQ ■},niV *.niv CJ ?■• t- i-t -# ci 55 CO -* 00 000 C:00^-* OlrtC". ICCO 0} CO 1-1 1— I ._ 01 T? 1-1 irl_ Oj-i o5 >H-H -^00 r" -^ •:^o:ct-;s f-l" SSiCOl I— 1-# c; C-. -iO OJi-lOl I i I •»«0O3 ■-l^Or-l -H Qi-I O 0--L OOiC OS OOr-l CCrHi-l 000 ... . o> OJi-i-H M-H i^CO i^ o-^ ei IM rH 10 c-l C5 CO »-^ tc 1-1 1-1 r-Ci-l05 c;coi--»~-a< coeotocoi-i Ni-fxicooi omoocoo 1— I 0> 07 ^H OJl-lClC> l-l T-!"■■•■ g>oi « 01 O N CO-*'* CO tcoscci-tco oso-^tco coi-it--^Tji oi30f-i«co lOoiotDo C2»ao:5»o ooii-ioi-< 000^^ oj 30 ii « CO ^Oi-<>^0 Oi-li-IOi-l t-l^i-lOO OMCO =C^ oi rH O oi 1-1 a^BQ eoTjiojot- co-*iMi-io C-if^i-IC'l Ot OiOICO ss 1—1 CO o 05 to o to !0 CO J 01 OlOiOlOIOJ OJ 1—1 CO ^^1— IC co;5>-j w^ <-i r-{ r^ -j:, co-*X) ».niv e ?* 01 ic -* C0N-*r-(O i-i M CO O gt »-li*iOCOiO M*t-t-K5oo i-i oi oi o o 01 :o o :c Oi 00 1— X/ 01 o OOOi-icO SS2^'* ^'^^izn tccsco o^co O 01 33 X 1-1 ooo:otD X Or-ii-if-io O l- rl Ml t- ... .00 1-1 01 1-1 CO 1-1 ajBo: OJi-Irl to to CO "OC 1-100 CO 1-H kO r-t 1-1 i-l tot-cot-o» ^^rte^cj-* to 01 00 xscocoxtc liOiv 1— i-^OMOt.. O tc Ol ■-( 00 to 01 01 1-1 I III »o CO 01 «0 10 ^OCsgsi— I C5XCOXCO -^COiOtCOi O-i^iOtOCO tOOlOOr^ OOtf^ tD»« 0505'*co-H "Ooicooo i-xi-ixo ooitoost- ^-^=^^3 Socio t-3» :i-lO^^^^ 1—1 01 ^^00 OO^^Oi-t i^^HO^^-.^ ■^-f 1— ( Oi to 1-i ^ -^ oi CO to OOKO t-o» . . .00 50 •■»^ ajBCr 5i— (to t-»0 •-^ t: S S ** I- 01 wOrl 1-t ^ 01 X CS -H O t-1-1 i-iiCNOl I I I I 005 -# too _ej 01 01^ CO ^ 01 o; tj.tO« tOlO t' l^rav nl2S52SS2 Si2SS2J^ S'^SSSJS ;r55'^Si5'° co-*oioob uoot-toi^^i xto x ■*ococo-a< -^-^looco Oi-;oi3DCO e-it-o-d'os -^xo^ti o3x-*oo3 ojSx OOi-ioii-< 1-I011-IOCO e^i t- o -d< OS oi 'H oi o 1-i ■*xo- ^Oi-<00 aiBQ IM^Ol-r-^ COCtOllMOl ^^Ol-^i-IOl ^nOl CO Ol ^- -H-HOi-liO Ol 01 O! -^ CO 01 kO CO g» >c — o too 0:0-* XCOXOl-4" 01 1— X to o 01 01 01 01 01 01 01 01 01 ^^ ^lOiv eSmSto° SlSSSr: SSSiS^ =£:S2i^S Hi-050i-l o4aji-<^i-i iOXi-CT*il oioi^o^ Or-tCO to 'f*! oi oi 1-i o o oi-o> a^BQ XM0O5 r-(Cjiej-* t-4i ci wo t-. .-( 1-1 OS Ol-^ j-iol-5 61 1-1 ice^ 00 OS tc cox OlOll-rtOI *itnv t- CO to to 00 o oi o o o to »o XCO1— i:otD to I— I to cs t 01 01 1 1—1 1-1 01 01 01 _^ _ ^ ^^ ^ «> .H OlC5-n 01 1-1 1-1 1-1 J-H- -- ' ■" *^■» 0000 coto Oli-lOSOlr-1 ooo'iio 01 1— 1 -H Ol CO 01 -H 01 01 1—1 01 -H 1-1 101 -H i-llCOl 1-1 01 tOOS@C>l— CO*Oiio COOOIOCS i-^Qi-io-i ci^oioo COQi-i^O 1-^ 01 to t- 00 iH-Oll-l cooco :^ CO— J»-li-H05 XC2-HOX 00 to 00 to O 01 -H to ' - 00 O -191 O 01 O^rv^ rt otct-:o«5 tct-cocooo tocoo^.* Sosoioco SSSSS! S2S c>to t- :o«3 OOi-i-i-4 i-ii-i-ioo t- CO CO 00 jrtO-H-Hr tOCOOi-l-* oirt-H-ij-i •aosoioco -tO-H-* 31^ t- 001 i-i 00 to i*OS •00 CO 1-1 CO 1-1 i-fosto 10 OS iO to t- O5t6ooo^- oscocoooo •«ji-*i-co'0 ^hou^osuo i-lr-IOl 1—11-1 0101i-li-li-(0101 22E;S?S S^J^Str rSSS;^:* 2S^o» M-ortoo H*l.^*^0 i-<00-Hi-i OS o cs 01 -* i-*0S05C0 l-ii-HQO-H ■*-HO 01 N 01 1-1 1- 1-1 coo-aito—i OlNi-l 1-1 ocsocso i-io^'oi-i i-INCO 01 rt-HOi CO 1-101 01 Iroos in -J' cotoo oscs -HdCOiinO tOt-XCSO 1— lOlCO^iO tCI^XOSO -H01CO**1»0 ■t-Lr'^ir trirtr'-oo xxbooooo xxooxS ixcscedsSs §^~c:o S^3 roxxoooo ooxxoox xxxxx oo^xoox SxxSS xooxxS §SS MARYLAND WEATHER SERVICE 199 In his " Climates and Weather of India/" * Mr. H. F. Blanford cites the following, among others, as the heaviest rainfalls on record in that locality : TORRENTIAL RAINFALLS IN INDIA. Place Amount in ,,.. ^'''^®- 34 hours. ^'™®- Cherra Poongee f Assam i 40.8 June 14, 1876. Purneah (Bengal) 35.0 Sept. 13, 1879. Nagina 32.4 Sept. 18, 1880. Danipur 30.4 Sept. 18, 1880. Rewah (Central India) 30.4 June 16, 1882. In connection with the above, Mr. Blanford writes : " These exces- sive falls are always the result of cyclonic storms. Not such as are of destructive violence, but the long-lived cyclonic storms in which the barometer is not greatly depressed, and only recognizable in their true character when the barometer readings and the winds are laid down on the charts Another noteworthy point is that they have frequently occurred in years of partial drought Thus in 1875, although in the Punjab it was one of the wettest years on record, the rainfall was very deficient in Bengal and all over the southern half of the Peninsula. . . . Scarcely less remarkable is the occasional heaviness of the falls even at places where the average rainfall is not by any means excessive. That upwards of 40 inches in 24 hours should have been recorded at Cherra Poongee will perhaps hardly be surprising, but falls nearly as great, from 30 to 35 inches, in the same interval have occurred on more than one occasion on the plains of the Ganges Valley, at places where the average for the whole year is not more than from 40 to 65 inches; and even in the extremely arid province of Sind as much as 20 inches fell in one day in 1866, at Doprbaji, where the annual average is probably less than six inches." Greatest Kainfall in 24 Hours. In Table XLVIII the greatest precipitation in any 24 consecutive hours of each month and year from 1871 to 1903 is recorded, together with the amount of the fall and the date of occurrence. The greatest - Blanford, H. F. Tlip Climates and Weather of India, 12mo, London, 1889, pp. 77 et seq. 200 THE CLIMATE OF BALTIMORE during the entire period for each month is also graphically shown in Fig. 55. Falls equalling or exceeding two inches in 24 hours have occurred in all months of the year, but the heaviest have been recorded in the summer and early fall months. The greatest fall recorded in the table, namely, 4.76 inches, was that of September 6, 1895. During the present year (1904), however, this record was broken by the excessive rainfall in connection with the severe coast storm of September 14-15, when 5.08 inches fell in 17 hours. An inspection of Table XLVIII shows that the heaviest precipitation of the month may be very small, sometimes falling below half an inch, but such instances are comparatively infrequent, especially during the months of active plant growth. The following list comprises all months without a fall of half an inch or more in 24 hours during the 33 years from 1871 to 1903 : MONTHS WITH A MAXIMUM RAINFALL OF LESS THAN HALF AN INCH IN 24 HOURS. January 1S71, 1872, 1890. February 1805. March None. April 1881, 1898. May 1872, 1896. June 1901. July 1894, 1900. August 1876, 1877, 1889, 1894. September 1878, 1884. October 1874, 1882, 1884, 1892. November 1882, 1890, 1903. December 1871, 1873, 1874, 1875, 187G, 1889, 1896. Fig. 55 shows the extent to wliich the heaviest precipitation in 24 consecutive hours occurring in each year from 1871 to 1904 has varied from year to year. The amounts range from a minimum of 1.47 inches in August, 1898, to a maximum of 5.08 inches in September, 1904. The tendency to a periodic fluctuation embracing a group of years is more marked in this diagram than in those representing the total seasonal or annual fall. Especially interesting and instructive, as well as striking, is the gradual and steady increase in the intensity of maximum rainfalls from the smallest of the entire period of 34 years in 1898 to the greatest of the period in 1904. It would seem to be a safe inference from these MARYLAND WEATHER SERVICE 201 facts that we have arrived at a maximum for this particular periodic swing, and that during the following two or three years there will be a diminishing intensity of precipitation in individual storms. This view finds additional confirmation in the grouping of excessive rainfalls of 2.50 inches and over as sho\^ai in Fig. 56. There seems to be no fixed relation between the annual amount of rainfall and individual rains in any given locality. Eegions with a high annual or seasonal precipitation do not necessarily have excessive rates of fall for shorter periods. While some of the phenomenal rains have occurred in the tropics, where the seasonal rainfall is generally greatest, there are many instances where record-breaking downpours have occurred in comparatively dry regions. H Fig. 55. — The Heaviest Precipitation in any 24 Consecutive Hours. (E.ypressed in inches anfl fractions of an inch.) In Table XLI will be found a record of the annual number of occasions on which the rainfall of a 24-hour period equalled certain stated amounts under and exceeding one inch. The precipitation records of the past 3-i years have been further examined for all days upon which the rainfall exceeded 2.50 inches (see Table XLIX). Eainfalls of the latter amount may be considered excessive for all but a few limited regions. They are not of frequent occurrence in the vicinity of Baltimore; since 1871 there have been but 42 all told, most of which occurred in the montiis from June to September. Their total monthly frequency in 34 years is shown by the following figures : an. Kcl.. Mur. Apr. May •luni" July Aujf. Sei)t. Oct. \ov. Dec. ^'(•al■ o ■> 1 1 a 9 3 9 li 1 :! 42 202 THE CLIMATE OF BALTIMORE In Fig. 56 their frequency and intensity are also graphically shown by months and years. The manner in which these excessive rainfalls are grouped is interesting. There are apparently three groups in the entire period of 34 years, of which the j^ears 1876, 1887 and 1901 are the JA'. 18 ■ I 1 7t 18 1 1 1 1 80 18 1 1 1 1 85 IJ 1 1 1 1 90 IS 1 1 1 1 95 19 1 1 1 1 00 1 1 1 1 Feb. 1 1 MCH 1 1 Apn 1 May 1 June 1 1 ll Juiv II 1 1 II 1 1 AJG 1 1 Sep- 1 1 1 ll 1 1 1 Oct. 1 1 II 1 Nov. 1 Dec. -I 1 nil 1 1 1 1 nil 1 1 1 1 1 i 1 1 ilM Fig. 56. — Rainfalls Equalling or Exceeding 2.50 Inches in a Day. The frequency and seasonal distribution of rainfalls of 2.50 Inches in 24 consecutive hours are shown in this diagram. The total amount of the fall is roughly indicated by the length of the heavy vertical lines, the shortest representing 2.50 inches. central years. These years coincide with considerable exactness with the minimum sunspot period of approximately eleven years. Further atten- tion will be given in later pages to the relation existing between rainfall and this well-known period of solar activity. MARYLAND WEATHER SERVICE 203 TABLE XLIX. — DATES UPON WHICH PRECIPITATION EQUALLED OR EXCEEDED 2.50 INCHES IN 24 HOURS. .Januarj-. February. March. April. May. June. u o 1 1 s a C 03 i^ 1 » f 1 © S ai < n Year Am't Date r- 1880 1883 1885 1SS6 1900 a < 2.66 2.66 4.47 3.18 2.62 1 1886 1896 2.60 3.48 10-11 5- 6 1881 1889 3.. 51 8-9 2.71 3-4 1889 .... 3.58 25-26 1886 2.99 7-8 11 26-27 28 22-33 16-17 July. August. September. October. November. December. 1875 1876 1880 1884 1887 1889 1891 1903 2.70 1.5-16 3.14 30 3.71 20 3.75 11 2.77 20-21 3.63 1- 2 4.02 30-31 2.. 59 8 3.99 13-13 1873 1885 1901 4.36 3.35 3.28 13-14 2- 3 6- 7 1874 1876 1891 1895 1899 1900 1902 1904 3.15 3.94 4.00 4.76 2. CO 2.90 3.61 3.82 5.08 15-16 16-17 5- 6 6 19-20 25-26 1.5-16 25-26 14-15 1873 187.! 1877 187S 1890 1902 3.42 19-20 2.64 27-28 2.74 4 2.75 22-2-3 3!04 "" 23 2.79 4- 5 1877 .... 2.85 23-24 18T8 1888 1901 2.85 2.. 56 2.88 .... 10 16-17 28-29 Table XLIX shows all periods of 24 consecutive hours during which rain fell to the depth of 2.50 inches or over, from 1871 to 1904. The day and year of occurrence are likewise shown, and the total amount which fell within the 24 hour period. 1 I- 1 1 1 1 MM MM 1 MM J ...,. , , I 1 1 r 1 1 1 II 1 1 1 II 1 1 1 III III ml 1 1 1 1 -I 1 Mil MM MM MM ® .^ .4^ « 9 ^ +j » 4) u -w e « e 6 eS a E a « OS a fc3 £ s d ^ < H o >* < H Q V* < H r- >• < H (5. 1889 1.20 1-0 20 1881 1.00 0-45 £0 1877 1.28 0-55 24 1873 1.30 1- 10 1903 1.49 0-37 24 1.16 1-10 £0 1884 1.40 1-15 31 1875 1.41 1-15 la 1888 1.10 0-35 33 1895 1.05 1- •1 1887 1.74 1-10 22 li-91 1.15 1- 4 1896 l.EO 1- 21 1888 1.03 1- 8 1893 1.S3 1- 27 1897 1.36 0-23 17 " 1.13 1-0 5 1901 1.77 0-25 35 1890 1.96 1-10 21 1903 2.87 0-33 12 1898 1.43 0-25 1 *' 1.00 0-24 September October. :::: 1899 1901 1.59 1.00 l.Ofi 1- U 0-44 0-35 36 6 1896 ' 1.00 0-40 19 1897 1.38 1- 13 13 1899 1.00 0-53 2.i 1903 1.40 0-39 5-6 1900 i 1.62 1 1- 1.1 " 1.23 0-25 37 1904 1 1.48 i 0-40 14 ' Table L is a list of all occurrences of rainfall equalling or exceeding one inch in one hour, from 1871 to 1904. The year, month and day of occurrence are shown, and also the amount recorded and tlie duration of the excessive rate of rainfall; the latter in the column marked "Time," expressed in hours and minutes. March, April, jSTovember or December. The distribution through the season is as follows : May June July Aug. Stpt. Oct. Year 5 8 13 4 1 as The monthly distribution here indicated associates this class of exces- sive rainfalls at once with the thunderstorm. Their frequency and intensity, arranged by months and years, are also graphically shown in Fig. 57. The grouping referred to above in the discussion of the rain- falls of 2.50 inches and over is here also evident, though less clearly. IMARYLAND WEATHER SERVICE 205 Excessive Eates of Precipitatiox. The rate of rainfall, or the quantity which falls per hour, or part of an hour, in the case of excessive precipitation, is one of great importance in large centers of population, as it involves the engineering problem of providing adequate means for carrying oi¥ the surplus water without damage to proj^erty or interruption to traffic. Especially is it desirable in this connection to know the maximum rate of fall. Hence particular pains have been taken to tabulate and chart excessive rainfalls under a variety of conditions. To facilitate the study of such practical problems in engineering. Table LI has been prepared, showing all the necessary TABLE LI.-EXCESSIVE RATES OF RAINFALL IN CUMULATIVE FIVE-MINUTE PERIODS. 59 :>> Total duration. Begin-! End- ning. I ing. Excessive rate. Excessive periods in minutes. o Begin- H ning. ?n^~ J ®i 6 I 10 , 15 20 I 25 I 30 j 35 10 45 60 1894. 6 6.6.5p S.lOp S.lOp " 20 8.20p 8.20p 1S94.23 ll.ir,!i 1897.21 1.43p " 24 6.2Sp 1898. Ki 4.08p 1899.1(5 «.50i) T.OOp 8.45p lO.Oop 9.22p 7.30p 1 .25 DN 1.47 DX L4- 21st 9.15a 1.53 2lst 9.15a 1.53 10.52P 12.30p .«) l:.'.00n 2.35p .72 ].4!tp 9.15p .85 (i.KJp fi.09p 1.20 5.09p 8.20p .74 7.0.^1) 1901 . 24 10. 15p DX .02 10.25p 1902.25 n.3.-)p 6.20p .51 ,'i.37p 1903.,24 2.80a 4.20a .l-OS^ 2.53a 7.04p T 1.25 ...... 8.57p .20 .30 1.35 |.40 ' .. 10.27p .GO .20 .30 .45 .55 ,15 i.30 .30 ,35 1.40 I .. 9.64p T .15 11.06p j.60 .15 l.>.2'.p .05 .20 .45 .60 .70 2.00p .01 .57 .67 .70 .71 7.01p .01 .11 .29 .42 .44 ri.S'lp .24 .29 .57 ..59 .59 7.2 2.43p 6.3.5P 1.42p 13.37p 6.40p 4.65P Duration. h. m. 0-21 1-20 .27 .20 .40 T .25 i.45 T .25 1.35 T .30 .45 ,03 .25 .26 ,38 .26 .6b .13 .36 .15 .34 .3? .66 .31 .80 .31 .50 .45 .50 .99 .99 991.04 1.05 .60 1.17 .65 .81^ .91 .65 .70 .95 .58 .50 .. .79j .86; 1.181.541.771.851.911.94 .55 .56 ..':.... .. 75 .86 1.20 .33 .98 1.722.23i2.52!2.693.87 .34 ..54 .61 .71 .72 .. .. .03 .18 .64 .9311.00, .. .. .03 .25 :.50 1.38 1.37 .98 .74 .88 1.32 1.732.232.622.692.87 1.63 1.23 .94 .951.201. 1.9411.041 .9511.201. MARYLAND WEATHER SERVICE 2or TABLE LI CONT. -EXCESSIVE RATES OP RAINFALL IN CUMULATIVE FIVE-MINUTE PERIODS. Total duration. 6 Excessive rate. 1- Excessive periods in minutes. Begin- ning. End- ing. Begin- End- uing, ing. 5 10 15 20 35 30 35 40 45 50 60 80 August. 1895 31 I 4.18P 5.00p .78 .78 1 1897 9-10 11.30P 5.10a l.«0 1.30a 2.20a .40 .03 .07 .13 .19 .25 .43 .52 .62 .77 .83 " 23 11.0:3a 11.5-b 11.15a 11.45a 1.08 .05 .09 ..33 ..5"? .77 .9'7 " 12 10.37a 2.05P 1.43 11.4.5a 12.20p .09 .OP. .10 ..31 .43 .78 1.00 1.10 V " 27 6.35p 7.30p .81 6.40p 6.55p .02 .:50 .62 .76 .78 V 1902 3 9.30p 9.50p .tit) 9.33p 9.43p T .m .64 .66 ]' 1902 5-6 11.40p D.V 1.43 11.46p 16.25a .01 .10 .30 .49 .56 .64 .89 1.34 1.40 Y " 11 :5.11p 4. CM) .55 3.46p 3.58p .14 .17 .35 .41 y' " 27 4.29P 5.30p 1.2: 4.3:p 5.02p T .10 .47 .89 1.12 1 .-i'i 1.24 y 1903 26 9.45a 10.3op .52 9.53p lO.llp T .11 .37 .49 .63 *^ U\nA5p 8.05a 1.53 4.30a 4.41a .98 .03 .37 .49 .60 Duration. Duration. h. m. h. m. Aver... 2-48 1.09 0-24 .18 .15 .35 ..51 .65 .85 1.2;^ 1.35 1.13 1.05 .91 1.06 Great.. 12-50 2.05 0-50 1.08 .46 .78 1.16 1.36 1.43 1.24 1.39 1.51 1.56 1.00 1.06 September.* 1894 8 1.27p 1896 3 2.46p '* 19 4.66p 1897 16-7 10.38p 1898 26 6.20p 1899 2 5.30a 25 5.20p 1900 15 1.20p 1.55p 3.431 1.51p 3.57p .i.i.ip ..in ^.»op a.oip 5.55pl.05 5.1.5p 5.39p 1.1.5a .71 li).45p 10.56p 6.45p .35 C.20p 6.35p 8.25a 8.05p DN 2.15 3.61 Aver. . . Great.. Duration. h. m. 1-35 3-66 1.16 13.61 C.20p 5.54a 5.43p 10.45P 6.14a 6.30p 11.30p Duration. h. m. 0-33 0-45 T T T .03 .11 1.87 .30 .10 .30 .30 .14 .13 .09 .18 .50 ..30 .65 .39 .31 .36 .31 .60 ..35 .90 .43 .35 .39 .38 .86 LOO .48 .40 1.03 I'Oo .50 .53 1.18 ;75 1.33 !87 1.46 ioo 1.49 !92 1.55 L53 1.61 " .35 1.87 .17 .30 .40 .65 .51 .90 7'' .83 1.18 1.04 1.32 1.16 1.46 1.20 1.49 1.24 1.55 1.53 1.53 1.61 1.61 October. 1897. 12 1900. 8 19a3. 8 5.35a 10.05a 12.45p 1.35p 9.10a 2.06p 1.93 6.59a 8.00a .40 12.54p 1.05p 1.07 9.11a 9.16a .14 T T .17 .20 .25 .36 .36 .31 .38 .48 .48 .48 .53 .53 .53 .67 .67 .67 .83 .82 .82 .93 .92 .93 1.05 1.05 1.06 1.21 1.21 1.35 1.35 1.35 1.45 1.46 1.45 V Aver.. Great. Duration. h. m. 3-25 4 55 1.13 19.2 Duration. h. m. 0-28 1-01 .06 .14 .21 .26 .31 .38 .34 .33 • Sept. 6, 1895, rain began 2.10 a. m., ended 6.45 p. m.. amount 4.76 inches. The gauge was not working, hence onlj' stick measurements were possible, and it is estimated that 1.06 inches rain fell between 4.00 and 6.00 p. m. 208 THE CLIMATE OF BALTIMORE TABLE LI CONT.— EXCESSIVE RATES OF RAINFALL IX CUMULATIVE FIVE-MINUTE PERIODS. u P Total a duration. es |2 Excessive o ^ rate. || Excessive periods in minutes Begin- 1 End- | ning. ing. p Begin- ning. End- la© ing- i*^ 5 10 j 15 20 25 30 35 40 1 45 50 60 80 November. 1896 '^^'' '' ^^^ ' ^ ^^" i Q.ti 3 nnn 4.00p .03 _. i . _. .93 1 ! i Table LI contains a complete list of all occurrences of excessive rainfall from 1894 to 1903, arranged according to months and years. The scale of excessive rates of precipitation adopted by the U. S. Weather Bureau, and employed in the above classification, is shown in the text. The above table shows the year, month and day of occurrence of the excessive rainfall, the time of the beginning and ending of the entire rain period, and also of the period of excessive rate of fall, the total amount of fall, and the cumu- lative amounts which fell in the successive 5-minute periods. The symbol { \/) in the last column indicates the occurence of a thunderstorm in con- nection with the rainfall; •.• indicates the occurrence of a thunderstorm on the same day at some other hour. details of every instance of excessive rate at Baltimore occurrintj since 1893, at which time the automatically recording raingage was installed at the Baltimore office. The arrangement is by months and years and shows the following facts: the year, the day and hour of occurrence, the total amount of the rainfall during the entire progress of the storm, the time of beginning and ending of the excessive rate of fall, the amount of rainfall before TABLE LII.-SDMMARY OF EXCESSIVE RATES OF PRECIPITATION IN CUMULATIVE FIVE-MINUTE PERIODS. E xcessive rains. o® V Excessive periods. d Dura- tion. Am'ts. 5 10 15 20 25 30 35 40 45 1 50 1 00 ' Averages. January March May June July August September. October — November . Average — h. m. in. min. 1 12-35 1.35 22 .05 .10 .25 .45 1 0-40 .34 6 .01 .26 .29 la 3-45 1.(12 19 .23 .40 .51 ..56 .6'; .83 .94 1.51 1..T4 i:{ 4-23 .75 18 .21 .35 .46 .."17 .fit- .78 l.H 1-56 i.on 21 .35 .m .74 .88 1.2S 1.43 1.62 1.23 .94 .95 1.20 1. 30 2-48 1.09 24 .15 .35 ..51 .65 .85 1 23 1.25 1.13 1.05 .91 1.06 8 1-35 1.16 92 .27 .40 .51 .72 .82 1.04 1.16 1.20 124 l.5h 1.61 3 3-25 1.13 36 .21 .31 .34 .48 .52 .67 .82 .93 1.05 1.21 1.35 1 1 3-10 .98 .93 78 3-49 .98 20 .16 .38 .45 .62 .79 1.00 1.16 1.20 1.16 1.16 1.23 1 80 45 :marylaxd weather service 209 TABLE LII CONT.- -SUMMARY OF EXCESSIVE RATES OF PRECIPITATION IN CUMULATIVE FIVE-MINUTE PERIODS. Excessive rains. Excessive periods. 50 60 Dura- tion. Am'ts. Si 5 10 15 20 25 30 35 40 45 SO Greatest. Ih. m. I in. |h. m. January 12-35 1.35 0-22 March 0-40 .34 0-6 May 11-55 1.63 0-37 June 15-45 1.21 0-30 July 6-47 2.87 1-20 August 12-50 2.05 0-59 September 2-55 3.61 0-45 October 4-55 1.92 1-01 November 3-10 .98 .05 .01 .57 .35 .37 .46 .30 .25 Greatest. Year Month Day.... Hour of beginning.. . 15-46 2.87 1-20 .57 .98 .46 1.7" 2 L161 .901 .36 ,80 .232 ,36:1 .021 1.22 1.722.2312.522 1.44 1.51 1.64 1.39 1.46 82 s : i "^ 1.49p 1903 July 12 12.04 p. m. 1.94 1.51 1.49 .92 1.94 1901 26 1.04 1.56 1.55 1.05 1.56 1899 .951.20 1.001.06 1.5S!1.61 1.211.35 .. I .93 1.581.61 1900 Sept. 15 1.80 l!45 1.80 1896 n 6.03p H.15p 10.45p 9.19p Table LII contains a summary of facts contained in Table LI. the excessive rate began, and the amounts recorded in cumulative five- minute periods during the continuance of the excessive rate. When the rain fell in connection with a thunderstorm this fact is also noted. It is a matter of record that in almost every instance these excessive rain- falls occur in connection with thunderstorms. In Table LII there is a summary of the preceding table, containing the average amounts recorded during cumulative 5-minute periods, and also the greatest amount for the same intervals during the entire period of ten years. The average and maximum rates of precipitation are also shown graphically for the entire year in Fig. 58. Excessive rates of rainfall as defined by the U. S. Weather Bureau, and employed in its published records, are indicated in the following table : TABLE OF RATES CONSIDERED EXCESSIVE. Amount. Time. 0.23 Inch In 5 minutes. 0.30 '• 10 •• 0.35 " 15 •' 0.40 •• liO •• 0.45 '• 25 " 0.50 •' 30 •• 0.55 '• 35 •• Amount. Time. 0.60 inch in 40 rtJinutes. 0.G5 " 45 0.70 " 50 0.75 " 60 0.80 " 80 0.90 " 100 1.00 " 120 210 THE CLIMATE OF BALTIMORE An inspection of Table LI will show that excessive rates of fall as defined above are confined almost entirelv to the warm months of the 10 20 40 50 60 80 100 MiN. ^ \ Inches 2.50 / / \ / // \ 2.O0 / \ ■/ / 1 I 1-50 // / \ ) y l\ 7 \ A y^ X / / X ^^c ■~-«.\__ ^.^ r.oo \ / 7 r y y' ^ 0.50 / ^y / D --- / ^>^ ^ ^ Fig. 58a. — Excessive Rates of Rainfall. A. The curve A represents the maximum rate of precipitation attained in any con- secutive 5, 10, 15, etc., minutes during the heavy rainstorm of July 12, 1903. B. Represents the rate during the first 5, 10, 15, etc., minutes after the beginning of the excessive precipitation of the storm of July 12, 1903. C. Represents the average rate in 78 cases of excessive rates of fall. D. Represents the lower limits of rates considered excessive by the U. S. Weather Bureau. year. Of a total of 84 instances recorded in eleven years, the seasonal distribution is as follows : Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year 10 1 1.3 14 19 21 9 5 1 S4 Over 90 per cent of all occurrences are credited to the months of May to September. None have been recorded in February, April and Decem- MARYLAXD WEATHER SERVICE 21] ber, while January, March and Xovember have but one each. Over 90 per cent of all excessive rains have occurred in connection with thunder- storms in the immediate vicinity of the station of observation. The average rate of fall based upon all excessive rains during a period of 10 years is indicated by the following figures for the five-minute cumulative intervals from 5 to 80 minutes. E F ^*«s,„^ 10 20 30 40 50 60 80 100 MiN, Fig. 58b. — Excessive Rates of Rainfall. E. Represents the rate of precipitation per hour during the heaviest 5, 10, 15, etc., minutes of rainfall in the storm of July 12, 1903. F. Represents the average rate per hour for 78 cases of excessive rates of fall. AVERAGE AMOUNTS OF EXCESSIVE RAINFALL. (In inches and hundredths.) Number of minutes from beginning of excessive rate. 5 10 15 20 25 30 35 40 45 50 60 80 Amount of fall Ki .33 .45 .02 .79 1.00 I.IG 1.20 1.16 1.16 1.23 1.62 Selecting from Table Lll tlic maximum rate of fall for each of tbe periods indicated, irrespective of the month in which it occurred, we have the figures below, whicli represent the maximum observed rates for the first 5, 10, 15, etc., minutes after the beginning of the excessive rate of fall. 15 212 THE CLIMATE OF BALTIMORE MAXIMUM RATES OF RAINFALL. Minutes. During first ; 6 Amount of fall i .57 Month May Tear l 1897 [ 15 1.72 20 I 25 2.23 2.53 30 36 2.87 July 1903 40 1.94 July 1901 45 1.56 Aug. 1899 50 60 80 1.58 1.61 , 1.80 September 1900 July 1896 Amount. Rate per hour. 0.80 Inches. 9.60 inches. 1.3.5 8.10 " 1.92 •' 7.68 " 2.32 " 6.96 " 2.58 " 6.19 " 2.75 " 5.50 " 2.87 " 4.31 " 2.87 " 3.44 " 2.87 2.87 " 2. ST •• 1.44 " The destructive storm of July 12, 1903, which swept over Baltimore and vicinity, was accompanied by a downpour, the rate of which was probably never equaled in the annals of Baltimore weather. The rate of precipitation as measured at the local office of the U. S. Weather Bureau is indicated by the following figures, and graphically shown in Fig. 58: RATE OF RAINFALL IX STORM OF JULY 12, 1903. For any period of 5 consecutive minutes 0.80 inches 10 " •' 15 '■ ■• 20 " " 25 " '• 30 " " 40 " " 50 " " 60 " " 120 " " Table LII, Maximum Cumulative 5-^linute Periods, shows a different value, namely, the precipitation of the first period of 5, 10, 15, 20, etc., minutes after the he ginning of the excessive rate of fall. The storm of July 12, 1903, showed a maximum rate for every period from 5 minutes to one hour. A rough calculation has been made of the amount of water which fell within the limits of Baltimore City during the storm of July 12, 1903. It was probably one of the most severe storms ever witnessed in the city. While the area of destruction in this type- of storm is fortunately of extremely limited extent, it may be safely assumed that practically all of the city had a rainfall approximately equalling the amount recorded by the official gage. The central path of the storm was about a mile dis- tant from the office of the Weather Bureau ; nearer the center of the path the rainfall was probably heavier than in portions of the city beyond its area of destructive winds. Hence we may assume the officially recorded MARYLAND WEATHER SERVICE 213 fall to be a safe estimate of the average for the entire city. Assuming the area of Baltimore City to be 31.15 square miles, we may readily compute the amount of water which fell during the progress of the storm : WEIGHT OF RAINFALL IX STORM OF JULY 12, 1903. (Within the limits of Baltimore City.) Depth Gallons per acre. o( fall. During the first 5 minutes 33 inch. 10 ■• 98 " 15 •• 1.72 " 30 " 2.69 " 35 " 2.87 " During the heaviest 5-minute fall 0.80 " 7,466 22,172 38,913 60,859 64,931 18,099 Tons within City Limits. 745.428 2,213,696 3,885,162 6,076,369 6,482,967 1,807,098 Frequency of Consecutive Days with Eain or Snow. In a large percentage of instances when rain or snow falls, it is con- fined to a single day. The exact percentage depends largely upon what is regarded as a day with rain. Including a light sprinkling rain, or a flurry of snow, in our calculations we find that in the past 33 years, the precipitation was confined to one day in 36 per cent of the total number of days with rain or snow. Considering only measurable quantities of precipitation (0.01 inch or more) the percentage is increased to 45. In 28 per cent of all cases the rain or snow extended into two consecutive days, and in 16 per cent, three days, when we take account of " traces." Counting only appreciable quantities, the percentages are respectively 31 and 13. The instances of precipitation on more than three days decrease rapidly with each successive day added. In the table below, the frequency, excluding " traces " of rain or snow, is shown for each month and for the entire year to the maximum period, namely, 14 days. frequency of consecutive days with rain or snow. (Expressed as percentages of the total number of cases of appreciable rainfall in 33 years.) Days. Jan. Feb. 42 41 3.5 35 13 13 7 8 O 1 2 1 1 1 "i Mar. Apr. 88 33 18 6 4 1 1 1 May June ' July Aug. Sept. Oct. Nov. ' Dec. Year 10 45 31 13 6 3 1 0.4 0.1 0.2 0.1 0.0 0.1 0.0 214 THE CLIMATE OF BALTIMORE Inchiding days with '' traces " of precipitation, the annual frequency is indicated by the following figures : ANNUAL FREQUENCY OF COXSECUTiyE DAYS WITH RAIN OR SNOW. (Including "traces.") Number of days. . . 1 2 3 4 1 5 6 1 7 2.5 0.9 8 1 9 ! 10 11 13 13 14 Percentage of pos- sible occurrence. 36 I 28 16 9 5 0.9 0.4 0.2 0.3 0.1 0.0 0.1 In the past 33 years there has been no single instance of rain or snow on more than 1-i consecutiye days, and but few in which the rain occurred on more than six consecutiye days. There haye been longer periods of " unsettled weather," but these will be found, upon inyestigation, to haye been interrupted by one or more days without eyen a " trace " of rain. (See Table LIV, Wet Spells.) Dry Spells. While the rainfall is quite eyenly distributed throughout the year, there are at times periods of many days without appreciable precipitation, or at least of amounts insufficient for the requirements of plant growth. During some seasons of the year this scarcity of rain is of comparatiyely little importance; during periods of critical crop growth, howoyer, it becomes a question of serio'us moment. What constitutes a dry spell is largely a matter of arbitrary judgment. It is not alone the number of days without rain ; pre-existing conditions enter largely into the problem, as well as the stage of deyelopment of yegetation. In the classification of dry spells noted in Table LIII, the selection includes, as a lower limit, all periods exceeding two weeks during which the precipitation was less than one-tenth of an inch. While this limit is a purely arbitrary one, a period of 14 days with either no rain or less than one-tenth of an inch is a long interyal considering the ayerage frequency of rains in this vicinity. For periods longer than two weeks, a proportionately larger quantity of rainfall was allowed, keeping in yiew the desire to select only such dry spells as fell yery far short of the normal quantity of rain for the dry interyal. In the course of 33 3^ears there haye been 58 periods answering the requirements of the definition, ayeraging a little less than two per year. A drought of this character cannot be regarded as seyere, but it may be followed by considerable loss to the farmer or trucker :marylaxd weather service 215 during certain critical periods. Ordinarily there are from ten to twelve days per month with rain to the extent of .01 inch, giving a ratio of one day with rain to two without. These are not uniformly distributed through the month, but are very likely to occur in groups of two or three days. The total monthly frequency of periods of this class, together with the average, the maximum and the minimum number of days included, is shown in the following tabular statement : DRY SPELLS IN 33 YEARS. I With less than one-tenth of an inch of rainfall in two weeks or more.) Jan. Feb. Mar. Apr. May June July Aug.' Sept. Oct. Nov. Dec. Year Total frequency " 3 3 3 7 3 3 3 6 12 5 8 58 Max. duration.. Min. Aver. 21 16 18 31 18 22 31 19 26 22 U 19 36 14 21 29 20 25 25 23 45 23 30 34 17 25 51 15 27 48 14 33 51 20 29 51 days 14 " 25 " Aver. rainfaU .. .05 .17 .12 .15 .11 .21 .10 .28 .12 .20 .17 .30 .18 inch J.. 18 1 1 1 1 75 18 1 1 1 1 8C 18( — r-i , 1 5 18 1 1 1 1 )0 18 — 1 — 1 — i— r 1 )5 19C -Till 1 1 ■] I 1 Fts 1 1 1 MC" ll 1 APn 1 1 1 »,. 1 1 1 1 1 1 JUVt 1 1 Juit 1 1 Ayr, 1 1 1 St>.T 1 1 1 1 1 1 Oct 1 ll 1 1 1 III 1 1 Nov 1 1 1 II 1 Otc 1 , , , II, , \l 1 1 : 1 III ill Fiu. 59. — Dry Periods. The diagram shows the frefjuency of occurrence and the seasonal distribution of all periods of two weeks or longer with a precipitation of one-tenth of an Inch or less from 1871 to 1004. The length of the period Is roughly shown by the length of the heavy vertical lines. The exact duration of the period is shown in Table LIII. 216 THE CLIMATE OF BALTIMORE TABLE LIII— DRY SPELLS. Tear. !2;b 1871. 1872. 1873. 1874. 1875. 1876. 1877. 1878. 1879. 1880. 1881. 1882. 1883 1884. 1885. 1886. 1887. 1888. 1889. 1890. 31 9 .23 18 0.25 23 SI .08 21 17 .01 14 .03 ir .29 29 .02 19 35 .04 16 SO .18 14 i .23 28 .05 15 .3: 20 .24 22 U .05 20 5 .17 34 9 .16 33 SO .17 19 9 .04 17 6 .10 23 20 .30 36 2i .59 38 .05 19 21 .39 51 .27 36 16 .10 21 17 .03 20 29 .06 ^3 9 .66 41 18 .31 22 is 1 3 1 2 3 1 2 1 MARYLAND WEATHER SERVICE 217 TABLE LIII CONT.— DRY SPELLS. Year. 189L 1892. 1893. 1894 1895. 1896. 1897. 1898. 1899. 1900. 1901 J. 1902 ■! 1903 < No. of dry spells. 2S .04 16 20 .06 31 20 .12 31 1 .16 19 .09 27 .36 36 18 .07 23 .26 38 .13 20 .10 16 7 .10 20 12 3 .26 22 SS .0 14 .13 39 .27 29 S7 .09 23 51 ^i; 58 Table LIII contains a list of all of the more pronounced dry periods occurring near Baltimore from 1871 to 1903. No strict definition of a dry spell has been adhered to in the selection of these periods, but the table contains all periods exceeding two weeks during which the precipita- tion amounted to less than one-tenth of an inch. The length of the dry spell in days is Indicated by the figures in heavy face type, the date of ending by italic figures, and the total precipitation during the period by roman figures. These dry spells are most frequent in the month of October, and hence after the harvest season. Their occurrence in May has been compara- tively frequent. Coming at a time when soil moisture is a matter of the highest importance, the dry spells of this period are serious matters. The 218 THE CLIMATE OF BALTIMORE seasonal distribution of these periods of deficient moisture is shown graphically in Fig. 59 for each year from 1871 to 1903. In another classification of dry spells, all periods of ten or more con- secutive days were included in which precipitation was less than .01 — r -r- 1 1 1 I I 1 1 1 1 I 1 1 1 1 1 1 I 1 1 1 1 1 1 I 1 T-r 1 1 1 1 1 i II 1 1 1 1 II II 1 II 1 II 1 1 1 1 1 1 1 1 1 1 h ii 11 1 1 1 1 1 1 1 II 1 1 1 II III II III 1 1 I 1. 1 il 1 iliiii II 1 1 il II 1 III II 1 1 1 1 liilll 1 , . ,1,, 1 1 1 1 1 l< ,1, 1 1 1 1 1 1 1 1 1 1 1 1 1 Fig. 60.— Dry Periods. The diagram shows the frequency of occurrence and the seasonal distribution of periods ^ith less than one-hundredth of an inch of precipitation in ten days or more from 1871 to 1004. The relative length of the period is approximately shown by the length of the heavy black vertical lines. inch. The total number of such periods, with the average and greatest duration, is shown in the following figures: DRY SPELLS IN 33 YEARS. (With less than one-hundredth of an inch of rain.) a •-s 4 14 18 fa 6 13 14 s 3 11 11 < 9 12 14 6 14 21 a 11 12 <-> 1 4J 0. o O O > o Year Total number 8 12 18 8 11 12 17 12 22 17 13 19 12 14 29 5 14 17 103 13 days og Greatest du ration :\rARYLAXD WEATHER SERVICE 219 The above table reveals the interesting fact that the longest period experienced by Baltimoreans in 34 years without rain was 29 days. This occurred in Xoveniber, 1874. September and October share the distinc- tion of having 17 periods each of this class of dry spells out of a total in 34 years of 102. The average duration of such periods is only 13 days. These facts are not in accordance with popular impressions. Scarcely a summer passes without some reference to periods of five or six weeks '' without a drop of rain." These statements, upon investigation, are generally reduced to " no rain of consequence." The dry spells of this class are graphically shown in Fig. 60. There seems to be no apparent periodic grouping of these periods of deficient rainfall, either in Fig. 59 or in Fig. 60. The following list comprises seasons with a marked defi- ciency in rainfall : SEASONS WITH DEFICIENT rRECiriTATION. (Departures below the normal in inches and liiuiciredths.) Winter. Spring. Summer. Autumn. 1829-30. .. .4.17 1822 5.11 1844.... 4. o4 1819.... 4. 57 18G4-65 4.70 1827 4.12 1849.... 4. 14 1825.... 4. 79 1870-71 6.00 1845 4.46 1864 6.66 1842 4.22 1871-72 5.73 1847 6.03 1869 7.(19 1863 4.32 1900U1. . . .4.80 1855 5.91 1870 4.85 1870 4.33 1856 4.87 1893 6.69 1879 5.06 1866.... 5. 85 1894 6.21 1884 5.23 1869 4.49 1903 4.56 1900 4.66 Wet Spells. While rain or snow storms do not usually exceed two or three days in duration, there are frequently periods of much more extended rainy or unsettled conditions. The more conspicuous "' wet spells " occurring since 1871 are emiiiu'rated in Table LIV, which contains a list of all jioriods of 10 days or less during which the rainfall or snowfall was equal to or exceeded the mean monthly amount. Longer periods were included when the precipitation was proportionately excessive. The last day of the wet spell and the duration in days are indicated in the table, together with the total amount of the precipitation. Such periods have been recorded on an average of less than two times per year, or, more accurately, 51 times during 33 years; the limits of variability are and 5. They occur in all months of the year, with a decided preponderance, however, in the 220 THE CLIMATE OF BALTIMORE warm months of July, August and September. Their frequency of occurrence and seasonal distribution are shown graphically in Fig. 61. One of . the most remarkable periods of unsettled weather was that accompanying the northeast storm of April 19-25, 1901. Rain began early in the morning of the 19th and continued during the greater part of I I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 h . 1 1 1 . ■ 1 . 1 II 1 1 h 1 1 ■ 1 1 1 I 1 1 1 1 1 1 1 1 1 1 1 1 ii 1 I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 _ Fig. 61.— Wet Periods. The diagram shows the frequency and seasonal distrilnition of periods of ten days or less in which the average monlhJij amount of precipitation was recorded. The amount recorded is roughly indicated by the length of the heavy vertical lines ; the exact amounts and the length of the periods are shown in Table LIV. seven days, or 162 hours between the beginning and ending of precipi- tation. The rainfall was not continuous, however, scattered showers falling on April 21, 22, 23 and 25. The total amount of rainfall recorded was 2.03 inches, not a large amount considering the great duration of the storm. Another noteworthy rainstorm was that of May 16-26, 1894. During this period of eleven days the rainfall was scattered and at times MARYLAND WEATHEPx SERVICE 221 heavy. The actual duration of precipitation was ahout 63 hours, total amount recorded durinij the entire storm was ■i.45 inches. The TA.BLE LIV.— WET SPELLS. "" 5 1873 } i 1873 \ 1874 -j 1875 -j 187fi -j 1877 -j 1878 ■/ 1879 1880 1881 \ 1882 j 1883 -j 1884 - i 1885 "I 1886 ^ 1887 I 1888 ■< 1889 ^ 1890 ' li 11 3.46 19 10 4.61 13 5.4S 19 17 3.71 13 11 5.65 52 3.83 29 23 6.39 30 6.31 16 4.25 11 11 4.07 i 17 7.63 1 7 4.33 23 14 S.81 29 9.61 23 4 14.73 31 9 5.20 6.37 11 6.9S 4 4 4.70 26 12 5.76 8 5.60 13 n 4.30 9 5.52 fi 4.19 2i 18 10.10 15 10 5.23 27 7 4.61 IS 12 4.53 26 5 3.79 4 12 1.09 13 3.39 3.77 S6 21 5.17 222 THE CLIMATE OF BALTIMORE TABLE LIV CON'T.— WET SrELLS. Tear. 1891 1893 < 1893 \ 1894 -j 1895 } 1896 ^ 1897 \ 1898 I 1899 - 1900 \ 1901 ^ 1902 1903 < No. of wet spells. 9 9 5.13 21 11 4,S1 2S 13 4.55 s 8 14.43 16 11 4.0a ■29 13 6.31 20 10 5.98 13 19 7.15 4.S6 t6 6.03 26 5 5.29 4.11 z"5 [' [» [' ll" ![' If- 30 \\ 34 3 61 Table LIV contains a list of all of the more pronounced wet spells occuring near Baltimore from 1871 to 1903. The table includes all periods of 10 days or less during which the average monthly amount was recorded. Longer periods were included when the precipitation was proportionately excessive. The last day of the wet spell and the duration in days are indicated by figures in Italics and Roman respectively; the bold face type indicates the amount of precipitation recorded during the stated period, in inches and hundreths. Table LIV comprises the more conspicuous wet spells of the past 34 years based upon excessive amounts of rain. Another table was pre- pared, but is not published in full, in which the basis of selection is the duration of unsettled, rainy weather. It includes periods of six or more MARYLAXD WEATHER SERVICE 233 consecutive days with a " trace " or more of precipitation, 8 days with not more than one day without rain or snow, or 10 days with not more than two days without precipitation. ^More extended periods were in- cluded when precipitation occurred on two in each successive period of three days. The total number of " unsettled periods '' of this descrip- tion comprised within the Si years is 164. The distribution throughout the year is indicated in the following summary; the last line indicates the number of intervening days without rain : PERIODS OF UNSETTLED WEATHER. (Six or more consecutive days with rain.) Total frequency Ma.xiraum duration in days Number of davs without rain d 03 11 19 i 15 17 4 S3 20 23 2 a < 14 13 1 oj 23 23 4 c 3 ►^ 13 18 3 "3 1-5 12 15 si 3 < 17 19 4 P. $ 9 12 § 7 10 > o is 11 10 1 o P 12 6 03 * a o — 2 -z , >1 Bi O >> jq Q 1) rt . 03 -.. < ^ ■;; "J r S 226 THE CLIMATE OF BALTIMORE cipitation by days and months is shown for the same years in Fig. 63. The depth of rainfall is indicated by tlie length of the heavy vertical lines. The dry year (1900) was deficient in rainfall frequency as well as in amount. The normal year (1888) had 154 rainy days; the dry year (1900) had 115, and the wet year (1889) had 164. The rainfall of the dry year was only about half that of the wet year, the amounts being 31.57 inches and 62.35 inches, respectively. The normal precipitation is 43.34 inches. The great excess in the wet year was due to the heavy spring and early summer rains of 1889. TABLE LV. -SUMMARY OF PRECIPITATION DATA. (1871-1903.) January . February March April May June July August — September October .. . November. December . Means. 3.20 3.70 3.99 3.27 3.63 3.78 4.6r. 4.20 3.8f) 2.99 2.99 3.07 Mean depart. Year 43.34 J. 11 1 47 1.44 1.19 l.fil 1.45 2.03 1.80 1.83 1.48 1.12 1.28 5.36 12 Monthlj' and annual amounts. Greatest. Least. lA. S O 1^ S fi.42 1892 7.07 1896 7.94 1891 8.70 1889 7.26 1894 8.08 1883 11.03 1889 9.49 1873 10.. ')2 1876 fi.85 1902 6.85 1877 7.07 1901 62.35I 1889 201 194 199 266 200 215 241 229 267 231 224 228 0.88 0.65 1.19 1.37 1.00 0.90 1.40 0.64 0.09 0.16 0.65 0.37 1872 1901 1894 1885 1900 1901 1881 1877 1884 1874 1882 1896 144 31.. 57 ! 1900 No. of days* with pre- cipitation. 131 164 104 in in 1 1889 1871 h. m 7:18 8:14 6:1:^ 6:25 4:23 2:45 3:03 2:42 4:0:^ n-.m 6:15 6:46 5:18 Great- est in 24hr8. 1.951896 3.481896 3.511881 3.581889 2.991886 4.471885 4.021889 4.361873 4.761895 .3.421^73 2.851877 2.881901 4.76 1895 * Omitting days with only a " trace " of rainfall or snowfall. Table LV contains a summary of the principal facts relating to precipitation and published in full in preceding tables. The first column of figures shows the normal monthly precipitation based on 3.3 years' observations; the second column shows the proportion of the annual precipitation which falls in each month; the third column shows the average amount by which the actual monthly fall differs from the normal monthly fall, either above or below; the next following column of figures shows the same fact expressed as a percentage of the normal monthly precipitation. marylaxd weather service 227 Snowfall. There are many difficulties in the way of securing accurate measure- ments of the amount of snowfall, difficulties which are inherent in the conditions attending precipitation in general, together with the additional one introduced when the temperature is at or near the freezing point of water. The method of exposure of the snow-gauge is of highest im- portance even under favorable atmospheric conditions for securing all the falling snow. When the wind is high, and especially when it blows in gusts, the snow is drifted and blown about to such an extent as to make it impossible to catch any but a small percentage of the total fall in the gauge. Under such circumstances it is necessary to resort to a different method of measurement. In an open and exposed area several measure- ments are made of the actual depth in inches of snow on the level ground at points which, in the estimation of the observer, represent most nearly the average depth in the vicinity of his station. The average of thc^o measurements is then accepted as the true depth of snowfall. In order to secure the equivalent depth in melted snow, the several measured depths are melted and the average depth of water obtained is computed. The amount of water yielded by a given depth of snow varies greatly with the temperature of the snow and the conditions under which it falls. A light fluffy snow may require 15 to 20 inches for one inch of water : on the other hand, a wet soggy snow of 4 or 5 inches may melt to an inch of water. In rough measurements, under average conditions, the ratio is about ten to one, and this is the relation generally assumed. With a slight change in temperature at or near the freezing point, the snow melts as it falls, or after falling for some time it may change to rain. These are some of the difficulties encountered in an effort to secure reliable snowfall data. The record of fairly accurate depths of snowfall at Baltimore begins with the year 1883. The record of frequency of snowfall begins much earlier, dating from the opening of the Weather Bureau Station in 18T1. The actual monthly and seasonal amount of snowfall recorded during each month and year from 1883 to 1904 is shown in Table LVI, together with the monthly and seasonal average amounts for the entire period of 21 seasons. The seasonal variations are shown in Fig. 6-1. The average 16 228 THE CLIMATE OF BALTIMORE u^ ^ _ o - in S <1 MARYLAND WEATHER SERVICE 129 230 THE CLIMATE OF BALTIMORE TABLE LVI.-MONTHLY AND SEASONAL SNOWFALL. (In inches and tenths.) Season. 1883-4.... 1884-.").... 1885-6... 1886-7.. 1887-8.. 1888-9 . 1889-90. 1890-1.. 1891-2. 1892-3. 1893-4. 1894-5. 1S95-6. 1897-8.... 1898-9 ... 1899-1900. 1900-1.... 1901-2 1902-3 1903-4 Average (1884-1904). Greatest Year Oct. Nov. Dec. Jan. Feb. Mch. Apr. May. Sea- son. 4.4 14.2 8.7 8.0 35,3 0.6 3.8 1.9 17.2 5.9 2.0 31.4 13.0 15.3 2.0 30.3 T 10.2 2.5 6.0 6.8 1.1 25.6 12.0 8.8 3.9 7.4 32.1 1.1 T 2.6 5.3 T T 9.0 T 0.1 2.5 2.3 T 4.9 T T 10.6 1.3 3.5 20.5 T 35.9 T T 14.5 4.2 25.6 T 44.3 o 3 4.3 8.1 11.7 4.0 T 30.3 0.2 3.1 1.0 11.7 T 5.0 21.0 T 3.0 5.0 9.3 0.6 17.9 T T 0.2 1.0 2.8 13.8 T 17.8 3.0 3.2 4.7 0.7 T 11.6 T 2.6 5.4 2.4 0.1 10.5 9.7 0.6 !.3 33.9 1.6 .M.l 0.7 2.5 13.0 9.5 T 25.7 T T 6.5 2.1 0.1 8.7 0.1 0.6 6.7 1.0 5.0 13.4 7.0 6.8 6.0 19.8 T 1.0 3.8 16.6 2.5 2.0 25.9 T 0.8 3.3 5.6 7.5 5.8 0.8 T 23.8 T 9.7 12.0 16.6 33.9 25.6 8.0 T 51.1 1898 1887 1904 1899 1892 1.HH4 1898-9 depth of snow recorded during each month and the greatest and least monthly amounts recorded, are as follows: MONTHLY SNOWFALL, fin inches and tenths.) Average . Greatest. Y''ear Least — Year (1884-1904) Oct. Nov. Dec. Jan. Feb. March April May *T 0.8 3.3 5.6 7.5 B.8 0.8 T T 9.7 12.0 16.6 33.9 25.6 8.0 T 1898 1887 1904 1899 1892 1884 0.1 1890 1898 1903 Season 23.8 51.1 1898-9 4.9 1889-90 * T represents a trace of snow. February is the month of greatest snowfall, followed by March, with January third in the order of depth. The annual fall has varied from a minimum of about 5 inches in the season of 1889-90, to a maximum of 51 inches in 1898-9. Every month of the year excepting January has at some period since 1882 been entirely free from snow. The greatest monthly snowfall occurred during February, 1899, when about 3-i inches MARYLAND WP:ATHER SERVICE 231 were recorded. Over half of this amount fell during the great blizzard of that month. Expressed in terms of the percentage of the total annual precipitation, the average annual snowfall at Baltimore is 5.6 per cent; that is, about one-eighteenth of the amount representing the total annual precipitation falls in the form of snow. The percentage has varied from 1 per cent in the calendar year 1889, to 11 per cent in 1892. Computing the rela- tive amounts which fall as snow and rain in the season of snowfall only, we have the following figures : RAINFALL AND SNOWFALL OF THE WINTER SEASON. (In percentage of total monthly precipitation.) Rainfall. Snowfall. November 97 per cent. 3 per cent. December 89 " " 11 " January 83 " " 17 " February 82 " " 18 " March 8G " " 14 " April 98 " " 2 " Average 89 " " 11 " Even in the mid-winter months of January and February, the amount of snowfall is generally less than one-fifth the total precipitation for those months. Dates of First and Last Snow. The first snow of the season usually falls about the 15th of Xovember, and the last about the first of April ; hence the average length of the season of snowfall is about four months and a half. These first and last snows are, however, usually only light fiurries. This is particularly true of the first autumn snows. In the 34 seasons since 1871, snow flurries have occurred as early as October 9, as in 1895 and 1903. The first snow of the season has occurred as late as December 17, as in 1883 and 1887. The early snows were not followed by either an abnormal amount or by an abnormal frequency of snows. The last snow of the season has occurred as late as May 6, as in 1891, and as early as February 22, as in 1903. Table LVII contains a record of first and last snows for each season from 1871 to 1904. 232 THE CLIilATE OF BALTIMORE TABLE LVIL-DATES OF FIRST SNOW IN AUTUMN AND LAST IN SPRING. (Including " traces " of Snow.) Year. 1871 1872 1873 1874 1876 1876 1877 1878 1879 1880 1881 1882 1883 18^4 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 Earliest. Latest .. Average First in Fall. Last in Spring. Nov 29 Mar. 4 " 29 •• 0'> " 13 " 21 " 13 8 Apr. 1 ^ 18 Oct. 15 Mar. 2 Nov 10 " 29 Dec 5 Feb. 25 Nov 5 Apr. 5 ** 13 Mar. 29 Nov 24 Apr. 4 " 26 " 11 Dec. 17 Mar. 31 Nov 3 Apr. 9 " 23 " 11 Nov 13 Mar. 8 Dec. 17 Apr. 5 Nov 24 Mar. 25 Oct. 23 Apr. 6 " 19 " 1 Nov 28 May 6 " 9 Apr. 15 " 15 May 4 " 30 Apr. 12 Oct. 9 Mar. 20 Nov 13 Ai)r. 7 " 33 Mar. 14 " 24 Apr. 28 Dec. 4 " 16 Nov 9 4 Nov 18 Mar. 6 Dec. 5 " 31 Oct. 9 Feb. 22 Nov 13 Mar. 28 Oct. 9 Feb. 22 Dec. 17 May 6 Nov 15 Apr. 1 Year. 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 190] 1902 1903 1904 The Frequency of Days with Snowfall. The frequency of days with snow, including " light flurries," varies greatly from year to year. The average number for a series of years is, however, fairly constant. Dividing the entire j^eriod of 30 years from 1871 to 1900 into three periods of ten years each, the average annual frequency was as follows : AVERAGE ANNUAL SNOWFALL FREQUENCY. (Including traces.) 1871 to 1880 16.4 days. 1881 to 1890 24.0 " 1891 to 1900 26.1 " Mean (1891 to 1903) 22.0 " MARYLAND WKATIIER SERVICE Oct. Nov. Dec. jan. Fes Mch. Apr. Ma 233 Inches 30 20 \ A \ / ^ B \ " ^^ ^ ^^ \; i^ c\ \ / ^ s. \ /^ N N \ / E ~^^ k ^ ^ A 30 Inches Oct. Nov. Dec. Jan. Feb Mch. apr May Fig. 65. — Monthly Frequency and Amount of Snowfall. A. The greatest monthly frequency of days with appreciable snowfall. B. The average frequency. C. The greatest monthly amounts of snowfall. D. The average monthly amounts of snowfall. E. The least monthly amounts of snowfall. With an average seasonal frequency of 22. the numher has varied from 5 as in 1875-6 to 40 as in 1892-3. The average monthly and seasonal frequency for 34 seasons, including light flurries of snow, or 234 THE CLIMATE OF BALTIMORE " traces," is indicated in the following table. The variations in the seasonal frequency are shown in Fig. 65. FREQUENCY OF DAYS WITH SNOW. (Including "snow flurries.") Average Greatest Least — Oct. Nov. 0.2 l.r, 3 5 4.1 10 Jan. Feb. Mar. Apr. May j Season 5.8 18 5.6 14 3.8 8 0.8 3 0.1 23 days 1 140 •' j 5 " If we do not take into account da3's with light flurries of snow but only days during which a tenth of an inch or more fell, or days with TABLE LVIII.— NUMBER OF DAY^S WITH SNOWFALL EQUALLING OR EXCEEDING 0.10 INCH. Season Nov. Dec. Jan. Feb. Mar. Apr. Season 1370 1 '6 5 4 6 1 3 6 o 3 3 9 3 8 3 i 1 3 1 1 3 3 6 3.0 3.6 2.2 2.3 3 5 2 o 3 3 1 1 4 4 9 3 4 3 6 2 1 3 4 7 3 5 1 6 3 4 3 o 3 3 8 2.0 3.8 3.6 3.1 3 3 5 5 3 2 2 1 6 3 3 3 o 6 13 3 3 5 3 2 3 7 8 4 3 2 9 4 3 1 2 4 3.1 3.5 4.3 3.4 5 1 3 3 3 2 3 3 5 6 1 3 4 4 6 3 1 6 1 1 4 1 2 3 1.8 2.6 2.5 3.2 1 1 1 1 1 1 3 1 3 1 0.2 0.7 0.3 0.4 6 1871 2 12 187" 3 13 1873 4 13 1874-5 1875-6 10 4 1876-7 1877 8 1 15 3 1878 9 10 1879-bO 1880 1 .... 4 6 18 1881 2 11 188" 3 4 18 188.3-4 1884-5 1885-6 1886 7 13 26 8 19 1887 8 16 1888 9 1 8 1889-90 7 1890 1 16 1891-3 . . 1893-3 18^3 4 4 1 13 23 15 1894 5 11 1895-6 1896 7 1 11 11 1897 8 • . 8 1898-9 . 3 18 1899-1900 1900 1 11 5 1901 3 . . 1 9 1903 3 7 1903 4 1 23 Average. 1871-1880 1881-1S90 1H91 1900 0.1 1.1 0.9 9.3 14.3 13.7 1871 1903 0.7 13.0 MARYLAND WEATHER SERVICE 335 what may be reaardeil as '' appreciable '' snowfall, the monthly and seasonal frequency is reduced considerably below the figures shown in the preceding paragraphs. A detailed list of such days is contained in Table LYIII, which gives a more satisfactory index of the snowfall con- dition of a season than the figures which include " traces." Basing our calculations upon "'' appreciable " snowfalls, Ave have an average seasonal frequency of 1"3 days. The season of 18T7-S contained but 2. while 26 were recorded in 1884-5. The variations in the seasonal frequency are shown in Fig. 65. The average per month for the S-i seasons since 1871 is as follows: FREQUENCY OF DAYS WITH SNOW. (Excluding traces.) Nov. Dec. Jan. Feb. Average.. . Greatest . Least 2.2 P 3.1 3.4 12 Mar. Apr. 2.2 e' 0.4 2 Season 13.0 days 26 Heavy Snowfalls. The heaviest snow noted in the oflficial records of the local otHce of the Weather Bureau fell during the great " blizzard "' of February, 1899. The fall occurred in connection with an Atlantic coast storm which reached Maryland at a time when the Middle Atlantic states w^ere in the embrace of the severest cold wave of the past 30 years. The ground was already covered by snow to the depth of about 10 inches, which fell from the 5th to the 8th, and to this layer 5 inches were added on the 12th and 15.5 inches on the 13th. At the close of the storm of the 12th and 13th, the depth of snow on the ground measured 30 inches in the city of Baltimore. Greater depths were reported from other parts of Maryland. The wind was high and the temperature was extremely low, rancriutl between 5° and 20° below zero within the state. As a result, the dry snow was very much drifted and settled in places to depths of 10 to 20 feet. The city was snoAvbound and all local traffic was blocked for two or three days. The greatest depth of snowfall for any 24 consecutive hours during 236 THE CLIMATE OF BALTIMORE this storm was 15.5 inches, according to the official measurements. Single snowfalls equalling or exceeding 10 inches in 2-i hours are ex- tremely rare in the vicinity of Baltimore. There was one on December 17, 1887, another on the 3d of February, 1886, and another on the 18th of March, 1892, in the 21 years since 1884. In Table LIX will be found a record of the heaviest 2-i-hour snowfall for each month and season from 1884 to 190-i. TABLE LIX.-GREATEST SNOWFALL IX 24 CONSECUTIVE HOURS. Season. 1883-4. 1884-6. 1885-6.. 1886-7.. 1887-8.. 1888-9.. 18^9-90. Oct. 1890-1 T 19 1891-a .. 1892-3 1 .. 1893-4 1894-5 189.5-6.... 1896-7.... 1897-8 ... 1898-9.... 1899-1900. 1900-1. 1901-2 1902-3. 1903-4. Greatest. Nov. Dec. Jan. Feb. 0.5 2.9 1.0 3.1 6 10.6 17 T 19 3.0 15 1.0' 28 T T 1.2 0.2 15 T 30 T i 20 3.0 30 6.5 9 13.0 3 1.5 5 4.0, 26 3.2 9 1.8 13 2.5j 20 I 2.3; 27 G.li 23 I 1.5: 2 T 0.1 29 1.0 29 4. 511898 10. 6 IBS' 3.0 26 0.2 13 2.3 22 2.0 26 0.6 12 0.71 27 T I 21 0.5 23 4.0 5 1.6 2 1.0 25 7.0 15 4.8, 12 0.5 27 2.6| 29 1.0 19 3.6 27 3.0 31 2.8 1 1.5 28 4.0 25 5.6 29 4.0 34 6.0, 29 3.0 26 4.0 6 7.8! 17 3.5! 25 4.3| 7 1.0! i 0.5 8 T 25 15.. 5, 13 6.0 17 I 2.0 3 1.0 17 5.0 17 1.6 19 Mar. 5.0 5 1.5 13 2.0 8 3.6 4 3.5 5 T 7 3.4 31 9.5 27 13.fl 18 3.6 4 '1' 26 0.6 11 April 0.7 6.0 11 T 14 2.4 2 1.6 7 4.5 15 0.1 4.0; 1.5i 18 .0!l892|l5.5!l899:i2.0l892 8.0 1884: May Season. 8.0 Apr. 5.5 Feb. 13.0 Feb. 4.0 Feb. 10.6, Dec. 2.5! Jan. 3.4 Mar. 9., 5 Mar. 13.0 Mar. 7.8! Feb. 4.0 Apr. 4.3 Feb. 6.0 Mar. 3.0) Nov. 3.0 Jan. 15.0! Feb. 6.0 Feb. 4.0 Jan. 5.6 Jan. 5.0 Feb. 6.0 Jan. 15.5 Feb. The first column shows the amount of snowfall, and the second the date of occurrence. Duration of Snowfall. An effort has been made to obtain a value for the average duration of snowstorms in this vicinity. For this purpose the records were carefully examined for times of beginning and ending of snowfall for the period from 1884 to 1889, and the period from 1893 to 1902. Nq g^eat accu- racy can be claimed for the results, as there is no method in use for automatically recording beginnings and endings of snowfall. However, the figures given are based on a tabulation of 266 cases of snowfall during MARTLAXD WEATHER SERVICE 237 TABLE LX.-SUMMARY OF SNOWFALL DATA. Means. ■S OD _j- .Mii- 5 ^ C O. o . ®_ = 5 , =i =i <»2 1 t.3 ^ ^^ o z Greatest monthly am'ts. Number of clays with snow. Greatest snowfall in 24 hours. 1871-1903. 1883-1903. Omitting traces. ^^^^^S^ 1883-1903. 0-5 . o *3o3 a e u §a a t,CW o >< - < ' c - .® ^ « ^ « c s s a o o > 03 a '^ < 'K < October T November 0.8 December 3.3 January 5.6 February 7.5 March 5.8 April 0.8 May. Season . T 33.8 T 9.7 12.0 14.5 33.9 25.6 8.0 T 1895* 1S9S 1897 1S92 1899 1892 1884 1893t 1212 364 259 452 441 1000 61.1 1898-9 215 0.2 2 T 0.7 4 1.6 5 4.5 2.2 9 4.1 10 10.6 3.1 9 5.8 13 7.0 3.4 12 5.6 14 15.5 o o 6 3.8 8 12.0 0.4 o O.S 3 8.0 0.1 1 T 12 26 o 22 40 6 15.5 in in in in 1884-5 1877-8 1892-31875-6 1895* 1898 1887 1892 1899 1892 1884 1893t 1899 in Feb. T Indicates a " trace ' • Also 1889 and 1890. of snow ; an amount less than 0.1 inch. + Also 1891. 16 years, and hence are fairly relial)le. Trace? of snow, or snow '' flur- ries," were included in the calculations. DURATION OF SNOWFALL. (Including traces.) Nov. Dec. Jan. Feb. March April Season 11 60 5.30 45 2;» 5.20 99 428 4.20 67 395 5.50 34 169 5.00 10 47 4.45 266 " duration in hours Average duration (in hrs. and miu.) 1S^8 5.00 Fogs. Fogs in the vicinity of Baltimore are confined mostly to the fall, win- ter and early spring months. The record contained in Table LXI applies only to dense fogs surrounding the local station of the U. S. Weather Bureau office. Their frequency is doubtless greater as the harbor or the bay is approached. A fog has been regarded as dense when it obscured objects at a distance of about 1000 feet, and it has 238 THE CLIilATE OF BALTI:M01{E been recorded only when it hung abont the station for one hour or more. The table includes only such fogs as have been described and which occurred since 1891, the earlier records being regarded as less reliable. TABLE LXI.— FREQUENCY OF DENSE FOGS.* Year. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Ann'l. 1891 3 i 3 1 1 1 5 2 1 i 4 23 1.8 1 i 4 1 1 i 4 17 1.3 1 i 3 3 3 1 '8 18 1.0 'i 1 1 i 4 0.3 1 1 2 0.2 1 1 0.1 i i o 0.2 2 1 2 6 0.4 i 1 3 "b 7 3 1 20 1.6 2 "i 3 5 1 i i 3 6 2 37 2T1 6 "b 1 3 2 1 2 i 3 6 31 2.4 13 1893. 1893 9 1894 1895 10 15 1896 13 1897 1898 1899 1900 8 9 13 14 1901 13 ]' 35 Oct. 19 JIarch 2 19 25 18 1S98 " 26 Nov. 1 3 Oct 1'' Dec. " 28 3 Nov. 2 Dec. 7 9 10 23 28 31 " 19 Jan. 6 Nov. 25 ** 5 April 1 " 28 11 Dec. 19 " 14 9 12 20 " 15 June 4 13 " 22 Dec. 3 Sept. 14 20 Feb. 10 16 Oct. 10 Nov. 3 Nov. 5 4 Dec. 20 24 21 Dec. 27 * Fog about station for one hour or more, and too dense to see objects at 1000 feet. the month of July during this period of 13 years, and but one in the month of June, two each in May and August. SUXSHINE AXD CLOUDINESS. Sunshine. In connection with a discussion of the amount of sunshine recorded at Baltimore, it is important to know the metliod employed in obtaining the record. The instrument in use at the local office of the Weather Bureau since 1893 is of the kind known as the electrical thermometric recorder. The essential parts of the instrument are the two glass bull)S, one of which is covered with lamp-black. The two Inilbs are joined by a tube, in the middle portion of which are the terminals of an electric circuit. The direct rays of the sun falling upon the Itlack bulb 240 THE CLIIMATE OF BALTIMORE will raise the temperature of the air within to a higher degree than that within the bright bulb. This difference in temperature sends a column of mercury to the terminals in the connecting tube. When the sun passes behind a cloud or below the horizon, or, in other words, when the direct rays of the sun do not fall upon the bulb, the temperature in both is presumably the same, the mercury column remains below the terminals and the circuit remains open within the instrument. A recording de- vice is placed in the electric circuit at some convenient point in the ob- serving station. While the sun shines upon the black and bright bulbs, a characteristic line is drawn by a pen upon the revolving drum of the recording instrument. While the circuit is open a straight line is pro- duced. The clock which forms part of the recording device closes the circuit every minute of the day and night. In this manner we obtain a record of sunshine or no sunshine once every minute between sunrise and sunset. At the close of the day we may then add up the number of minutes of sunshine. With these figures and knowing the exact number of hours and minutes between sunrise and sunset, we may obtain the percentage of possible sunshine for each day. The hourly records for ten years show that there is a steady increase in the amount of sunshine in all months from sunrise to a maximum at about noon. The maximum hourly amount increases from 04 per cent in January to 81 per cent in September, and then again decreases to a minimum in January. The hourly distribution is shown in terms of percentages of the possible amount for each hour and month of the year in Table LXIII. The same distribution is graphically shown in Fig. 66, in which increase in the intensity of shading rei^resents an in- crease in the amount of sunshine. In Fig. 67 the average increase from hour to hour for the entire year is indicated by means of a single curve ; this shows a rapid and very uniform increase from sunrise to noon, and a similar decrease to sunset. This law of variation is common to all months of the year. The fact should not be overlooked that the amount of sun- shine recorded as described above is not a complement of the amount of cloudiness. Sunshine may be, and frequently is, recorded when the sky is, to a great extent, clouded. The instrumental record only indicates MARYLAND WEATHER SERVICE 241 TABLE LXIII.-AVERAGE HOFKLY DURATION OF SUNSHINE. Hours. » 2 a, >> G "3 3 +3 a > §' = •^ tn i < S »-5 HS < E» •^ ~' "1 4- 5 a. m 29 40 44 9 5-6 '• 33 41 .S2 42 42 38 49 23 6-7 " 33 31 44 ,54 38 .50 49 61 48 61 41 55 48 57 38 43 28 .. 33 7-8 " ... 9H 29 44 g^- 9 " m 4fi .51 64 .58 70 68 66 67 53 40 38 55 9-10 '• +9 m 63 68 65 75 77 73 74 64 ,52 54 64 10 IJ •• 6.S 68 74 67 70 72 71 70 78 79 79 SO 80 81 69 70 61 64 61 ; 70 11- Noon 65 : 72 Noon- 1 p. m fi+ 7"' 72 74 71 79 80 80 80 72 65 65 1 73 62 71 70 ■j-o 70 7-* 80 77 77 72 64 63 ; 71 2-3 •• !SX 66 66 68 66 75 75 72 74 68 57 56 , 67 3-4 " JO .ST 61 f,0 63 53 62 ,56 .56 49 70 60 67 59 66 54 70 .59 60 48 43 34 41 31 59 4-5 " 49 &- 6 " m 43 39 43 m 46 47 :{8 47 47 35 6- 7 " 36 83 25 m 30 27 46 19 7-8 ■• •• 18 31 27 30 » Mean dailj- number of hours of sunshine. . 4.9 6.4 6.8 7.9 7 7 9.2 9.2 8.4 8.4 6.7 5.0 4.9' 7.1 possible numlier 9.8 10.7 12.(1 13.2 14.314.E 14.6 13.7 12.511.2 10.1 9.512.2 Percentage of possible numV)er .50 .59 57 eo .54 02 63 61 67 60 M 51 58 Table LXIII shows the average duration of sunshine for each hour from sunrise to sunset, expressed in percentage of the possible amount of sunshine. The values are based on the continuous record of a self-registering thermo- metric sunshine recorder during the ten years from 1894 to 1903. The average daily duration is also given in hours and tenths and in percentage of the possible number of hours. Fig. M. — Mean Hourly Sunshine. The diagram show.s the mean hourly sunshine during each hour of the day and mouth of the year, expressed as a percentage of the liighest possible amount for the season. It is based on the ten years' record of a self-registering thermometric sunsliine recorder. The dotted lines S. R. and S. S. show the time of sunrise and sunset, respectively. The heaviest shading shows the time of occurrence of the higliest percentage of sunshine. The curved lines mark intervals of 10 per cent in the amount of sunshine. 242 THE CLIMATE OF BALTIMORE U'hetlier the face of the sun is or is not obscured at the moment of record- ing. It is only approximately an index of cloudiness. There is, in all seasons of the year, an almndance of sunshine at this station. The amount varies considerably in different months, but in all months the average is above 50 per cent of the possible amount. Janu- ary and December have the smallest amount in actual number of hours Fig. 67. — Mean Hourly Sunshine for the Year. The diagram is based on the ten years" record of a self-registering thermometric sun- shine recorder. The figures to the right and left of the diagram show the amount of sunshine, expressed as percentages of the highest possible amount. as well as in percentage of the possible amount. The amount increases from 4.8 hours in December to a maximum of 9.2 hours in June, per day. September, with but 8.1 hours of sunshine, has a higher per- centage than June, the value for the latter being 62 per cent, and the former 65 per cent. The average monthly and annual amounts of sunshine are indicated by the following- frg-ures : MARYLAND WEATHER SERVICE 243 Average Daily Sunshine. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year Average in hours. . 4.9 " percen- tage of possible amount 50 6.4 69 6.8 57 7.9 60 7.7 54 9.2 62 9.1 62 8.6 63 8.1 65 6.8 60 5.5 61 4.8 50 7.2 58 The sunshine of an}^ given month may vary greatly, however, from that indicated by the average figures given above. In the following table the months of the period from 1893 to 1903, during which the greatest and least amount of sunshine prevailed, are indicated, together with the monthh' ranges. The years in which these amounts were recorded may be found by consulting Tables LXIY and LXV. TABLE LXIV.— AVERAGE NUMBER OF HOURS OF SUNSHINE. (By months and years.) Year. Jan. Feb. Mar. Apr. May June July Aug-. Sept. Oct. Nov. Dec. Ann'l 1893 .... 1894 5.7 189.5 1 5.2 1S96 4.0 1M97 3.7 1898 5.5 1.^99 6.4 1900 i 5.2 1901 ! 4.4 1902 4.6 1903 4.6 Average 4.9 6.2 8.4 4.8 4.3 7.1 7.8 5.8 7.0 6.3 6.3 7.9 8.4 6.2 6.0 6.8 8.8 5.7 6.1 6.4 5.9 6.4 , 6.8 I 9.0 6.2 7.1 8.7 11.0 5.5 7.4 7.4 7.9 4.6 8.2 9.1 5.8 8.1 5.4 8.8 9.7 7.7 11.8 9.0 7.3 12.4 9.5 10.0 9.8 6.1 9.2 12.4 10.1 7.1 6.0 8.8 10.3 9.3 8.0 8.9 10.8 9.1 9.0 9.9 11.0 6.5 7.2 9.7 8.7 6.6 9.2 6.7 9.2 9.9 7.3 8.0 8.4 7.6 6.9 8.6 7.7 7.9 6.7 9.1 8.1 7.0 6.5 8.0 5.0 5.8 7.5 7.0 4.8 8.7 7.3 6.1 5.8 7.0 4.5 2.9 5.6 5.9 10.0 4.0 4.9 4.3 5.6 5.5 6.2 3.4 3.5 5.1 5.7 4.8 4.8 4.8 4.6 5.7 4.8 8.2 8.0 5.5 6.2 8.0 8.1 6.8 fi.8 6.9 7.0 TABLE LXV.— PERCENTAGE OF POSSIBLE SUNSHINE. (By months and years.) Year. 1894. 1H95. 1896. 1H97. iMOtl. 1S99. 1900. 1901. 1902. 1903. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Ann'l Average 60169 67 60 6462 62 63 66 60 64 17 244 THE CLIMATE OF BALTIMORE HIGHEST AND LOWEST MEAN MONTHLY SUNSHINE. (In percentage of the possible amount.) p ti C ^ e c >, to 4a +5 > d § » as P. cS a 3 P u o » 1.® >-> ^ S 100 " 208 1895 to Auk. 1. 189G 136 " 173 1896 to Apr. 30, 1902 82 " 185 1902 to Dec. 1903 117 " 220 'i.yi THE CLIMATE OF BALTIMORE The exposure of the anemometer was very satisfactory during the entire period, excepting from 1896 to 1902, when neighboring buildings obstructed the free movement of the atmosphere over the station. The elevation of the anemometer above sea level was approximately the same from 1871 to 1889, namely, between 90 feet and 100 feet; from 1891 to 1904, with the exception of September, 1895, to July, 1896, the sea- level elevation was increased by approximately 100 feet. Changes in elevation above sea level affect the velocity of movement of the atmosphere no less than changes in elevation above ground. The abrupt increase in the velocity shown from 1890 to 1891 is doubtless due to the change in the sea-level elevation of the anemometer. Since 1893 a continuous record of wind direction has been maintained without interruption excepting for a few hours at a time when difficulty was experienced with the recording instrument. The hourly changes in wind direction discussed in the following pages are based upon the ten- years' record from 1893 to 1902, unless otherwise stated. Average Hourly Wind Movement. The recorded hourly velocities for the twenty-year period from 1881 to 1900 have been reduced to average hourly values in order to determine the periodic variations in velocity during the day. The results are shown in Table LXX, and graphically in Figs. 70 and 71. In Fig. 70 the hourly changes in velocity are given for the months of January, April, July and October, and the average for the entire year. The curves for all months are similar in form. There is a minimum velocity in all months just before sunrise. The velocity rises rapidly to a maximum between two or three in the afternoon, which it maintains approximately for two or three hours, then decreases rapidly to 8 p. m. or 9 p. m., and more slowly to the minimum for the day just before sunrise. The same hourly variation is shown for all months of the year in another manner in Fig. 71. The influence of the diurnal variations in temperature upon the coincident variation in wind velocity is strikingly exhibited in the table and diagrams; the increase in velocity accompanies the increase in temperature throughout the course. The time of maximum rate of MARYLAND WEATHER SERVICE increase and decrease in velocity is coincident with the time of maximum rate of change in temperature, the most rapid increase occurring between 1234.56789 10 II tn ftlttS - / \ /« / \ \ ^ I 1 ; 1 __^_ J Ju Fu Ken lor May June Juljf Au^. Stu Oct No Dec Annual Vabiations or Wind Velocity Fig. 70. — Hourly and Annual Variations of Wind Velocity. Expressed In miles and tenths of miles per hour for the months of January, April, .Tuly and October, and for the entire year. 8 a. m. and 10 a. m.. and the most rapid decrease between 6 p. m. and 8 p. m. In the annual fluctuation in velocity, however, a similar relationship does not exist. On tlie contrary, there is almost a direct inversion of 354 THE CLIMATE OF BALTIMORE TABLE LXX.-AVERAGE HOURLY WIXD MOVEMEXT. (In miles and tenths.] 4 p 'u < ^ S 2 3 >. ■^ bib 3 <5 a. X O o 2; ® 3 <5 Midn't to 1 a. m. . . 5.5 5.9 6.0 5.3 4.5 4.2 4.0 3.5 4.0 4.6 4.8 5.1 4.8 •J •■ 5.0 5.8 5.9 5.2 5.0 4.1 4.0 3.5 3.9 4.5 4.8 5.0 4.8 3 " .... 5.3 5.7 5.8 5.1 4.3 3.6 3.9 3.5 3.9 4.4 4.8 5.0 4.6 i " .... 5.3 5.8 5.8 5.0 4.1 3.8 3.8 3.6 3.9 4.6 4.9 5.0 4.6 5 " .... 5.2 5.8 6.0 4.8 4.3 3.9 3.7 3.7 3.9 4.6 4.8 4.9 4.6 6 " .... 5.1 5.b 5.9 4.8 4.3 4.0 3.8 3.7 4.0 4.6 4.7 4.9 4.6 7 " .... 5.4 5.9 6.0 5.3 5.0 4.8 4.3 3.9 4.3 4.7 4.8 5.0 4.9 8 " . . 5.6 6.2 7.0 6.4 5.7 5.7 5.3 4.7 4.9 5.4 5.2 5.2 .'1.6 9 " .... 6.1 7.0 ■ 8.5 7.6 6.6 6.3 6.0 5.5 5.8 6.2 6.0 5.7 6.4 10 " .... 6.9 7.8 9.0 8.2 7.1 6.9 6.5 6.0 6.5 7.3 7.1 6.6 7.3 11 " .... 7.5 8.3 9.3 8.7 7.7 7.4 6.8 6.5 6.8 7.7 7.7 7.3 7.6 Noon 7.8 8.7 9.6 9.0 8.2 7.7 7.2 6.8 7.2 7.9 8.1 7.7 8.0 1 p. m — 8.1 9.1 9.9 9.4 8.6 8.1 7.8 7.1 7.5 8.2 8.4 7.S 8.3 O " 8.3 9.1 9.9 9.7 8.8 8.2 8.0 7.3 7.6 8.2 8.4 8.0 8.4 3 " .. . 8.2 9.1 10.0 9.7 8.9 8.3 8.2 7.5 7.6 8.2 8.3 7.9 8..'. 4 " .... 7.7 8.8 9.9 9.5 8.6 8.3 8.0 7.6 7.6 8.0 7.9 V.5 8.3 5 " .... 7.1 8.3 9.6 9.1 8.4 8.1 7.6 7.2 7.0 7.1 6.9 6.6 V.8 6 '• .... 6.2 7.3 8.4 8.3 7.6 7.3 7.1 6.4 6.0 5.7 6.9 6.8 6.8 7 " .... 5.8 6.6 7.3 6.8 6.4 6.1 6.0 5.2 4.8 4.9 5.4 6.6 5.9 8 " . 5.6 6.3 6.6 5.8 6.5 5.1 4.7 4.2 4.4 4.9 6.3 5.4 5.3 9 " 5.5 6.0 6.5 5.7 5.2 4.6 4.4 3.9 4.4 4.8 5.1 5.3 5. J 10 " .... 5.5 5.9 6.3 5.5 4.8 4.4 4.2 3.7 4.3 4.7 5.0 5.2 5.0 11 " ... 5.4 6.0 6.3 5.4 4.8 4.3 4.0 3.7 4.3 4.7 6.0 5.2 4.9 Midn't. ... 5.5 5.9 6.1 5.3 4.8 4.2 4.0 3.6 4.1 4.6 4.9 5.1 4.8 Means 6.3 7.0 7.6 6.9 6.2 5.8 5.6 5.1 5.4 5.8 6.0 5.9 6.1 Table LXX is based on the continuous record of a self-registering anemom- eter during the 20 years from 1881 to 1900. t 2 3 4 5 6 7 8 9 Fig. 71. — Average Hourly Variations in Wind Velocity. The heaviest shading shows the time of occurrence of the highest average wind veloci- ties for the day. The curved lines mark intervals of half a mile in the average velocity. The dotted lines marked S.R. and S.S. show the time of sunrise and sunset, respectively. The diagram is based on hourly values for a period of 20 years. MARYLAXD WEATHER SERVICE 255 the relation existing between temperature and wind velocitA'. The light- est winds occur in the months of greatest heat, while the highest veloci- ties occur in March, with a slight secondary increase in October and November (see Fig. 71). The annual fluctuations are due to the varia- tions in cyclonic activity at different seasons of the 3'ear, The highest average hourly wind velocities occur between 2 p. m. and 3 p. m. in the month of March, when they attain an average velocity of 10 miles per hour. The lowest velocities occur in the early morning hours of June, July and August, when the average falls to about 3.5 miles per hour. This law of increase and decrease is remarkably constant through- out the year and is recognizable at any time when not interrupted by the presence of a well-developed cyclonic or anti-cyclonic disturbance. Average Daily, and Total Monthly Wixd Movement. In Table LXXI the total monthly wind movement for each month of the year from 1873 to 1903 is shown, together with the average daily movement for each year during the same period. As the elevation of the anemometer was changed several times during this period, it is essen- tial to bear in mind the fact in discussing the variations in wind veloci- ties as shown in Table LXXI. Xo attempt has been made to reduce the records to a single elevation; the changes in elevation are distinctly traceable in the monthly and daily values for the wind movement. In- ferences as to fluctuations in the annual velocity should be made with caution. The average daily wind movement is approximately 145 miles for the entire year. The velocity varies from a minimum of 122 miles in August to a maximum of 175 miles per day in March. The following figures represent the average daily wind movement for each month, as derived from hourly observations from 1873 to 1902, a period of 30 years : AVERAGE DAILY WIND MOVEMENT. Jan. Feb. Mar. 175 Apr. May June July Aug. 149 142 134 122 Sept. 129 Oct. 137 Nov. 143 Dec. Year Miles 146 162 160 142 1 14S 256 THE CLIMATE OF BALTIMORE TABLE LXXL— TOTAL MONTHLY AND AVERAGE DAILY WIND MOVEMENT. Year. © PR 1 < e >-> 3 1-5 bo 3 < ® Oi Q c 5 S i m 1873 3665 4848 3571 4510 3764 5335 5.33t> 3R.3S 4180 4136 4636 3671 4021 633S 5S11 3977 5766 4854 5124 4158 6295 6519 4466 4753 4615 4876 5847 5893 5495 5.336 4964 6033 7326 6904 5872 7070 8038 4.33S 4226 5040 4340 E007 4594 4938 5330 5.514 5543 5435 175 513S 5565 4891 4718 4065 4870 5149 6398 5045 4423 4583 5139 4.3a3 4085 4963 4736 5455 3997 44.30 64.33 6604 6351 6308 6056 3974 4S.39 4077 3574 5-387 4963 6449 4936 4815 5303 4983 166 4389 4670 4990 4495 4016 4344 4350 4764 3841 4310 4993 4483 3917 4335 3783 4041 419.S 3970 4838 5987 65S8 5S80 5435 5669 4196 4235 36.50 4101 4359 5933 6336 4397 4444 5004 4616 149 4136 3966 3.f;73 3680 4032 3353 4764 3775 3937 3610 3629 3253 3774 4110 3576 3531 3443 3184 3570 5140 3386 4008 4867 .5063 4731 4857 3453 3395 3425 3362 3176 3383 4754 4549 3780 3881 3940 3867 139 4103 3802 3687 41^0 4317 4600 .3461 3939 .3456 33.33 4213 3905 3926 3796 4093 4626 4669 3975 6125 5484 5360 5,S0S 5645 3589 41.36 4080 .3415 3585 3074 6197 6339 3878 4481 4389 4349 137 4256 3665 3267 4079 4715 4617 4715 3640 4135 3542 4061 3934 4199 4708 38.33 3,S31 3342 5937 6567 5422 6213 5373 3133 3463 3559 3197 3732 4299 4781 4619 4062 4474 4316 4384 143 4181 45S6 3959 4014 43S5 5765 3905 4661 3795 4439 3634 3928 5047 4038 4176 4376 3714 4440 5766 6567 6798 5054 6033 3343 3526 4180 4036 3450 3699 5940 6330 4359 4458 4505 4407 143 4254 4543 3780 4474 3994 4575 4355 4318 4067 4a36 4166 4119 4294 4338 4331 4205 4333 3863 5033 5796 5837 5.364 5643 5031 3769 ■3913 3780 3729 3931 6014 5483 4240 4425 4598 4421 140 1874 4529 4646 41,SS 3793 J 34:55 , .3310 4433 4341 3769 4479 1 4083 , 3907 4479 4189 3T17 4326 4518 4026 4644 3949 3876 .3906 3,867 ' 3073 4733 4213 3.523 4076 4184 3861 3655 4074 3237 4354 , 3740 4073 3803 i 3.508 3680 4133 4073 1 3653 3949 3.SO6 3690 3900 3948 3945 3281 3506 3631 5112 5502 i 4399 5635 1 4451 : 4536 4867 4957 5233 4654 4960 3880 4561 ; 4.541 4890 4955 ! 5397 3133 3440 3551 3908 3754 3765 3130 34S9 3.3.55 3918 149 1876 124 1876 1877 147 131 1878 1879 150 143 1880 3443 ■IR.HR 143 1881 3654 4.335 3471 4609 5090 4789 4600 4551 3938 4191 3740 6393 6510 5413 6030 6399 4559 4097 3840 3933 3793 3874 6410 4135 4536 4834 4498 145 4362 3567 4054 4273 4439 4831 4407 :^iH8 3.S08 3661 4385 6410 6628 5553 7090 7297 3840 3668 3946 4547 4036 4839 6073 4070 4365 5143 4636 163 134 1883 1883 1884 1885 133 137 135 141 1886 139 1887 139 1888 1889 138 143 1890 137 1891 ... 1893 165 191 1893 193 1894 176 1895..... 1896 186 165 1897 134 1898 139 1899 134 1900 1901 3715 3397 3363 3689 3304 3163 123 139 1903 5698 1 fiOQK 4606 4731 3696 3870 3807 3791 133 165 1903 5115 4334 4189 4230 4351 143 5007 4131 4079 4361 4154 134 180 Average 1873-1883 1883-1893. . . . 1893-1903 1873-1903 Average ) daily V . . 1873-1903 \ 139 145 151 145 Table LXVIII shows the total monthly and average daily wind movement for 31 years, from 1873 to 1903; also the average daily movement for the entire 30 years ending 1902. The figures are based on the continuous record of a self-registering Robinson anemometer. The average daily movement for an entire year has been as low as 124 miles, as in 1875, and as high as 150 miles, as in 1878, confining our choice of limiting values to the period from 1873 to 1890, during which the elevation of the anemometer remained practically unchanged. a es C O ••iia « cj ■* ce c5 ■* eo S'l >o a) i-i!M>-i i-i i-ii5>-ii-ira 1-1 i-i b-oa ■I^Jl^" -l^^5"r® 5'o3=,5 »°5®5 S5h?;0< S5aK,-^5ai-;-— i ra'^cc'*:^ C'iC''!CJT!ec jjcccc*: • — CSIOtO 0iX^}»O3D h »5 cc »o cc p; ci cc -^ »o c-j t I- M 05 Xi m t- ffi •* rt t- ^3"*"^ *OOlOCC zzzzz ZZZZ Zx.x>'>^ '^zz *®A?2'^T*' J??!r ■M '-*< tc ;c Xi iC tC rj tC'f** -* t--H.-^ 808 Ztc ZZ iC ip — « ^j* :C' xzzzi^i rioi-io —rtc; — — • -^-t^x- z'^'^^z :;S;fa; z:z''z »> SI re » cc M ?5 ?J « cc X » lO ?i ^ X « ?'t X X ?i as o -H 0^1 ~i"ds 15 ?j ?i « ?} ei ■* •>* M » r! rj II s! ?i is cc cc 01 2-8S o»o-^rrx cc-^ooii rcio-H^^t— CT ^ ?) I? 01 ?1 re ?? ^ - H g; ZZZMZ x-*oiO'* -^csaj-^o -*ooo-* «T!re->»-i, : o o :£■ ^H C5 ^ ?-l 2-iZ ot— rerex o»o-*gi-^ g»-H —I ?! ?;?!?! —I C5 s :r -^ o re»ret ■ X c^re 03i^ Zx-yiZ- =^?^Z^Z Z>: z H>>:-C>: :kZ; :o Oi-:=- 682 ?CSl^^^^g5 lO^-tXC'lO iOOSlXl- CIdOCSX »-i?i»oxre IMCC: C-? re iM re ?} ^zzS= ZIS^M^; Z''2;^!2; ''ZOJOQ occcooo CO-^Ot- Xt--*»OX ^CCiOC'tO O^lOO COt— re gi c^ ?! 1-— 1-*0 CSIT!: I- 10 ^ X := ^H ?!?!: -S«-^ g^^-^ Z m'^Z s- "530003 o<-rerere cjoosjo reciJire^ SI -H ?!?>?! -^ SI SI SI re SI re SI SI re s5?;ss? 8-96 ■*i-05«re o>03:iore 551-1 OS -«SISIg Z "^ VI >: Z Z Z X Z Z Z 7: x z z '' •z''z' '^ZZz z ZZZZZ SI 1 - o o re X3J'* — Si siore-^ •"^ '"*' "*' ** i^ *^ p^ 53 ^^ C2 *^ ^ ^ "^ ^ "'^ !>■ ^ ^ ^ zzz zzzzz zs^zz" Z^^Z" '''^M'' z zzz j'«?«-n«? MSI SI SI SI SI sire I- O 10 ■* = 10 lO 1- s» re •OS! SI re SI SI sj re :,-63 re SJ s> »o SI ^reo-*x oS"U5i- —X SI 35:c x-^sj:cs ;-*siiO ■/.zz''z J § ?; i.- ?, ?, ZZZZZ Z ZZZ '^'^'^'f^'^ '^m'^'^ zz ?.?iii2;-' zi^zz JosisiS?! SSwS-^ SreSSeS SSeSSil t'68 - => s 5 y ? rsii iO¥'- X : ^i^^ T681 81,81 S68I 8881 poer 8681 6i8I 9681 — « -^?» ^c^i ^TO gi e^t 0^ ?t o^ _c^_»r^Mcc : Tool JSKJ:^^ ?=^W==^ ^S^5=^ ^^W^ ^ ^ bo O C tn " •" 3 -o o 5 ^ •r; w ^ M "S CI o W) '" 'H T-l Ci +J <" bp 10 » t>> a 111 ^ a.9 ^a »J3 5 a ^ 5 0) a 03 ai-i *J ►-1 ^ "O X ^ iJ 1-5 fe 3 ■— I a) ra o rt 358 THE CLIilATE OF BALTIMORE Maximum Wind Velocities. In the preceding paragraphs the total wind movement over the station for an entire month, and the average hourly and average daily move- ment were alone considered. In Table LXXII a record will be found of the highest velocity of the wind attained in any 5-minnte period during each month and year from 1875 to 1903, together with the accompanying direction of the wind and the date of occurrence. In determining the maximum wind velocity for the day, the sheet contain- 60 ■ - - 40 - - - - - - - - - - - - 20 Fig. 72. — The Frequency of Storm Winds. The diagram shows the variations in the annual frequency of winds exceeding 25 miles per hour. ino^ the continuous record of the anemometer is examined and the five- minute interval selected during which the velocity is greatest. The number of miles or fractions of a mile registered during this 5-minute period is then multiplied by 12 in order to obtain the rate of movement per hour, or what is usually termed the hourly velocity of the wind. All of the daily records since 1875 have been carefully examined and the highest velocity recorded during each month selected and entered in Table LXXII, at the same time noting the date of occurrence and the direction of the Avind during the selected 5-minute period. An examina- tion of the table shows that high winds are not confined to any particular season of the year, but have occurred in all months. The high winds of MARYLAXD WEATHER SERVICE 259 the winter months occur in connection with the well-defined cyclonic disturbances, while the high velocities of the summer months accom- pany the thunderstorms, or the tornado, in the rare instances of its occurrence in this vicinity. The annual variations of the maximum velocity are shown in Fig. T2. The highest velocity of the wind recorded at the Baltimore Station of the U. S. Weather Bureau since 18T5 occurred during the storm of July 20, 1902, when the wind blew at the rate of TO miles per hour for five minutes. Further particulars of this storm, which was one of the most destructive ever visiting this vicinity, will be found in a later section of this report. Selecting the highest recorded velocities, in miles per hour, for each month of the year, we have the following comparative figures : HIGHEST MONTHLY VELOCITIES. > 6 o « Z. Q 48 5-t S E 1891 1898 28 4 28 30 Highest vel.... 48 45 50 , fiO 4.S i-2 70 45 38 45 48 , 54 70 Direction W NW S NW | W SW W SW NW SW S E W Year 1894 1893 1896 1879 1893 i 1893 1902 ' 1888 I 1892 1878 \ 1891 | 1898 1902 Day .30 19 19 3 23 I 27 20 8 ■ 26 | Av.vel.of ma.\.- 29 , 30 30 i 30 26 1 26 28 24 24 27 28 30 28 The average of all maximum velocities during the 28 years from 1875 to 1902, as shown in the last line of the above table, indicates a remark- ably uniform value for this factor, throughout the year. The highest monthly average velocity (30) differs from the lowest (24) by only 6 miles. The lowest velocities occur in August and September, and the highest in February, March, April and December. The fact that the September records show the lowest average velocities for storm winds is significant in view of the popular association of the so-called " Equi- noctial " storms with this month. As already stated, wind velocities are generally expressed in terms of the rate per hour based upon the actual velocity during a five-minute period. By basing the rate per hour upon the duration of the mile made in the shortest time, we obtain what is officially designated as the extreme velocity. By this method we are more liable to obtain the 18 260 THE CLIMATE OF BALTIMORE velocity in brief gusts of wind, velocities which are lost when the hourly rate is based upon the movement during a period of five minutes. As much of the destruction due to high winds is wrought during these brief gusts, or squalls, the extreme velocity is a factor of great importance. It is, in nearly all cases, higher than the maximum; it cannot be lower. There is no fixed relation between the two velocities ; it may be of inter- est, however, to show to what extent they have differed from one another. Basing our inquiry upon the official record of the monthly maximum and extreme velocities during the period from 1888 to 1903, we have the following comparative figures: RELATION BETWEEN MAXIMUM AND EXTREME VELOCITIES. (In miles per hour.) Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Deo. Maximum 48 13 45 55 50 60 42 50 4.3 48 42 50 70 75 45 52 38 50 12 43 50 8 48 60 12 54 60 10 10 8 5 8 5 7 6 This relationship may be expressed b}' another method. In place of selecting the highest maximum and highest extreme velocities for each month, we may examine all cases of high winds occurring in a stated time and note the difference between the maximum and extreme veloci- ties. This has been done for a period of three years with the following result : DIFFERENCES BETWEEN MAXIMUM AND EXTREME VELOCITIES. (In miles per hour.) Jan. Feb. Mar. Apr. May .lune July Aug. Sept. Oct. Nov. Dec. Year Average diff 4.5 Greatest " Least " No. of cases. 4.5 10 1 29 4.3 4.3 3.4 5 3 11 3.6 12 14 7.8 3.9 17 I 10 I 1 1 15 7 4 8 16 1 13 3.6 10 1 17 5.0 4.2 4.5 17 195 The highest wind velocities generally occur in connection with a northwest wind in all months of the year. These winds usually accom- pany a rising barometer and occur a short time after the shift in tlie MARYLAND AVEATHER SERVICE 261 wind which follows the turn in the barometer. While northwest is the usual direction of the storm wind, all directions of the compass are rep- resented. In Table LXXII there are 348 records of high winds covering a period of 29 years; placing these in the order of frequency of the directions from which they came, we have the following relative positions for the entire year : RELATIVE FREQUENCY OF HIGH WINDS. Direction of wind NW W SW NE N SE E S Percentage of frequency 41 20 12 8 1 4 4 4 The same order of frequency obtains practically in all months of the year. In nearly three-fourths of all instances of storm winds, the direc- tion is from some point between southwest and northwest. In only 12 per cent of instances is the direction from some point between east and south. High winds from the north or from the east are of compara- tive] v rare occurrence at Baltimore. Frequency and Duration of Stated Wind Velocities. The hourly wind velocities during a period of five years (namely, from 1893-96 and 1903) were tabulated into groups in order to deter- mine the relative frequency of stated velocities. The result is shown in the following table, in which the frequencies are expressed in terms of percentages of the total number of hours in each month : frequency of stated wind velocities. (In percentage of possible number of hours per month.) Miles per hour.. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. 41.8 Nov. Dec. An'l 5 37.6 31.2 33.7 28.8 35.3 41.5 44.3 49.0 45.3 39.3 40.7 39.1 ti-10 35.7 .35.2 3(i.4 40.7 4.3.0 44.4 43.1 39.7 38.5 .37.6 35.6 .35.3 38.8 11-15 15.8 15.7 16.2 19.9 15.6 12.2 10.8 9.1 13.2 13.8 16.1 15.7 14.5 16-20 6.4 , 9.9 8.3 . 8.2 4.9 1.8 1.3 1.3 2.5 4.6 6.5 6.3 5.2 21-25 2.8 1 4.9 3.5 1 1.9 O.T 0.1 0.1 0.3 0.4 1.7 2 2 1.5 1.6 26-30 1.2 2.1 1.4 0.5 0.4 0.1 0.2 0.2 0.3 0.2 0.4 0.6 .31-40 0.3 : 0.9 0.3 0.06 0.1 0.06 0.1 0.2 0.1« 0.1 41-50 0.03' .... 0.03 0.0 262 THE CLIMATE OF BALTIMORE Winds of 10 miles per hour and under prevail during about 78 per cent of the total number of hours of the year; winds of 11 miles to 20 miles during less than 20 per cent. Hence the total duration of veloci- ties exceeding 20 miles per hour is only about 2.3 per cent of the entire year, or about eight and a third days. Storm winds, or winds exceeding 25 miles per hour, prevail during about 62 hours in an average year. Average Duration of Storm Winds. It will be seen from the statements in the preceding paragraph that winds having a velocity exceeding 25 miles per hour are of brief duration. The duration decreases rapidly with increase in wind velocity. The rate of decrease may be readily judged from the figures in the above table representing the annual relative frequency of stated velocities. Basing our calculations upon the same period of five years employed in deter- mining the relative frequency of stated velocities, v^e obtain some inter- esting figures defining the average duration of storm winds. AVERAGE DURATION OF STORM WINDS. (In hours and minutes.) 1-5 i 03 b < >> 03 0) c 3 1-5 >> be < P. o O > o I? 6 OS >> Total annual duration . . . 23.36 43.35 28.25 12.25 5.35 0.26 0.60 4.10 3.40 11.10 9.50 13.06 1.56.40 Averag-e frequency.. 6.2 9.2 7.6 5.0 4.4 2.0 2.6 1.4 2.6 4.2 4.8 4.4 54.4 Duration per storm 3.50 4.40 3.40 2.30 1.20 0.12 0.20 3.00 1.25 2.40 2.00 3.00 2.50 Greatest duration — 19.00 46.00 23.00 16.00 7.00 .36 .30 18.30 7.00 9.10 6.00 19.00 46.00 In the winter and early spring months a storm wind usually continues from three to four hours. The duration rapidly diminishes on the approach of summer, reaching a minimum in June and July, when the average duration is only a few minutes. The high winds of summer usually occur in connection with thunder squalls of brief duration, while those of winter, spring and fall accompany the passage of well- defined cyclonic storms. The comparatively long duration of August storm winds in the above table is due entirely to the severe gulf storm of August 28-29, 1893, during which the wind blew a gale for many hours. MARYLAND WEATHER SERVICE 263 a duration which would be considered long even for a winter gale of the severest type. Xeglecting this storm, the August average duration for the remaining four years is about 35 minutes. During the passage of the Gulf storm of February 7-10, 1895, the wind blew at Baltimore with a velocity exceeding 25 miles per hour for about 46 consecutive hours. It then fell below the storm velocity for about 12 hours and again went above 25 miles per hour for another period of 12 hours. It is one of the longest storm periods on record at Baltimore. The storm originated in the Gulf of Mexico on the 6th, and moved rapidly eastward and northward along the Atlantic coast from Florida to the Gulf of St. Lawrence, the center passing just east- ward of Baltimore on the 8th, with a maximum velocity of 42 miles per hour from the west. The barometric gradient between the center of the storm and the center of the area of high pressure to the west and north- west was very great throughout its course, amounting at one time to about two and a half inches. The extreme velocity of 50 miles per hour was reached on the 8th at about noon. As the summer high winds occur mostly in connection with thunder- storms, their time of greatest frequency, and hence greatest probability, is from 3 p. m. to 4 p. m. The winter, spring and fall storm winds, ' accompanying cyclonic disturbances which occur at any hour of the day, also have a well-marked tendency to fall within the early afternoon hours. This may easily be explained by supposing that the cyclonic winds are augmented at these hours to a maximum extent by the diurnal wind movement. Gales. A gale is technically defined by a wind velocity of 40 miles, or over, per hour. Such winds have been recorded on 42 occasions at the Balti- more office of the U. S. Weather Bureau since 1873. They have been of comparatively great frequency in some years, notably in 1893, which is credited with nine; there were seven in 1903. In half the years since 1873, none were recorded. The highest velocity registered in the years from 1880 to 1887 was 39 miles. As is the case with most high winds, 364 THE CLIMATE OF BALTIMORE TABLE LXXIII.— SUMMARY OF WIND VELOCITIES. (1873-1902.) January . . February. March April May June July Aug'ust — September. October November. December . Year 6.1 8.0 Means. 6.0 6.7 7.3 6.9 6.3 6.9 5.6 5.1 5.4 5.7 6.0 5.9 10.8 10.8 9.3 8.9 7.8 7.4 7.0 7.1 8.2 9.1 8.1 1893 1891 1891 1893 1895 3.7 3.9 5.3 5.0 4.9 4.5 4.5 3.9 4.4 4.1 4.4 4.0 1877 1877 1875 1900 1901 1899 1897 1900 1901 1896 1875 5.1 1900 28 Maxima. c3 sj 1879 1893 1893 1903 1888 1893 1878 1891 70 1903 14 1881 1883 1901 1885 1891 1900 1881 1884 1886 1890 1900 1883 1881 1883 1898 1880 Storm winds.* as 2-5 sa 4.4 10 5.5 12 6.7 13 5.3 13 3.9 9 3.3 5 1.8 4 1.2 4 1.4 5 3.0 7 3.7 9 3.9 8 42.0 70 1878 1895 1881 1880 (1878 '11893 11877 11879 11878 -^1896 (1901 1887 (1889 11896 1894 1886 j 1885 11887 * Winds exceeding 25 miles per hour. gales blow mostly from the northwest or west. Of the 42 instances referred to above, the relative frequency of the points of the compass from which they blew is as follows : DIRECTION OP THE WIND IN GALES. NW W S SW SE NE N E Total Number of gales. . . 15 13 4 3 2 3 3 1 42 The distribution of gales by months shows that they have been most frequent in February. The month of " equinoctial storms " is the only month without a gale to its credit in 30 years. FREQUENCY OF GALES IN 30 YEARS. Jan. Feb. Mar. 3 Apr. ; May June July Aug. Sept. Oct. j Nov. Dec. Year 2 10 2 2 3 4 2 3 5 6 43 MAKYLAN'D WEATHER SERVICE 265 Prevailing Hourly Wind Directions. We have seen in preceding paragraphs that there is a well-defined diurnal fluctuation in the velocity of the wind. Without a close obser- vation of diurnal changes of direction in the locality of Baltimore, a well- marked periodicity would scarcely be suspected. Such is the fact, how- ever, as demonstrated by a reduction of the hourly observations for a period of ten years. The results are shown statistically in Table LXXIY and graphically in Plate XI, Fig. 73. A well-defined diurnal period was TABLE LXXIV.-PREVAILING HOURLY WIND DIRECTION. Hours a S3 1-5 i i a. >> 03 c >. sib 3 *> u o 1 pc ^ < s ►-5 i-s < CQ O (z; P "^ lA.M NVT w NW W w SW sw sw sw N w W W o sw w NW w w SW sw sw NW N w w SW 3 sw w NAV NW SW sw sw N NW NW w NW NW 4 w NW W NW NW sw sw NW" N NAV NW w NAV 5 w W W W NAV sw sw NW N N NW w AV 6 w W W W N sw sw NW N NW N W W NW NW W W NE sw sw N N N NW AV NW 8 W W W E N sw sw N N JN W W AV 9 W W w SAV NAV N N SW SW sw N sw sw SW SW NE E NW E N N W SAV SW 10 SW 11 w w E SE SR w sw N N E SW SW SW w w w w SE SR SE SE SE SE SE SE sw SE SE SE E SE E SE w sw NAV NAV SE 1 P.M SE 2 w NW SE SE SR SE sw SE SR SE w A\" SE 3 ... w w W NW SE SE SE SK SE SR SE SR SE SR SE SE SE SR SE SE sw SE SW AV SE 4 SE 6 w W SE SR SE SE SE SE SR SR SE AV SE fi w W SE SE SE SE SW SE SR SK SR AV SE "r NW W W W E W SE SE SE SR SE SE SW sw SE SR SE SE SE SE SE w NAV AV SE S SE 9 w NW NW SE SR S sw SW SW E w NW SW 10 \w W NW SE SR SW sw SW SW K w AV sw 11 NW W NW SR SR sw sw SW SW NW w NW sw Midnig-ht SW W NW W SW sw sw SW SW N N W sw Prevail, direction. . W W NW SE SE sw sw SE SE SE w W SE Table LXXIV, showing the prevailing direction of the wind at the hours stated in the first column, is based upon a record of ten years, extending from 1893 to 1902. not at first expected to be revealed by the average hourly values which included all the observations of the year, or e\'en all of any particular month. Hence the first attempt to detect a periodic movement was made by selecting days in the months of January, April, July and Octo- ber, during which the skies were prevailingly clear, and the wind move- ment was light. This was done with a view to eliminating the influence 266 THE CLIMATE OF BALTIMORE of neighboring .cyclonic disturbances. The result of such classification is shown in Fig, 74 for January, the diagram being based on ten selected days during which the sunshine exceeded 90 per cent of the possible amount, and the wind movement was less than 100 miles. The wind direction observations were classified into morning and afternoon winds, the former class including the hours from midnight to noon, the latter from noon to midnight. A prevailing westerly wind during the morn- ing hours and an easterly wind during the afternoon hours was so clearly revealed in all months in these diagrams that the hourly obser- vations for each hour and for the entire period of ten years were tabu- lated and charted, with the result shown in Table LXXIV and Fig. 73. These tables and diagrams reveal some interesting and probably unsus- pected facts concerning the daily fluctuations in the wind direction at Baltimore. A well-defined diurnal periodicity appears in all seasons of the year when the local conditions are not influenced by the presence of cyclonic disturbances. This is quite as well marked on cold winter days as in the summer time. Even by employing all observations, the average of all conditions of the weather, this periodic movement is conspicuous excepting in the winter months of December, January and February, when the cyclonic winds almost completely mask the periodic movement. An examination of Fig. 73 shows a prevailing wind from some quarter between northwest and southwest at all hours between midnight and 11 a. m., with a very few exceptions when they are from the north. This is true for all months of the year. In January, February and December these westerly winds continue throughout the day. In all other months there is an abrupt change in the direction to the southeast about noon; a little earlier in March, April and October and a little later in July and November. The southeast wind then continues without interruption to an early evening or a night hour, when the direction returns quite as abruptly to the southwest or west. The hour of return to the morn- ing direction varies more than the change from the morning to the afternoon direction. The southeast returns to southwest in July as early as 6 p. m. ; in April and May as late as 11 p. m. The southeast or afternoon direction is maintained, accordingly, for a minimum period MARYLAND WEATHER SERVICE. VOLUME 2, PLATE XI. o z o y- "^ 1 "l >l < 1 >|r vl^^^-Sr^^^^r^ HH^ ^'-n^'-^- ^^^^r v?^ -?^^^^^^ i^ )(}()( k k k A A \ >{ ^r \ < 1 - -4-^^- ^s^ v^^ -^^ % Y "" ' ^ ""^ V*^ v'^ V'^ ^^ ri^-A- ^^ v*-v^ ^^ V- V- s r? .2 M a >5 ■- >» .a bij S a; ■— 2 -^ 71 OJ 2 3 MARYLAND WEATHER SERVICE 267 of 5 hours, as in July, to a maximum of 13 hours, as in April, May and October. For the year as a whole, the southwest wind changes to a southeast at noon, maintains this direction until 8 p. m., and then returns to the southwest. The southwest becomes a west or northwest wind from 1 a. m. to 8 a. m., and then southwest again from 9 a. m. to 11 a. m. These hourly changes are surprisingly uniform throughout the year when prevailing directions for a long period are considered, or on quiet days, for short periods of only a few days. Fig. 74. — Prevailing Morning and Afternoon Wind Directions in January. The heavy black lines Indicate the prevailing: winds during the hours from midnight to noon ; the light lines show the prevailing winds during the hours from noon to mid- night on selected days in January with a light wind and bright sunshine. In the figure based on the rougher grouping into morning and after- noon directions, the percentage of frequency of occurrence of the wind from each quarter is also shown. Fig. 74 indicates that even in mid- winter, represented by the month of January, the morning winds are distinctly west of the north and soutli line, and the afternoon winds mostly to the east. In Fig. 73, which is based on all observations during a period of ten years, the winds are westerly in January, February and December, both morning and afternoon, as stated above. A feature especially worthy of note is the abrupt change from southwest to south- 368 THE CLIMATE OF BALTIMORE east about midday. The change from northwest or west to southeast, and in the reverse order, is made without lingering in the south. A prevailing south wind is not revealed in the diagrams or table for even an hour in any month of the year. It is difficult to assign a satisfactory cause for this daily periodic movement in the vicinity of Baltimore. The first explanation which is suggested is that it is a land and sea breeze effect. The station is, how- ever, too far removed from a body of water sufficiently large to produce the effect, even at the season of the year when contrasts in temperature between land and water are strongest. The harbor presents a compara- tively small water area in Patapsco Eiver, which is in turn twelve to fifteen miles from Chesapeake Bay, while the station is fully a mile distant from the harbor. These facts of local conditions make it ex- tremely improbable that the winds are the effect of an interchange of air between land and water areas. The suggestion arises whether the fluctuations are an integral part of the diurnal cyclone described in the preceding section on pressure changes. To demonstrate this would require a similar discussion of the hourly changes in direction at many widely scattered stations, especially at points somewhat nearer the path of the center of the diurnal cyclone, and on both sides of the equator. Prevailing Monthly and Annual Directions. In view of what has been presented in the foregoing paragraphs con- cerning the hourly changes in the direction of the wind, it becomes obvious that the choice of hours of observation is an important matter in determining the prevailing monthly and annual direction of the wind at any given locality. Most systems of observations, before the days of continuously recording instruments, provided for three eye observa- tions : one about 7 in the morning, another about 3 in the afternoon, and the third at about 9 in the evening. This combination yields a very fair value for the average direction for the day. The prevailing direc- tions based on two daily observations from 1893 to 1903 are placed alongside of the prevailing directions computed from three daily obser- vations and from hourly observations covering the same period. The Fig. 75. — Relative Frequency of Prevailing "Wind Directions. The diagram shows the relative froquency of the prevailing directions of the wind in the months of January, April, July and October, and in the year. For example, for the month of July, the prevailinK directions during a period of ten years were confined to southwest and southeast winds ; in January, the prevailing winds were always westerly during the same period, etc. 270 THE CLIMATE OF BALTIMORE WARM Year. NORMAL YEAR. Colo Yt'- 1893-4 DEC JAN FEB 1903-4 \ \ \ 1903 \ \ \ 1893 \ 1900 JULV y'V AUG. JUNE y^ 1903 1900 Kr., \_ 1890 D N , D 1893 J Fig. 76. — Prevailing Monthly Directions of the Wind in Warm, in Normal and in Cold Seasons and Years. differences are marked only in August, September and October. By the system of two daily eye observations we obtain a prevailing north MARYLAND WEATHER SERVICE 271 wind in September and northwest in Octol^er, whereas the hourly obser- vations show a prevailing direction from the southeast during both months. The resultant prevailing directions based on three daily observations agree somewhat more closely with those derived from hourly observations, the chief divergence occurring in August and October. The annual path pursued is best represented by the diagram in Fig. 76, which is based on 24 hourly observations. The prevailing monthly directions derived from the three series of observations are as follows: PREVAILING DIRECTIONS. Jan. Feb. Mar. Apr May June SW sw SW July SW SW SW Aug Sept. Oct. Nov. Dec. Year 7 a. 3 }). 8 a. 8 p. . 3P W W W W W W W NW SB SB SB SB SB SB SW SW SB N SB N NW SB W NW W W NW W W NW Hourly ob ser%'t'ns NW SB W.SE There is a fair degree of uniformity from 3'ear to year in the prevail- ing directions of the wind for the same months. The extent of the departure from the average direction is indicated in Fig. 76, in which the prevailing directions are shown for seasons and a year with a normal temperature, a well-marked temperature below the normal and for those well above the normal in temperature. In each case these seasons and years have been selected from the period from 1893 to 1904, and hence the prevailing directions are based on hourly observations. An inspec- tion of the figure will show that in nearly all cases there is an unusual percentage of northwest winds in the cold seasons, and a predominance of southerly winds (southwest to southeast) in the warmer seasons. This is in harmony with the results obtained by determining the average temperature of winds from each quarter. Selection was made of a number of days in each of the months of January, April, July and October, during which the wind blew from the same quarter all or most of the day. This was repeated for each of the eight points of the com- pass. The average temperature of these days was when computed from the hourly observations. It was not always possible to find days during which the wind blew from the same quarter more than half the day; in 272 THE CLIMATE OF BALTIMORE Buch cases it was necessary to admit days with a direction 45° on either side of the desired point of the compass. In the winter months, the southeast winds are the warmest; in the TABLE LXXV.- PREVAILING MONTHLY AND ANNUAL DIRECTION OF WIND. Year. a 03 1-5 OS a < 83 d 3 D 1-5 si < D. 0) o O c a < 1871 NE NW NW NW SW NW NW NW NW" W^ NW NW W" NW N NW NW NW SW" NW NW NW NW W w w^ w SW" SW W" w W" NW" NW w NW NE NW NW" NW" NW NW NW" W" W NW NW NW NW NW" NW NW NW" NW^ NW NE NW" NE NW N NW^ W w W" W" W" W" w^ w NAV NW W NW NW NW NW NW SE NW NW NW NW NW^ W NW^ NW" NW" NW NW NW" NW NW^ NW" NE NW NW NW NW NW E E E W W" W SE NW NW NW" NW W W" NW NE NW NW NE NW^ NW NW W NW N NW^ NW NE NW NW NE NE NW NW SE NW SE SE W w SE NW N W NW NW NW" NW NW NE NW^ NE SE SE SE NW" NW" SE S SE s SE NW NE NW SE SE SE S NE NW" W" SE SE SW NW^ SE SE W" E SW" SE SE SE SE SE SW SW SE SW s SE SE W" SW W" w s s SE NW^ SE S NW SW s NW SW" E SW^ E SW NW SW N SW^ SB NW^ SE SW s SW SW NW SW SW SW S NW SW SW s s W" s SW NW" SE SW s s SW s SW SW SW SW NW SW w SR SW SW" SW SW" w SW s SW SW" SW w NE N NE SE SE NW W S N S N N SW NW N SW" SW s NW NW SE SE SW N NW SW NE AV S SW NE SE N NW^ SW N N N NE SW" NE SE E SE NW SE N NE S N S N N NE NE NW" SE W NE N SW N SW SE SE N NW" NW^ N N SE N NW^ NW NW^ NW NW" NW^ NW NW" SE SE S NE N N N NW NW NW" NE NW NW NW NW NW N N N SE E E SE NW NW NW NW" NW NAV NW NAV NAA" NAV N NAV NAV W^ NW NW NW NE N NW NW NAV NAV NAV NAV NAV NAV NAV N NW N SW AV AV NAV NAV W N NW NAV NW NAV NAV SAV AA" NAV NAV NE NW SAV W^ E NAV AV NAV NAV N NW^ NAV NW NW NE NAV NAV NW SAV NW N W NW AV AV AV NAV NAV NW NAV NAV NAV NAV NW 1873 NW 1873 1874 NAV NAV 1875 NW 1876 NAV 1877 1878 1879 NAV NAV NAV 1880 NAA" 1881 AV 1883 NAV 1883 NAV 1884 NAV 1885 NW 1886 1887 NAV NAV 1888 1889 NAV NW 1890 NW 1891 NAV 1893 NAV 1893 NAV 1894 NAV 1895 N 1896 SAV 1897 AV 1898 AV 1899 1900 SE AV 1901 AV 1903 AV 1903 1871-1880 NAV NAV 1881-1890 NW 1891 1900 NAV 1871-1903 NAV January, 1871-Oct6ber, 1879, from eye observations at 7.30 a. m., 4.30 p. m., and 11.00 p. m. November, 1879-December, 1886, " " " " 7.00 " 3.00 " " 11.00 " January, 1887-June, 1888, " " " " 7.00 " 3.00 " " 10.00 " July, 1888-November, 1893, " " " " 8.00 " and 8.00 p. m. December, 1893-December, 1903, " hourly record. spring, the south winds; in the summer, the southwest; in autumn, the winds from any quarter between east and southwest have about the same temperature. The relative position of the winds, arranged according to temperature, Avith the Avarmest first, is indicated below: MARTLAXD WEATHER SERVICE RELATIVE TEMPERATURE OF THE WINDS. 273 Warmest. Coldest January SE E sw NE W s NW N April S SW SE w N E NE NW July sw S w SE NW NE E N October SE S sw E NE NW W N Tear SE sw s E W NE NW N COMPARATIVE PREVALENCE OF STATED DIRECTIONS. (In average number of hours and minutes per day.) NE January 3.54 April 1.24 July 2-3« October 4.18 Average '2.54 1.24 3 24 1.42 1.54 0.43 1.42 E SE S SW W 3.06 2.24 1.13 2.36 2.36 7.24 3.18 3.24 0.30 0.30 3.12 1.24 3.06 3.24 7.12 3.06 5.30 3.24 3.24 2.54 3 24 4.06 1.13 4.06 3.48 NW 4.36 3.36 8.24 4.36 4.06 In the above table the figures show the number of hours and minutes during which the stated winds prevailed during the five years from 1893 to 1897. For example, in January a north wind prevailed on the average for the five years, during 12 per cent of each day, or a little less than three hours. This is equivalent to about 3.7 days for the entire month. A south wind is in all months of the 3'-ear of decidedly shortest duration and of least frequency. Monthly Frequency of Stated Directions. A four years' record of hourly wind directions was examined and the observations tabulated in such manner as to show the number of days per month upon which the wind blew from each quarter. The monthly number of days for the entire year is as follows : .\ viTHgc no. of days . . Highest number NE j E I SE 19.0 16.0 15.9 I 16.4 21.8 19.5 19.2 22.0 IX)weBt number 16.2 14.0 14.0 i 9.5 1 I s SW W 15.8 18.7 19.7 19.5 23.8 22.8 9.3 11.0 16.5 NW 20.5 15.8 274 THE CLIMATE OF BALTIMORE The above figures indicate that the wind blows from nearly every quarter once in about two days. Take for example a northwest wind; on the average it blows on 20.5 days per month the year round; in Janu- ary, February, March and July it has an average frequency of 22.2 days, and in September 15.8 days. We may also learn from these figures that the wind blows from four to five different directions every day, on the average. The exact figures, based on the four years' record, are as fol- lows for each month of the year : AVERAGE DAILY NUMBER OF WIND DIRECTIONS. Jan. Feb. Mar. Apr. May June July Aug-. Sept. Oct. Nov. Dec. Year 4.6 4.3 4.4 4.8 4.8 5.2 4.9 5.3 4.6 4.1 4.4 1 4.6 4.7 These figures are in harmony with the facts recorded in the discussion of the diurnal periodicity of wind direction. It was there shown that the wind backed daily from a westerly direction in the morning to south- east or east in the afternoon, and then returned again at night to the west or northwest ; in other words, that the wind shifted through four or five points by noon and returned to its original position at night. The Direction of Upper and Lower Clouds. Table LXXVI is inserted at this point simply to show the prevailing direction of the wind at the level of the upper and lower clouds. The observations cover a period of five years and indicate the directions at 7 a. m. and 3 p. m. The upper clouds include all cirrus forms, the alto- stratus and alto-cumulus; the lower forms include the cumulus and stratus forms. The upper clouds move from the west throughout the year, both in the morning and afternoon, with an occasional exception in the way of a northwest or southwest direction, especially at the afternoon observation. The lower clouds are also mostly from the west at 7 a. m. from May to December; from January to April they are generally northwest. In the ^klARYLAXD WEATHER SERVICE 275 TABLEILXXVU.— PREVAILING DIRECTION OF LOWER AND UPPER CLOUDS. 1 ^ 1 'C 1 2 -. bl X 1 z -< 7 o. m.... tr ir jr ir ir tr ir i««3if^:m:::: .... tr w tr .sir w ir sw tr X ir w . 3 p. m.... \\" \\ \\ w w X w 7 a. m.... xir Air tr .vir tr tr ir ir A' ir ir 3 p. HI.... fr s ir xtr tr E tr If ir ir ir Air ir ir 1884 - 7 a- m.... N NE X w NW w NK SH \\ w w w NW w w 3 p.m.... >> s s\\ w xw xw E w w X w xw xw xw xw 7«. HI.... fr .VK ir ir .sir Air tr ir Air .sir tr ir .s'lr ir Air ir 3p. m.... 1885-1 Ta. m.... }r (r ir xtr sir tr E tr tr II' tr ir Air ir ir sw w N \\ \\ xw \\ w \\ sw w xw w [sp.m.... w NW w sw xw xw xw xw sw \\ sw NW xw xw 7a. HI.... »r STF Nir tr ir A-ir Air tr .s' tr xtr ir ir »r $p. in.. . Tr Sir Nir NW xtr tr Air tr .sir Air XE tr ir ir ir 1886 7a. m.... w NW NW NW NE NW NE w w w w X w sw xw w 3 p.m.... sw w xw w XW NW X \v xw sw w w xw xw \v NW 7a. HI.... ir IC ir tr ir ir tr tr tr ir ir ir ir 3 p. HI.... w ir ir X tr ir ir tr tr tr ir ir tr ir 1887 ^ 7 a. m.... sw SE sw xw XE NW s XE sw NE sw w w sw w sw 3 p.m.... NW NE NW NW SE NW E w sw NW N SE xw NW xw xw j 7 a. »!.... w ir Air )(' tr tr IK 1888 -{ 3 p. m.... 1 7 a. m.... NW w ir w IC w tr sw tr sw .... W W 13 p.m.... w w NW X \v sw \v W 1 7a. HI,... w ir ((• tr tr tr tr tr tr tr ir ir ir 1 3 IJ. III.... jr tr tr xtr tr ir tr ir tr tr tr ir ir Prevail, i 7 a. m.... sw NW N E xw NW \\ w w w w w w w w 3 p.m.... w NW w NW xw w NW N w \v w xw X w NW NW a. ] a. s. 1 Vpiiff chpiiih. ■{ Ci. Cit. y in Italic. I .1. Ci(. 1 L A. S. J Lower clouds. •{ |- ^'"• f Cu. ! S. ■ I s L N. in Roman. afternoon, a northwest direction prevails during eight months of the year in the lower cloud layer, with a direction from the west in January, March, August and September. 19 276 THE CLIMATE OF BALTIMORE ELECTEICAL PHENOMENA. Thunderstorms. The intimate relation existing between thunderstorm formation and temperature is demonstrated by an inspection of Table LXXVII and Fig. 77. The maximum frequency occurs in the month of greatest heat TABLE LXXVII.-HOURLY FREQUENCY OF THUNDERSTORMS. Midn't to 1 a. m. 1- 2 2-3 3-4 4- 5 5- 6 6- 7 7-8 &- 9 9-10 10-11 11-Noon.. Noon- 1 p. m 1-2 2-3 3-4 4- 5 5- 6 6- 7 7-8 8- 9 9-10 10 11 ll-Mldn't. Totals 24 97 141 164 100 38 61 66 69 38 34 31 16 13 610 Table LXXVII shows the total number of thunderstorms recorded as begin- ning within the stated hours in the 27 years from 1876 to 1903, during each month and during the entire year. and at the hour of the daily maximum temperature. In tabulating all thunderstorms which passed over Baltimore during a period of 28 years, a total of 678, we find the following distribution by months: Jan. Feb. Mar. Apr. Maj' June July Aug. Sept. Oct. Nov. Dec. Year 13 11 20 33 107 1.56 179 Ill 43 7 6 2 1 678 \ IMARYLAXD WEATHER SERVICE 277 - i ' 1 1 Na ON 1 2 3 4 5 6 7 8 9 1 ) 1 Mdt 1 1 ' — ' s — i • t . ^>-. N ^^^-^/T\: •i^ x > Ir 4 ^^ s , • • • " u V ^ ^ \! — LV '. • — 1 • ( ^fe—V^y ^\ — -^ • "^ ■ Fig. 77. — The Frequency and Distribution of Thunderstorms. The diagram represents over 650 thunderstorms which passed over Baltimore in the 30 years from 1871 to 1900. The density of the shading increases with the frequency of the storms, showing a maximum frequency between 3 p. m. and 5 p. m. in the month of July. For the hours of the day and month during which less than five storms were recorded in 30 years, the actual number is indicated by the small dots. The figures attached to the curved lines represent the total frequency in 30 years. 1 1 1 1 ,ll 1, 1 1 1 1 Fio. 78. — The Average Monthly Frequency of Occurrence of Thunderstorms. 278 THE CLIMATE OF BALTIMORE About five-sixths of the total annual number occur in the months of May, June, July and August. In the winter months they occur only at rare intervals, generally in connection with cyclonic storms which ex- hibit strong contrasts in temperature. The summer storms occur mostly in connection with shallow and not very well defined cyclonic depres- 1880 1885 1890 1895 1900 30 20 10 Fig. 79. — The Annual Frequency of Occurrence of Thunderstorms from 1871 to 1904. sions. That there is a strongly marked diurnal periodicity in the for- mation of the summer thunderstorms is shown bv the following figures: HOURLY FREQUENCY OF THUNDERSTORMS. Hours ending Morning' Afternoon 1 22 3 1 I 3 45 76 78 5 fi 7 8 9 10 n 12 I 3 6 fi 6 8 9 1.5 61 .55 69 3d , 31 , 31 . 16 13 :XIARYLAXD WEATHER SERVICE 279 The hourly distribution for all months, expressed in terms of the total frequency in 30 years, is shown in Fig. 77. The monthly and annual distribution by years, from 1876 to 1903, is shown in Table LXXVIII. The annual changes in frequency are also shown graphically TABLE LXXVIII.— NCMBER OF THUNDERSTORMS PER MONTH. Year. 1876 .. 187T 1878 1879 1880 1 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 I89I 1892 1893 1894 1895 1896 1897 1898 1 1899 1 1900 1901 .. 1902 1903 Totals, 1876-1903... 3 Averatre 0.1 11 0.4 •? — 30 0.7 33 1.2 107 3.8 166 5.6 6 15 3 6 5 5 2 5 8 7 5 1 « 4 3 3 8 5 11 7 7 6 8 7 8 6 6 7 179 6.4 111 4.0 4H 1.5 26 16 14 8 20 21 18 36 20 37 19 24 17 16 18 25 31 21 43 24 25 0.2 0.1 678 24.2 in Fig. 79. The average annual number is appro.\iinately '■^4, with a maximum frequency of 42 in 189-4, and a minimum of 8 in 1879. The following figures express more exactly the average monthly and annual frequency (See also Fig. 78) : .lull. Fob. Mar. 0.7 A,,r. •May .June July Aug. Sept.! Oct. 1.6 0.2 Nov. 0.3 Dec. Yiar 0.1 0.4 1.2 3.8 6.6 6.4 4.0 0.1. 24.2 280 the climate of baltimore Thunderstorm Probability. The probability of the occurrence of a thunderstorm upon any desig- nated day may be expressed in a very rough way by finding how many times a storm occurred on that day in the past. By examining all records of thunderstorms for 27 years and arranging them according to the day of the month upon which they occurred, we may roughly obtain a percentage of probable occurrence. Not much reliance should be placed upon such a method of forecast- ing, but some interesting relative values are brought out. In the past 27 years one or more thunderstorms have occurred on every day of May, June and July, and on all but one day in August (the 20th) ; no thun- derstorm has occurred on September 6, 13, 21 to 23, 27, 28 or 30. In April there is no record of a storm on the following days : 1, 3, 6, 7, 13 to 15, 21 to 23, 25, 30. The highest number occurring on any stated day in May is 7, on the 21st; in June, 11 on the 21st; in July, 11 on the 5th; in August, 7 on the 12th. Hence the highest probability of occurrence upon any day in the year is only eleven twenty-sevenths, or about 41 per cent. The average probability for a day in May is 14 per cent; in June, 20 per cent; in Jnlj, 23 per cent; in August, 15 per cent. The probability for the Fourth of July is only 17 per cent, or 5 per cent less than the average for July days. According to the Baltimore records a thunderstorm has passed over the city on July 4 only five times in 29 years. One has occurred 11 times on July 5, in the same period, making the maximum probability 41 per cent of the total number of such days in 29 years. Consecutive Days with Thunderstorms. Thunderstorms generally occur as isolated storms in the vicinity of Baltimore. In over 80 per cent of all instances, a second storm does not occur on the following day. In only 14 per cent of all cases have there been thunderstorms recorded on two successive days, and in only a little over three per cent have storms occurred on three successive days. Only on one occasion have as many as 5 occurred on 5 successive days. These percentages vary in different months but they are not MARYLAND WEATHER SERVICE 281 large in any month. The following table shows the figures for each month and for the year: CONSECUTIVE DAYS WITH THUNDERSTORMS. (Total number in 28 years.) s 1-5 si p. < o a 3 si < OQ O > o 03 3 11 16 i 26 2 65 13 2 87 18 4 1 1 94 21 6 3 6T 12 4 26 6 1 T 6 2 410 72 17 4 1 On 2 consecutive days "3 " " "4 " " "5 " " s. w. Fic;. 80. — The Direction of Movement of Thunderstorms. The diagram shows the actual and relative frequency of thunderstorms from each direction of the compass. The total number of storms represented is nearly 400. Direction of Thunderstorms. In the vicinity of Baltimore, thunderstorms usually come into view from some point between northwest and southwest. Out of a total of 282 THE CLIMATE OF BALTIMORE about 400 storms, nearly 90 per cent moved from some one of tliese points. (See Fig. 80.) The order of frequency of direction is as follows : THUNDERSTORM DIRECTIONS. (1876-1902.) Jan. Feb. Mar. Apr. 8 3 I May June July Aug. 21 18 21 3 2 2 '3 Sept. 18 6 'i i Oct. 1 2 2 Nov. "i Dec. Year NW to SE 1 i 3 3 3 4 i 26 19 13 2 2 2 20 29 31 '3 1 1 33 32 20 3 6 '5 3 .. 126 \V •* E .. 115 SW " NE S •• N .. 104 s SE " XW E " W NE " SW 1 14 4 N " S 13 Pressure Changes Duking Thunderstorms. A thunderstorm usually occurs with a falling barometer; the bar- ometer rises during the first few minutes after the storm has begun, falls slightly before the close of the first hour, and then maintains a steady pressure for several hours, eventually rising slowly (see Pig. 81). In other words, the storm usually breaks in the trough of a cyclonic dis- turbance. The following table shows the average hourly barograpli PRESSURE BEFORE AND AFTER THUNDERSTORMS. (Station readings; not reduced to sea-level.) No. Hours Preceding. Rise. Hours following. Month. 5th 4th 3rd 3nd 1st Begin- P. 'H 1st 2nd 3rd 4th 5th 29.45 ning. H ss .30 .20 Feb.... (3) .43 .40 .36 .34 29.31 0.25 29.36 .34 .26 .18 March. (3) .67 .63 .60 .56 .51 .49 0.25 .53 .51 .51 .51 ! .53 ..>t April.. (2) .38 .38 .38 .38 .38 .38 MS .47 .44 .44 .43 .44 .44 May... 118) .78 .77 .76 .76 .74 .74 0.42 .79 .77 .77 .78 .79 .80 June... (16) .79 .78 .77 .76 .75 .74 0.43 .77 .75 .74 .74 .75 .75 July... (19) .77 .77 .75 .74 .73 .72 0.51 .78 .74 .74 .74 .74 1 .76 Aug... (11) .80 .80 .79 .79 .79 .79 0-40 .84 .81 .80 .79 .79 .80 Sept. . . (6) .76 .75 .74 .73 .72 .70 0.53 • .76 .74 .75 .77 .78 .80 Aver... (76) .68 .66 .65 .63 .62 .61 .45 .66 .64 .63 .63 .63 .6?$ * Time from beginning of the rise to its maximum, in hours and minutes. readings before and after about 75 thunderstorms selected from the records of the past three or four years. The readings of the barograph 3»ia ' M«.rck 3o, 130I M».rt.lv 2}, ?9e3. flpr. 2,6, I9»2.. Moy i, <»o2.. Muy iS. 1902. F«.l).a8, (>«3 /Mav, i*. f9»3. iM»3 (1, I9» + . Ju N 3, 1901. 7u^y 20, '^oi. 7u/y ', 'J't. flgg fc, i9o2 /?uj. i 7, ( j«z ffw) 2 9 j9oS. Fio. 81. — Some Typical Barograms During Thunderstorms and Squalls. 284 THE CLIMATE OF BALTIMORE are given for the five hours preceding and following the breaking out of the storm. The minimum reading is also given, just before the begin- ning of the " hump/' which constitutes the characteristic feature of a barograph curve during the passage of a thunderstorm. The duration of the rise in pressure, from the minimum to the maximum point attained in the " hump " is given in hours and minutes. Of the 76 thunderstorms examined in the above table, about one-third began with a value between 29.60 inches and 29.69 inches for the bar- ometer reading, assuming the beginning of the rise in tlie barometer to be the beginning of the storm. Tabulating the barometer readings ac- cording to the pressure at the breaking out of the storm, we have the following comparative frequency of stated values : FREQUENCY OF STATED READINGS OF THE BAROMETER AT THE BEGINNING OF THE STORM. Barometer Re 29.30-39 in 4049 ading. ;hes Actua 3 5 Frequency I. Pel •centage. 4 7 50-59 5 7 60-69 70-79 80-89 90-99 24 14 17 7 32 18 22 9 30.00-09 1 1 7G 100 The thunderstorms in the above table were confined almost entirely to the months of May to August. We see that the storm broke most fre- quently when the pressure registered some value between 29.60 inches and 29.69 inches; this was true of 32 per cent of all cases tabulated; in 72 per cent of all cases the barometer reading was between 29.60 inches and 29.89 inches. In only one instance was the pressure above 30.00 inches, namely, in July, 1900. The lowest pressure recorded in any case was 29.30 inches, in February, 1903. See Fig. 81 for the character of the rise in pressure during thunderstorms. Hail. The phenomenon of hail formation is so intimately associated with the dynamics of thunderstorms that the treatment of the subject is taken MARYLAND WEATHER SERVICE 285 up in connection with these storms rather than with the subject of pre- cipitation. Hail is not of frequent occurrence in the vicinity of Balti- more. During a period of 28 years it has been recorded but 49 times, or less than two times per year. The annual number has varied between and a maximum of 6. The monthly and annual distribution is shown in Table LXXIX, and the hourly distribution by months in Fig. 82. TABLE LXXIX.-FREQUBNCY OF OCCURRENCE OF HAIL. Year. 1876. 1877. 1878. 1879. 1880. 1881 1883. 1883. 1884. 1885. 188fi. 1887.. 1888.. 1889.. 1890.. 1891. 1892. 189:5. 1894. 1896. 1896. 1899. 1900. 1901. 1903. Total in 28 years... 9 I 6 66 Tlie liourly distribution for the entire year is as follows: HOURLY FREQUENCY OF HAIL. (Total number iu 28 years.) Time A. M. 1 9 10 11 Noon. 1 Z 8 4 6 6 7 8 9 10 11 P.M. Frequency, i 2 1 1 2 8 < 1 i 4 6 13 7 4 8 2 1 286 THE CLIMATE OF BALTIMORE 5 6 7 8 9 10 II Noon 12 3 4 5 7 8 9 10 tl MO ■ 1 — ' ....i. .it' (1 - 1 ■ • •[ Fig. 82. — The Frequency of Occurrence and the Hourly and Seasonal Distribution of Hailstorms. The diagram show.s all of the hailstorms recorded as occurring in Baltimore during a period of 28 years. Each black dot represents a storm. Mdt. Noon Mot Noon Mot. Noon Mdt Inches 1 2 3 30.00 ^ -^ .50 29.00 30 00 s ~--— 4 5 6 29.50 — ■ -..^ — Fig. 83. — Barograms During Hailstorms. Each barogram represents a period of 24 hours, from midnight to midnight. The time of occurrence of the hail-storm is indicated by the sharp temporary rise and fall in the curve. Dates of the storms represented : 1. .July 7, lOiH. 2. February 28, 1002. .3. August 27. 1902. 4. June 8. 190.3. ."). May It). 1904. 6. .July 5. 1004. MARYLAND AVEATHEU SERVICE 287 The hourly frequency rises to a maximum between 4 p. m. and 5 p. m. The dates of all recorded occurrences of hail in the vicinity of the Balti- more station of the Weather Bureau are siven in Table LXXX. TABLE LXXX.— DATE AND HOUR OF OCCURRENCE OF HAIL. Date Time - First precip.* Date Time First precip. 1876, Mar. 2» 7.1.5 p. ra. 1.S95, July 5 12.25 p. m.-12.30 p. m. " May 12 2.1.5 p. m.- 2.30 p. ra. " 16 + 4.'36p " 21 9.35P " Aug. 11 " " 31 4. .50 p. m.- 4.52 p. m. 4..30p. m.- 4.35 p.m. 1879, June 11 2.00 p. m.- 3.03 p. m. ., 28 + 4.'29p " Sept. 19 + .... 3.10p 1880, Apr. 17 Early a. m. 1S96, July 27 8 13 p. ra.- 8.17 p.m. " July 20 t 4.40p 1897, Mav 21 " Aug. 23 1.45 p.m.- 1.55 p.m. 11.47 a. m.-11.52a. m 1881, June 8 1882, •• 19 t 3.15p Noon. 1898, May 16 \ 4.10 p.m.- 4.13 p.m. 1 5.14 p. m.- 5.17 p. m. 1884, July 11 4.00 p. m.- 4. .50 p. ra. 1899, Mar. 12 J 8.27 p. m.- 8 35 p. m. 1887, Feb. 18 5.02 p. m.- 5.05 p. ra. •' 28 8.07 a. m.- 8.09 a. ra. " May 26 t 5.26p '• Apr. 16 jl0.25a.ra.- .... 1 12.20 p. m.- .... " June 18 + 4.35p " July 18 5.10p " May 16 t 6.50p 1888, June 16 2.30 pVin.- 2.35 p. m. " June 6 t 7.22p " 23 t ll.'o.^a " Aug. 21 7. '5:3 p. m.- 7.3S p. m. " Aug. 8 5.45 p. m.- 5. .50 p. m. 1901, May 25 t 10 15p 1890, Apr. 27 3.45 p. m.- 4.00 p m. " July 7 7.00 p. m " May U t e'.srlp 1902, Feb. 2.s 9..50a.m " June 12 3.55p " Aug. 27 4.43 p. m.- 5.05 p. ra. 1892, Mav 2:! i 4.08 p. m.- 4.]6p m. 19C3, May 24 3.24 p. ra.- 3.27 p. m. " June 30 12. 02 p. m.-12.10p m. " June 8 3.27 p. m.- 3.33 p. m. 1893, July 3 5.25 p. m.- 6.35 p m. J, t 4."l6p 1894, May 6 fi..57 p. m.- 7.03 p m. " June 12 4.37 p. m- 4.40 p m. " 24 t 4.'l6p * 1 n the absence of the exact time of occurrence of hail the time of beginning ot precipi- tation is given. + Hail in city or suburbs ; none at station. i Not accompanied by a thunderstorm. In Fig. 83 a few typical barograms are reproduced showing the char- acteristically sharp rise and fall of the atmospheric pressure during the passage of a hailstorm. In the thunderstorm curve the summit of the " hump '' is usually more rounded, as shown in Fig. 81. 288 THE CLIMATE OF BALTIMORE Auroras. The following brief list contains all occurrences of the aurora borealis reported in the records of the U. S. Weather Bureau at Baltimore since the establishment of the station in 1871 : Date. Duration. Date. Duration. 1872, Feb. 3 8 p. m. to 9 p. m. 1892, Feb. 13 6.30 p. m. to 9 p. m. Apr. 11 8 p. m. " 10 p. m. May 18 8 p. m. " 11 p. m. Auf?. 3 8.40 p. m. " 10 p. m. July 16 10.30 p. m. " 11.30 p. m. Aug. 4 About 9 p. m. 1893, Feb. 4 9 p. m. " 12 md't. Aug. 8-9 9 p. m. to 3 a. ra. 1894, Feb. 23 9 p. m. " 10 p. m. Oct. 14 6.30 p. m. " 7 p. m. Mar. 30 7.20 p. ra. " early a. m. Nov. 1 10 p. m. ." 11 p. m. 1897, Jan. 23 Evening. 1873, June 36 About 10 p. m. 1898, Sept. 2 About 10 p. ra. 1882, Apr. 16-17 10 p. m. to 3 a. m. 1903. Oct. 12 7 p. m. to 7.30 p. m. Apr. 20 12.30 a. m. to 3 a. m. SuxspoTS AND Weather. The effort to extend the period covered by weather forecasts has ever been one of the chief aims of the practical meteorologist. The limit of time for which forecasts are now issued by American and European official weather services is about three days. The forecasts made from day to day generally cover from 24 to 48 hours ; under favorable conditions the time is occasionally extended to three, or even four days, but this is only done in exceptional cases. The three or four day limit is probably the utmost that will be realized from present methods, and with the material now at our disposal. The only hope of extending the period lies in the discovery of some new laws of weather sequences. The search for periodical recurrences of similar weather conditions has long been one of the most interesting, and, at the same time, one of the most elusive problems in cosmical ph3^sics. The investigations have usually been along two lines : A series of observations has been subjected to close examination and critical analysis in order to discover any periodic change which may be hidden in the constant fluctuation of values; or a periodic movement has been assumed and the weather observations examined for synchronous changes. There is but one undisputed source of terrestrial weather changes — ; namely, the sun. While no one doubts the influence of the sun upon MARYLAND WEATHER SERVICE 389 the earth^s atmosphere many claims have been, and are still being made in favor of attributing to other heavenly bodies, such as the moon or the planets, a considerable effect. The champions of the moon's influence are legion, and they never grow less; but the arguments of several centuries, including much serious and intelligent effort, have not suc- ceeded in securing for lunar or planetary forecasts a position more exalted than the pages of the perennial almanac. It is now approximately 100 years since a definite period was dis- covered in the increase and decrease of sunspot frequency, and less than 50 years since the flames emanating from the surface of the Sim, or the solar prominences, were first observed. The first definite relation between sunspot frequency and terrestrial changes was the discovery of the synchronous activity of the magnetic needle. There is now no question about the coincidence of these phenomena whatever may be the true relation existing between them. In attributing terrestrial changes of the weather to " sunshine," we have until comparatively recent times assumed a constant output of radiant energy from the sun. In view of the fact that our present knowledge concerning the physical condition of the sun indicates a surrounding atmosphere composed of incandescent metallic vapors, is it not more rational to suppose that the temperature of these highly heated gases is var\ang constantly, than it is to think of them as at a constant temperature? If the temperature does vary, the fluctuations must necessarily affect to a greater or less degree the physical condition of our own atmosphere. The question then becomes one of degree of influence. There are many observed facts which point to a varying output of solar radiant energy, and quantitative measurements will not long remain unknown. Just what the nature of this influence is has certainly not yet been demonstrated. One obstacle in the way of more rapid progress toward a solution of these problems may be found in the crudeness of much of our observational data, and the lack of uniformity in the methods and hours of observation. Moreover, in the middle lati- tudes where most of our best observations, and the longest scries which we possess, have been made, the non-periodic fluctuations are so much 390 THE CLIZilATE OF BALTIMORE greater than the periodic changes sought that the latter are separated out from the former only with the greatest difficulty and care. The most favorable regions of investigation for periodicities based on solar changes are the tropics. Here the daily, seasonal, and incidental changes in weather conditions are more uniform and less pronounced, permitting of more ready detection of the periodic changes of longer duration. It seems highly probable that changes in terrestrial temperature, in rainfall, storm frequency, etc., may be due to changes in the physical constitution of the sun's surface. It may also be that these solar changes are not reflected directly in the conditions above mentioned. Similar weather conditions should not be expected in all parts of the earth at the same time. The results of efforts thus far to find a direct connection between the sunspots and weather changes have apparently failed largely as a consequence of dissimilar weather conditions found in different localities during similar phases of the solar period. These contra- dictions may be only apparent, not real. Let us suppose, for instance, that the normal distribution of pressure over large areas is disturbed as a result of changes in the quantity of heat received from the sun from year to year. We would then have excessive heat in some places and at the same time abnormal cold at others; or we would have excessive rains here and droughts there; or an increase in storm fre- quency in one place and a decrease in another, when compared with average conditions. Such variations cannot be looked upon as contra- dictory; they are the natural results of changes in the distribution of pressure, changes such as we see upon our weather charts every day. The present status of the problems concerning the relation between the varying physical conditions of the sun and synchronous changes in our terrestrial atmosphere is well stated in the following extract from a recent paper by Professor Bigelow,* who is one of the most active and able investigators in this most promising field of cosmical physics. " The numerous studies during the past fifty years into the apparent *Bigelow, F. H. Synchronism of the Variations of the Solar Prominences with the Terrestrial Barometric Pressures and the Temperatures. Monthly Weather Review, Washington, D. C. November, 1903. MAETLAXD WEATHER SERVICE 291 synchronism between the solar variations of energy and the terrestrial effects, as shown in the magnetic field and the meteorological elements, have been on the whole unsatisfactory, if not disappointing. Just enough simultaneous variation has been detected in the atmospheres of the sun and the earth to fascinate the attentive student, if not to justify a large expenditure of labor, in view of the great practical advantages to be obtained in the future as the result of a complete understanding of this cosmical pulsation. The attack upon the problems has really consisted in rather blindly groping for the most sensitive pulse in the entire cosmical circulation, and in disentangling the several interacting t}-pes of impulses. It is e^-ident that the partial failures hitherto attending this work have been due to two principal causes: (1) The comparison was made between the changes in the spotted areas of the sun and the terrestrial variations, but these solar changes were not sensitive enough to register a complete account of the action of the solar output. Discussions of the spots are being replaced by others upon the solar prominences and faculse, which respond much more exactly to the working of the sun's internal circulation: (2) The magnetic and meteorological observa- tions have not been handled with sufficient precision to do justice to the terrestrial side of the comparison. It is evident that all these physi- cal data at the sun and at the earth must be computed with an exactness comparable to that of astronomical observations of position, if meteorology is to be raised to the rank of a cosmical science. When one considers the crudeness of the meteorological data, taken the world over, due to the character of the instruments employed, the different local hours of observation, and the divergent methods of reduction, it as no wonder that small solar variations have been swallowed up in the bad workmanship of meteorologists. The prevailing methods have been sufficient for forecasting and for climatological purposes, but they are entirely inadequate for the cosmical problems whose solution will form the basis of scientific long-range forecasts over large areas of the earth — that is, for forecasting the seasonal changes of the weather from year to year. It is perfectly evident that if secular varia- tions f)f any kind, such as the annual changes in terrestrial pressure, 20 292 THE CLIMATE OF BALTIMORE temperature, or magnetic field, are to be attributed to solar action, the original observations must be finally reduced to a homogeneous system. The local peculiarities of each station must be carefully eliminated, and the data of numerous stations must be concentrated before anything like quantitative cosmical residuals can be obtained. When we consider that .there have been numerous changes in the elevations of barometers, various methods of reducing the readings, and many groups of selected hours of observations entering into the series at the same station, how could it be expected that anything better than negative results in solar problems would be obtained? The skeptical attitude of conservative students, who declare that the many indecisive results already obtained mean that there is no true and causal solar-terrestrial synchronism, is, of course, quite fallacious until it has been demonstrated by the use of first-class homogeneous data that the suspected physical connection is imaginary. There is but little question that the existing uncertainty is in ^f act based upon the use of the ver}^ imperfect methods of observation and reduction which have prevailed in meteorological offices, rather than upon the unreality of the phenomena in nature." The results of a comparison of Baltimore weather observations Avith the sunspot and solar prominence frequency curves have not differed from those arrived at in similar investigations elsewhere — they neither prove nor disprove an intimate relationship. As pointed out in preced- ing paragraphs there are synchronous changes here and there in the constantly fluctuating terrestrial conditions, but on the whole the evi- dence is negative. In view of the complicated character of the weather conditions, especially in our middle latitudes, a close agreement in phase of any periodic changes need scarcely be looked for, but the length of the period of the terrestrial and solar changes should harmonize. In Fig. 84, the sunspot and solar prominence curves, constructed from Wolf's tables as printed in the Monthly Weather Eeview* are shown in connection with curves representing the actual annual changes at Baltimore in: (a) the mean pressure, (&) the mean temperature, (c) ♦Monthly Weather Review of the U. S. Weather Bureau for April, 1902. MARYLAND WEATHER SERVICE 293 the total rainfall, (d) the frequency of thunderstorms, and (e) the fiequency of storm winds (exceeding 25 miles per hour). In Plate XII, these facts have been presented again in a modified form, the annual values for the climatic conditions having been smoothed, eliminating some of the irregular fluctuations in order to show more clearly any periodic occurrences of longer period. The values employed in the con- struction of the curves of Plate XII were computed by means of the following formula, ^^^^ TL^ in which a, h, and c represent actual 4 values for three successive years. In this manner a smoothed value was computed for each year of the entire series. Plate XII contains in addition a record of all excessive rains at Balti- more from 1836 to 1904, an excessive rain being defined as a fall of 2.50 inches or more in 24 consecutive hours. These excessive rainfalls were taken from two distinct records; those occurring from 1871 to 190-4 are a part of the official record of the U. S. Weather Bureau; those of the period from 1836 to 1870 are from the record of the Army Medical Department at Fort McHenry, with very few exceptions. The rainfalls of the earlier period are apparently too frequent, owing to the fact that there was more uncertainty in noting beginnings and endings of precipi- tation than in the later series. The earlier record doubtless contains excessive rains in which the time limit was extended to 30 or 36 hours. However, the grouping and relative frequency of these excessive falls are the features to which especial attention is called ; the actual frequency is of less importance. In a preceding paragraph reference was made to the fact that the periods of excessive frequency of heav}^ rains coincided very closely with the periods of minimum sunspot frequency from 1871 to 1901. The earlier observations were not then at hand. On extending tlie series of observations back to 1836, the same nice agreement does not hold good; there is a gradual change to a reversal in the phase of the sunspot period. However, the grouping is very striking, and the average length of the periods from maximum to maximum, or minimum to mininnun. agrees very well witli the average length of the sunspot and solar prominence periods. (Smoothed annual values for weather condifo"'' • excessive rainfalls, which are observed values.) MARTLAXD WEATHER SERVICE 395 These selected Baltimore observations are reproduced in Fig. 84 and Plate XII, not so much to call attention to any particular cycle of changes as to place tliein into convenient form for a critical study by some who may later find a more suitable clue to the solution of the difficult problem of the relationship between solar activity and terrestrial weather changes. General Character of the Seasons. The average values of climatic conditions and the departures from these values have been discussed in considerable detail in the text and tables of the preceding pages. To all but the expert in the study of statistical tables it is a difficult matter to derive, from a table of figures, no matter how perfect the arrangement, a satisfactory conception of the general character of a selected period, be it a day, a month, a season, or a year. The graphic method of presenting results appeals to a greater number because it enables the eye to take in at a glance relations between groups of values which would be more or less obscure to the casual reader when presented in tabular form. While recognizing the limitations of the graphic method for representing such a complex conception as the general character of the season, it has yet seemed profitable to resort to the use of a diagram for grouping such factors as are deemed most important in characterizing the weather conditions of a season. The results arc siiown in Plates XIII to XVII. in which eight selected factors, expressed as departures from the normal climatic conditions at Baltimore, are presented for each season and year from 1871 to 1904. The choice of factors, as well as their arrangement, was a purely arbitrary matter, hut they, as a iiiaitcr of course, remain the same for each season and year. Tlir method of cliartiug may be briefly described here, although the legends on the plates will be found sufficiently clear for this purpose. The normal value of each of the climatic factors selected is represented by a point in the circumference of a circle, at the inter- section of a radius representing one of the eight points of the compass. Taking for example the mean seasonal temperature: A departure above the normal would be represented by placing the point a given number of units beyond the circle along the extension of the radius representing 296 THE CLIMATE OF BALTIMORE temperature ; for a departure below the normal, the point would be placed within the circle along the same radius. By applying a similar method for each of the factors and joining the points thus located, we have a char- acteristic octagonal figure. A season having all its points located in the circiimference of the circle would be represented by a regular octagon. The degree of departure from the regular octagon shows at a glance the amount and the character of the departure from normal climatic con- ditions of the season inspected. The unit adopted for measuring the amount of departure is the same in the discussion of all seasons and years, namely, the average variability of the factor. The average varia- bility was obtained by adding up the individual departures for each season or year and dividing by the number of years employed. The normal value for each factor is shown by the figures given below the designation of the factor in the key accompanying each set of diagrams. To bring the form representing the character of the season into further relief and separate it from the scale, the former is tinted. The comparison of the seasons at Baltimore from 1871 to 1904 will be found to be greatly facilitated by the use of these diagrams. OBSERVATIONS AND INSTEUMENTAL EQUIPMENT. Historical Notes. Volume I of the Eeports of the Maryland Weather Service (1899) contains a report upon the progress of meteorology in Maryland. Eefer- ences are there made to the early records of the weather which came to the notice of the writer, and to the development of systematic instrumental observations. While temperature records were regularly kept as early as 1753 in Prince George's County, we find no evidence of any instrumental observations within the limits of Baltimore City, or in the immediate vicinity, until the series made by Capt. Lewis Brantz, referred to in the preceding pages in connection with the discussion of temperature and rainfall. To the best of the writer's knowledge the very complete and accurate observations of Capt. Brantz were the earliest made within Baltimore City. VOLUME 2, PLATE XIII. PRECIPITATION 10 INCHES corner of the plate. A departure above ( + ) or below ( — ) the f departure (indicated by the figures 1, 3, 5 along the radii) is the MARYLAND WEATHER SERVICE General CHAnACTEn of the Seasons.— Winter. average seasonal variability of the factor. °"' ■ '■'=^P«'="^«'y. =■"• ^'o^g a radius or its extension. Tlie unit of departure (indicated by the figures 1. 3, 6 along the radii) is the VOLUME 2, PLATE XIV. DAYS WITH ECIPITATPON PRECIPITATION 10 9 INCHES lor of tho plate. A dpparture above (4-) or below ( — ) the jparturc (indirated Ijy the figures 1, 3, 5 along the radii) is the General Character i Seasons. — Sprin Pointe in the oircumferencp of the circle, at the Interaection of the railii (0), represent the average value of factors enumerated In the key In the lower right-hand corner of the plate. A departure above ( + ) or below (— ) the average (or normal) value of the factor is shown by the position of points .beyond or within the circle, respectively, and along a radius or Its extension. The unit of departure (indicated by the figures 1, 3, 5 along the radii) is Ihe average seasonal variability of the factor. VOLUME 2, PLATE XV. PRECIPITATION '»« INCHES coriipr of Iho plate. A departure above ( + ) or below ( — ) the of departure (indicated by the figures 1, 3, 5 along the radii) is the General Chakactee of the Seasons. — Sujimer. Points In the circumference of the circle, at the intersection of the radii (0). represent the average value of factors enumerated in the key in the lower right-hand corner of the plate. A departure ahove ( + ) or below (— ) the average (or normal) vahie of the factor Is shown by the position of points beyond or within the circle, respectively, and along a radius or its extension. The unit of departure (indicated by the figures 1. 3, 5 along the radii) is the average seasonal variability of the factor. VOLUME 2, PLATE XVI. DAYS WITH ECIPITATION PRECIPITATION 98 INCHES cornpr of tlio platp. A doparturp ahovp (-f ) or 1)p1ow ( — ) thp of rlepartiirp ( indicated by the fiKiires 1. 3, 5 along the radii) is the Seasons. — Autumn. Points in thp rircunitcrenrp of the rircle. at the inlorspotion of the railii (( veragp (or normal) value of tlie factor is .shown by the position of points tje verage seasonal variability of the factor. below I — ) the represent thp average value of factors enumerated in the key in the lower right-hanil corner of the plate, A rleparture above ( + ) ntl or within the circle, respectively, and along a radius or its extension. The unit of departure (indicated by the figures 1. 3. 6 along the radii) is the VOLUME 2, PLATE XVII. cornor of tlio platp. A departure above (-|-) or below ( — ) the if departure (indicated by the figures 1. 3. 5 along the radii) is the PRECIPITATION « INCHES Point average ( average seasonal I the circumferpnoe of the circle, at the intersection of the radii (0), represent the average value of factors enumerated in the key in the lower right-hand corner of the plate. A departure above ( + ) or below ( — ) the normal) value of the factor is shown by the position of points beyond or within the circle, respectively, and along a radius or its extension. The unit of departure (indicated by the figures 1, 3, 5 along the radii) is the annual variability of the factor. MARYLAND WEATHER SERVICE 297 Some years later, in 1831, a station of the U. S. Army Medical Depart- ment was established at Fort McHenry, where observations were main- tained with but little interruption until 1892. Between 1830 and 1840, two or three individual Baltimore observers sent occasional reports to the Maryland Academy of Sciences, or to the Franklin Institute in Philadelphia. The Smithsonian Institution was established in 1847, and very soon after numerous voluntary observers reported weather conditions regularly from different parts of the State. The Baltimoreans who cooperated with the institution between 1850 and 1860 were Dr. A. Zumbrock (1850-52), Dr. Lewis H. Steiner (1853), and Prof. Alfred W. Mayer (1857-59). In the year 1870, the U. S. "Weather Bureau (then known as the IT. S. Signal Service) was established by act of Congress. The Baltimore station was among the first to be opened, and observations were main- tained without the interruption of a single day until the present time. The instrumental equipment was always of the first order and was steadily increased with the growth of the Bureau. During the past ten years, continuous records of air pressure, temperature, wind velocity and direction, sunshine, and rainfall, by means of self-recording instruments, have been maintained. In 1902, a Richard hygrograph was added, the property of the Maryland State Weather Service. Details concerning the history and equipment of the U. S. Weather Bureau station are given in the following pages in tabular and statistical form for ready reference, including changes in the location of the observing station, in the elevation of instruments and in tlie personnel of the station. In 1891, the Maryland State Weather Service was organized; the ])urpose and method of organization are fully described by the director in Volume I, of the Reports of the Maryland Weather Service, issued in 1899. From the beginning there has always been an intimate and harmonious cooperation between the National Service, the State Service, and the Johns Hopkins University. The offices of the U. S. Weather Bureau and of the Maryland Weather Service have been in the buildings of the University since the establishment of the State Service. The Board of Control of the Maryland Service comprises a representative of 298 CLIMATE OF BALTIMORE MARYLAND WEATHER SERVICE 299 METEOROLOGICAL OBSERVERS AND OBSERVATIONS IN BALTIMORE, MD. West part of city 1 Private Fort McHenry Observer Capt. Lewis Brantz Post Surgeons, L'. S. A. Dr. G. S. Sproston, U. S. N. Md. Academy of Science and Literature Dr. T. Edmondson, .Jr. 1 West part of city Dr. A. Zumbroclt Dr. Lewis H. Steiner Baltimore Medical Institute Prof. Alfred M. Mayer Wiu. Lutber Woods ! .Johns Hopkins Hospital Auspices North Longitude Elevation West of Latitude Md. Acad. Sci. and Lit. Johns Hopkins *U. 8. Weather Bureau; {U. S. Signal Service un-j University til July 1, ISiil) Smithsonian Institution Smithsonian Institution Smithsonian Institution Maryland S. W. S. and U. S. AV. Bureau United States Government 16 ' 76 17 76 I 17 I 76 feet Period. 50 113 +123 1817-1834 Jan. to Aug., 1839 Nov., 1S36 to June, 1837 1S31 to June, 1859 Apr., 1«61 to Jan., 1863 1864 to Feb., 1893 B6, 21 and 33, June, Sept. and Dec. 1835 to 1837 1846 to Oct. 1853 Apr., 1850 to July, 1853 Jan. to Oct., 1853 Hours of Observation Items Observed 8 a. m., 3 p. m. and \ Temperature, rainfall, winds, ' Published in pamph- 10 p.m. clouds; Barometer and hygro. i let form from 1818 meter added in 1836. i to 1825 ; reprinted in I the American Alma- nac, 1834, and in I other publications. 7 a. m., 3 p. m. and Temperature, humidity, wind 9 p. m. direction and weather; rainfall added in May, 1836. I., 2 p. m. and 9 p. m. Hourly Sunrise, 10 a. m., 3 p. m. and 7 p. m. ■ a. m., 3 p. m. and y p. m. a. m., 3 p. m. and 9 p. m. Temperature, and rainfall. Nov.,1857toAug., 1859| 7 a. m.,3p. m. and 9 p. m. ^o^; 1894 to dat ■'«"■. 1871 to date Self-registering maxi- mum and minimum thermometer See page 303 for houri of observation ' For earlier locations and elevations see page 304. ometer elevatioi" Occasional reports to Franklin Institute of Philadelphia. Pressure, temperature, wind | direction and velocity, hygro- | Published in Trans. meter, cloud movement and Md. Acad. Sci., Vol. I, state of weather. 1837 Pressure, temperature, wind Printed copies of re- direction and force, dewpoint, ports for Jan., Mar., rainfall and weather. July and Aug. pre- sented to Maryland Academy of Science. Temperature, wind direction and raiufall. Pressure, temperature, winds, clouds, rainfall, relative humid- ity. Temperature, wind direction and rainfall. Temperature, wind direction, weather and rainfall. Daily observations at stated hours, since Jan. 1, 1&71 of pressure, temperature, wind direction and velocity, humid- ity, clouds and rainfall; tem- perature of water in harbor, Sept. 1881 to March, 1887; at- mospheric electricity, 1882 to 1887. Continuous automatic record of wind velocity since Jan., 1871 ; wind direction since 1874; rainfall since Nov., 1893; pres- sure and temperature since Jan., 1893 ; sunshine since July, 1893; humidity since Feb., 1903. 300 THE CLIMATE OF BALTIMORE each of the three institutions. Since 1896, the system of voluntary observing stations established in 1891, and later, by the State Service, and the publication of weekly and monthly reports, have been under the con- trol of the National Service. The special appropriation by the State has since been devoted to the investigation of special climatic problems and the publication of the results. The local office of the U. S. Weather Bureau is intimately associated with the commercial organizations of the city. The results of observa- tions and specially prepared reports are quickly put into the possession of those most benefited thereby through the instrumentality of the public press, by special bulletins, by telephone, or otherwise. Special efforts have always been made to place the daily telegraphic reports of weather conditions from all parts of the countrv' before the commercial interests at the earliest possible hour each day. These reports, as soon as received, are placed upon a large glass map upon the floor of the Chamber of Commerce, by means of symbols and lines; the Baltimore station was among the earliest to adopt this method of publishing the daily weather conditions. One of the most important duties of the local office of the Weather Bureau is the prompt distribution of the daily forecasts to the public, and of special warnings of the approach of storms, or cold waves, or the occurrence of frost. This is accomplished by a liberal use of the telegraph and the telephone, and by the hearty cooperation of the public press and the U. S. Postal authorities, especially through the instru- mentality of the recently established system of rural free delivery, by means of which the daily weather forecasts are being placed regularly each day in possession of all who dwell along the established routes, even those in remote farming communities. The shipping interests are informed directly by telephone of the approach of a severe storm; the owners of the smaller craft in the neighboring waters depend mostly for their information upon the dis- play of storm warnings at a conspicuous point along the harbor by means of special storm flags by day and lights by night. For many years, these . special warnings were displayed from the flag staff on Federal Hill ; they MARYLAND WEATHER SERVICE. VOLUME 2, PLATE XVIII. OFFICE OF THE U. S. WEATHER BUREAU, JOHNS HOPKINS UNIVERSITY, No. 532 NORTH HOWARD STREET. MARYLAND STATE WEATHER SERVICE. MAETLAXD WEATHER SERVICE 301 are now displayed from the top of a steel tower, 50 feet in height, erected upon the roof of the seamen's home, known as " The Anchorage," at the intersection of South Broadway and Thames Streets in East Baltimore. Eeports on the weather and crop conditions throughout the States of Marjdand and Delaware have been printed and widely distributed from the local oflSce each week during the season of crop growth from 1891 to the present time. Monthly and annual reports have also been issued during the same period, showing the daily weather conditions and the monthly and annual average values, together with the progress of crop growth and farm operations from month to month. A printed daily weather map (single sheets 11 in. by 16 in. in size) was issued from the local oflBce from October, 1896, to N'ovember, 1898. These received a free and wide distribution in the business portions of Baltimore and among educational institutions. The maps showed the weather conditions over the entire country, based on over a hundred telegraphic reports of obser- vations made each morning at 8 a. m., 75th meridian time. They reached the public about 1 p. m. each day. In November of 1898, the printing office connected with the Baltimore station was closed and the daily weather map was replaced by the larger and more complete litho- graphic map issued from the Central Office at Washington, D. C. Observers and Observations. The table on pages 298 and 299 contains a list of the individuals and institutions responsible for systematic instrumental observations of the weather within the city limits of Baltimore. The geographical location of the station, the period, and the character of the observations are also stated as far as known. There were doubtless other observers but their contributions to the observational literature of meteorologv' in Baltimore have not come under the notice of the writer. Instrumental Equipment. The instruments in use at the local station of the U. S. Weather Bureau since the organization of the Service in 1871 are enumerated in the following list; the dates of installation and discontinuance of tlie instru- ments are also jrivcn: 302 THE CLIMATE OF BALTIMORE Instruments In Use Since Anemometer, Robinson Jan. 1, 1871 Anemoscope Jan. 1, 1871 Barograph, Richard I Dec. 30, 1892 Barometer (Mercurial) Jan. 1, 1871 *Hygrograph, Richard Feb., 1902 Hygrometer, stationary Jan. 1, 1871 Psychrometer (whirling) July 11, 1886 Rain Gage (ordinary) Jan. 1, 1871 Rain Gage (self-registeriug, float) May 30, 1891 Rain Gage (self- registering, tipping bucket) , June 13, 1897 * Rain recorder, continuous Jan., 1903 Register, single (wind velocity) i Jan. 1,1871 Register, double (wind velocity and direction) Feb. 10, 1874 Register, triple (wind velocity, direction and rain). . Nov. 10, 1892 Snow Gage Jan. 1, 1871 Sunshine Recorder (added to triple register) June 30, 1893 Shelter, standard window ■ Jan. 1, 1871 Shelter, standard roof I Oct. 1, 1885 Thermograph, Richard j Dec. 30, 1892 Thermometer, dry bulb Jan. 1, 1871 Thermometer, maximum July 22, 1872 Thermometer, minimum July 22, 1872 Thermometer, water i Sept. 1, 1881 * Property of the Maryland State Weather Service. Use Discontinued July 11, 1886 June 20, 1897 Feb. 10, Nov. 10, 1874 1892 Sept. 30, 1885 Mar. 31, 1887 Hours of Observation. The prescribed hours of observation, as well as the kind of time em- ployed in the National Service, have been changed from time to time since the organization of the Bureau in 1870. Local time was in use at all stations from January 1, 1871, to July 31, 1881 ; on August 1, 1881, Washington time was adopted, making a difference, of two minutes between the local times of observation in Washington and Baltimore. Since January 1, 1885, observations have been made on 75th meridian time. The difference between Baltimore local time and 75th meridian time is six minutes, the former being slower than the latter by this amount. The combination of selected hours for observation has varied con- siderably in the thirty-four years since 1871. From 1871 to June 30, 1888, at least three direct observations were made daily, one at an early morning hour, from 7 a. m. to 7 :30 a. m., another in the middle of the afternoon, at 2 p. m., 3 p. m., or 4 :30 p. m., a third at night between 9 MARYLAND WEATHER SERVICE 303 p. 111. and 11 :3U p. m. Five observations were made daily for some years. Since July, 1888, there have been but two observations, one at 8 a. m., another at 8 p. m. The exact hours of observation and the duration of their employment are indicated in the following tabular statement : HOURS OF OBSERVATION. (U. S. Weather Bureau, 1871-1904.) Time of Observation. 11:87 p. u). (1. t, 11:02 p. m. (1. t 7:00 a. m. (1. t.) 2:00 p. m. (1. t.) 9:00 p. m. (1. t.) 12:02 p. m. (1. t.) (special) Oct. 22, 12:02 p. m. (I. t.) (daily) Feb. 23, .Telegraphic 11:02 a. m. (1. t.). 7:02 p. m. (1. t.) 7:02 a. m. (1. t.) 7:00 a. m.(w. t.) 3:02 p. m. (1. t.) 3:00 p. m. (w. t.) 11:02 p. m. (1. t.) 11:00 p. m.(w.t.) 2:00 p.m. (w. t.) 2:00 p. m. (7.5th m. t.) 7:(I0 a. m. (7.5th m. t.)^ 3:00 p. m. (75th m. t.) 11:00 p. m. (75th m. t.) 10:00 p. m. (7.5tli m. t.)J Water temperatures. . Tehii^raphic 11:00 a. m. (75th m. t.). 7:00 p. m. (75th m. t.). 8:00 a. K:00 p. (75th m. (75th m, !:;} Telefj^raphic Local lime (1. t.) 6 min. slower than 75tli m. . . . Wasliin^^ton (w. t.) 8 iiiin. slower than 75th ni. 7.5th meridian time Ended. Oct. 31, 1879 Oct. 31, 1879 Aug. 24, 1872 Dec. 31, 1884 June 30, 1881 Feb., 1872 Dec. 31, 1879 Dec. 31, 1884 Dec. 31, 1884 July 31, 1881 Dec. 31, 1884 July 31, 1881 Dec. 31, 1884 July 31, 1884 Dec. 31, 1884 Dec. 31, 1884 Mar. 31, 1887 June oO, 1888 June .30, 1888 Dec. 31, 1880 June 30, 1888 Aug. 3, 1886 Aug. 3, 1886 Cnrrent Current July 31, 1881 Dec. 31, 1884 Current 304 THE CLIMATE OF BALTIMORE Changes iisr the Location of the Station. Changes in the location of the observing station of the U. S. Weather Bureau have necessitated changes in the elevation of instruments. During a period of 34 years^ five different stations have been occupied. The office has always been in the thickly settled portion of the city; from 1871 to 1889 in the heart of the business section, later in the buildings of the Johns Hopkins University, No. 532 North Howard Street. The details of the changes experienced in station and instruments are indi- cated in the tabular statement below. With unimportant exceptions, the instruments have always had a fairly free exposure, as good, perhaps, as can be had in the midst of a large city. location of u. s. weather bureau stations and elevation of instruments. ci Above Ground (Feet). o o Observations begu 3 o 12; a) a -c 3 '3) a a ^•» — V < Location. a 2 03 P5 • a o a n J3 33 76 9 6c 08 a "3 P3 69 © a o a § > < _l < Q < a. O 05 q: Q CD Li- o O z 1- z o o CE < - ^ uT o 2 < a. O O I h- o 05 z MAEYLAXD WEATHER SERVICE 305 VARIATIONS IN THE ELEVATION OF THE BAROMETER AND CORRECTIONS TO THE EPOCH, JANUARY 1, 1900.* Barometer above reference plarte 25.3 feet. Reference plane, top of iron pipe underneath sidewalk on west side of Howard street opposite Center street, transverse station No. 482 of the City of Balti- more Topographical Survey, above mean sea level 98.0 " Station elevation of barometer, Jan. 1, 1900 123.3 " Date of change. Building's Occupied. 1870, Dec. 23 1889, Jan. 1 ! 1891, June 1 • I 1895, Sept. 7 1896, Aug. 1 1899, Jan. 1 Southwest corner of South & Water streets Southwest corner of Baltimore and Hollidaj- streets Johns Hopkins University (Physical Laboratory I 178.8 Southwest corner of Calvert and Fayette streets U1.5 Johns Hopkins University (No. 532 North Howard street) 45.2 75.9 123.3 ^ 3fe +30.7 +102.9 -37.3 -18.2 la 111 ;g|| Sgc +78.1 +47.4 -.054 -55.5 -18.2 + .063 + .020 -.015 -.015 -.015 -.015 -.015 -.103 -.069 + .048 + .005 -.015 .000 * For further information regarding the reduction of barometric observations see : Bige- low, F. H. The Reduction of Barometric Pressure Observations at Station of the United States Weather Bureau. Vol. II of the Report of the Chief of the L'. S. Weather Bureau for 1900. OFFICIALS IN CHARGE, U. S. WEATHER BUREAU OFFICE, BALTIMORE. (1871-1904) Name. Official title. Cowan, J. E. . . Penrod, II. J. . Boyd, W. T. . . Kabernagle, J. Bell, R. J McCiann, E. W. Black, W Seyboth, Robt. Felfrer, G. W. . Cronk, C. P. . . Marburv, J. B. Hunt, G. E. ... Walz, F. J. . . . Fassig, O. L. . . Sergeant Local Forecast Official Section Director Date of assignment. Jan. May Sept. April Dec. May June Oct. Sept. Oct. July July May July 1871 1871* 1874 1875 187.5 1877* 1879* 1879* 1882 1888 189.5 1896 1897 1900* Date of relief. Mar. 27, 1871 Sept. 23, 1874 April 24, 1875 Dec. 20, 1875 May 21, 1877 June 6, 1879 Aug. 29, 1879 Sept. 6, 1882 Oct. 26, 1888 July 6 July 1 May 22 May 29, 1900 In charge 1895 1896 1897 • During intervening intervals the station was in charge of the first assistant. 306 THE CLIMATE OF IJALTIMOKE 3 T " £■ S iz iC ii N CO 8 C5 s O £2 CO 1 ^ t- s 30 ^ o 1- le lO ^ t- -* O ^ OO s 5 (M s ^ -* TO •^ f2 lO s CO 0 X — r; « 5 '^ ?! N IJ — . lA ■^ .=< " *~ o Q «c O 00 55 o o o o ao X o 30 = = ^ CO t^ d -H i;; »o ss CO 1-1 >0 O C'J • - CO t- o > o s O C5 30 5^J o o o o CO d «c c: m m c: 3C cs N d d o CO ^ g^ 00 t- s: 2 2^5 • . O C5 O o O » g f-l o N O is o d I- « w d d Oi CO »l S^-Ss ^ : E^ H ® ft o (» o E S t- 00 o o o OS N ■* o »0 ^^ C5 30 lO O CO d d C5 s « sgg s O CO -* CO o ^ »! CO O CO -* O CO CO f- ^ ^ ^ N July o CO « c? ^ 2 ^ CO -* o S I- ^^^ X t- 8S S ~ S '^ "^ "^ >o c ■ s ° c lO 3: o i- fi t^ ■* " ci 30' d o CO co S5S s lO ^ >* N '^ o 5s CO « d t- o F-l 30 ©1 OS o N 55 S o CO t^ rH »3 r. lO 3 t^ »i CO : EH H => ft < 05 o °5 E c^ d S.I o -* o >0 .-1 OJ d d CO t- o >- (N I- CO CO 30 r^ -H CO •* 8S 00 o • d 00 o lO o ^ 3i lO o o o o =g CO CO d 05 ■* OS 05 » -; CO t^ -^ M C3 00 8S S (M d CO ^ 2 00 CO • lO d o "zi cc O 5= !C o o o o «o o 2 oc 30 O t- lO l- O CO CO t^ d -^ S ■* °^ lO 05 -. s O uj 30 o o o o 30 t- 5SS «s CO d 30 CO d d ©J 2 >* ^ _ CO CO r^ ; lo to d 5 + O Si + c ^ > o C ^ 2 § ^ „■ 5 £ ° *2 2 i- c • 2 ^ £ ^ o I d a, g w 5 -2 § K 03 OS i- Oi 0/ ei J* & £ OJ o S sj o 'Z 1 S a CJ . =s C3 . — r- ■ - c . 53 C3 • ^ ^ • >i >. • a ■ «— > 1— t ^ s . S^ o . X) ^ ■ a. O . • r^ ' V- V- ' C O . 'X tc M • c • • a ■ A J3 • • U ^ : • , . o o . — . tn c „ < =.< ■.>» r i c .. 5, -■^^— -^' -f-o - ^ - - - 1) S c S if = s .t; H 2 q - c — - ^ a ~ o o ~ CS fl Q t; "" a - .£' § ° M £ S = -2 >. a B C =3 -Z 5 ® I i I I S B .= - I E X X a X c :: a 1) « i 308 THE CLIMATE OF BALTIMORE 0 CB CD ■* rt 1-1 "-I N O >0 03 CO >0 o is 05 O -* O -^ 00 CO ?2 § 1 to o »o o 03 05 o CD e» OS CD CO OS CO ■* ■t-3 O O CO -t. ■ 00 l- 00 o -* ei t- : : : : EH CO s 2 ^ CO 00 ■* US «■» g? t- o CO ■* 00 «o o -p ■* (M i-H C! >a CO OS lO to * r-l 1— I C-l lO CO t- CO f-1 : H CO oi o cc t- I- CO t-^ OS ■* 00 lO CO lO OS >0 CO 'H OS lO OS ^- -^ ^' -* o ^ s CR o IC tr- « ■X. rt p. N o Ol o X § g ■-"" - to io OS ^ -* .-1 ^ CO OS 00 o 00 -* ■* « 00 lO o CO lO -* u: lO ^1 ■* -* CO o -* OS OS CO CO s o lO s s CO 00 -* »o 00 >o CO o s ■<* 00 CD e) d (S >-5 CO OS o o L^ 58 OS -* CO co OS CO 00 to 00 U3 CO o CO CO o E-( I & S ..^ u o " t- o • -5 -S a cJ rt >-; "3 "3 S t^ ^ u "? 5 •=> S 'S 9 c: c3 S g 5£ §S So O O 02 03 m (^ o o K 0)0 a s j; z z s m Z ci '5; ^ es > -^ o ■~' ^ 2Lzi o SL >-i C3 -l-i -1-1 w * n t. 1^ cj m t^ Q) o a> ci s- 2 to -o Q a a >> ■y * t-i a; r* a o a a o a /3 3 - — !^ a Z Z a « Z f* « Z Mo cj M t- eS K o) c3 a) o ci U QJ > U <2^ O iJ c-1 o t- ci »© O t- lO t- !C Ir- CI -^ iO «■ o 22.0 1.9 .00 3.00 NW ^ 1** : d -1 o 1-1 > o ^H (M "-H «o 5- S 8 > ^ 1^* : : d N O rt .w ^ -<* « O 05 '^ t- 1-^ W ^ 1** 00 t- s ■* !0 2 25 35 d I-} o o 4i c © CO t- t- S n «0 O ' rt ^ ^ s»s= rt lO O 1-1 C'^ C-l ..* IC x< »1 ^ O ^ 3C -* » S S ;> 2° ■'^^ cs » s* : • '■ o ^ o o o 3 ^ I- CC O I- rr t; o I- X 2° °S^ ^ 1^* = ~ >o ;^ = s -# M- ^ is^ 1-1 C:^ 1-5 OS ^^1^^ : ■ • to lO .-1 r-l CO 1 03 C3 00 » 00 21 « 5 "* "^ >0 M O N ^ ^ l.. 00 CO 00 O r-c (4 < lO § 1 2 g § ^ ^ 1^1 o ao o S *J '«' .* >o IN i-i «0 O CO t- lO O-l lO t- •O M § S ^ -* >o d re ^ ^ ^^ : • : d -* o CO 2 o o o ca o -* ^ N 1^ Jz; i^^ ■■ -* d ci o CO CO >0 O S5 CO TO (N 05 S S fe g3 ^ -= " z ^^^ : IN t- CO -^ N 2 CO a> d 1-1 o -* a o o t, 2 S s a S ■ 3 o cj *-* O "5 "^ ^' O O 1^ 3 r-l n O „ 3 3 S ^ rt ^ ^ y u o o r J J P p 5 ^ w 3 i a 5 ^ '^333 ?i = 3 3 _o _o _o g •- — "qj "o "c c o * o > > > ^ ^ £ O O O g -I "-I 5 a Q o t« to bi: !i< 3 3 3 3 3 a> ^ Z (» 02 M |S Z i-' r' ^ •-' ^ w Z ** 5 « 5 ^ " •sj ^ i« .< fc< 3 3 ^ S 2, a « a) m f.* t." 3 fa b C a! £ a a s 3 2^ am 7J oa 3 c e u 3 3 -S -3 t^ r >. H 5 ja §. 2:' >■ § §1 P 3 „ = S5 I s = H S = o 3 - .3 -r Z 03 X e ~ g O) ts i! c MARYLAND WEATHER SERVICE. VOLUME 2, PLATE XXL to -r O I- -- UJ < . E Oil ^ oo ^ §? = uj2 > Oz I I- — °- THE WEATHER OF BALTIMORE INTEODUCTION Leaving now the discussion of climate^, or the average and extreme values of the principal factors which constitute the sum total of atmos- pheric conditions, we come to a consideration of some of the more im- portant types of weather characteristic of the geographical horizon of Baltimore. As stated in the introduction to this report, the term weather is restricted in its use to the actual state of the atmosphere as regards temperature, humidity, wind movement, etc., at any given instant, or short period of time. The method employed in the discussion of cli- matic conditions is not applicable to descriptions of weather; the various factors cannot be considered separately, but must be studied in their relations to one another at a given instant of time, in order to afford a proper mental picture of actual conditions in nature. The past fifty or sixty years have witnessed a gradual but radical change in our views of weather conditions and sequences. Before the days of the telegraph, observers were isolated and independent of one another; we had but vague ideas as to synchronous weather conditions prevailing at distant points. Here and there in the eighteenth century we find a suggestion of the importance of co-operation in the methods and time of making observations ; but intercommunication was slow and the important discoveries of Franklin and Jefferson in America, of Bran- des in Germany, and others, met with rather tardy recognition, or were entirely overlooked and had to be rediscovered when times were more propitious for utilizing the results of new discoveries. The rich collection of weather proverbs, based upon natural signs — changes in the wind, forms of clouds, the habits of animals — are based upon the accumulated experience of individual effort. For centuries 21 312 THE CLIMATE OF BALTIMORE weather changes were minutely observed and carefully recorded. Espec- ially was this true of changes in wind direction, as this factor was almost universally regarded as the underlying cause of variations in the other elements. Not until the use of the telegraph became general, making it possible to gather reports from an area covering thousands of square miles, and to obtain a picture of actual weather conditions at the same instant of time over this area, was the true meaning of weather changes gradually revealed to the student of meteorology. Change in the direc- tion of the wind, while it still holds a conspicuous place in weather prognostics, is no longer regarded as the fundamental factor in the weather situation. The weather map, so familiar to us to-day, shows us that atmospheric pressure, or the height of the barometer, is the key to the problem of coming weather — not the actual height of the barometer at a single station, or at a number of stations, but the relative heights over a large area. Having given the relative distribution of pressure over a given area, the remaining weather elements can be supplied with a fair degree of accuracy by the expert student in weather forecasting. The Synoptic Weather Chart. The development of the synoptic weather chart, a chart showing the actual physical condition of the atmosphere at the same hour over an area of thousands of square miles, forms one of the most interesting chapters in the history of modern meteorology. Conceived before the close of the first quarter of the nineteenth century, the middle of the century witnessed a remarkably rapid development and application of the idea. This was brought about by the rapid spread of the electric telegraph and the recognition of the vast commercial importance of such a chart. The successive steps of its progressive development in this country and in Europe have been carefully traced by Abbe ^ and Hell- mann," to whom we are indebted for much of our accurate history of meteorology. ^ Abbe, Cleveland. The Development of the Daily Weather Map. Vol. I, Maryland Weather Service, Baltimore, 1899, pp. 225 et seq. ^ Hellmann, G. Neudrucke von Schriften und Karten iiber Meteorologie und Erdmagnetismus. No. 8, Berlin, 1897. MARYLAND WEATHER SERVICE 313 To-day the great majority of the nations of the world support a national weather service and issue such charts daily. We now have daily charts of synchronous observations for most of the land area of the northern hemisphere excepting Asiatic Eussia and China; and of the North Atlantic. In the southern hemisphere, there are charts for Australia, the Indian Ocean, South Africa, and the Argentine Eepublic. The time will soon come for the realization of one of the fondest hopes of the meteorologist, when we shall be able to construct such maps for practically the entire globe. The importance of the constant endeavor to extend the area of observations becomes apparent when we realize that there are no definite boundaries in the atmosphere of the globe, and that no extensive disturbance can take place in any portion of its vast extent without affecting, sooner or later, every other portion. We proba- bly do not yet realize fully the nice adjiistment of atmospheric forces, and we are still ignorant of many of the important laws underlying the larger atmospheric movements. Cyclones and Anti-cyclones. Before the advent of the synoptic weather chart, the constant changes in the direction of the Avind, the increase and decrease in cloudiness, the occurrence of rain with a certain wind and clear skies with another, were very little understood. Certain sequences in weather changes had long been accurately noted, but why these changes should follow a defi- nite order was incomprehensible. The weather map, however, revealed the clue to the interpretation of the changes in the relative distribution of atmospheric pressure over extended areas. It was soon learned iliat differences of pressure, or of the height of the barometer in neighboring localities, set the air in motion, causing it to flow from the area of high barometric pressure to the areas of lower barometric pressure, much as water is transferred from a higher to a lower level. This flow of the air from place to place in an effort to restore a disturbed equilibrium is what is termed the wind. Changes in wind direction in turn bring about changes in temperature, the wind blowing warmer with a change from a northerly to a southerly direction in the northern hemisphere, with inter- 314 THE CLIMATE OF BALTIMORE mediate changes in temperature wlien blowing from tlie east or west. A study of the weather map soon led to the formulation of a new set of rules of weather changes, more general in their application and more in- telligible than those based on the study of observations at a single sta- tion. Two distinct types of weather were soon recognized. The most conspicuous of these, and the first to be investigated was the storm area, an area in which the readings of the barometer decrease rapidly from all sides to a minimum in the center of the area. Such areas were observed to be accompanied by cloudy skies, and more or less rain, by winds in- creasing in force as the center was approached, and blowing approxi- mately toward the point where the barometric pressure was lowest. As the types of weather first investigated were naturally well developed storm areas, the winds high and blowing in paths nearly circular, or at least spirally inward, the term cyclone was applied to them; a term first used about the middle of the nineteenth century by Captain Henry Piddington to describe revolving storms in the Indian seas. Later, the term cyclone was given to all atmospheric disturbances of wide area in which the winds blow toward a central point or line of low pressure. Another weather type which later claimed the attention of students of meteorology was the fine weather type, in which the barometric pres- sure is highest in the central area, decreasing in all directions from the center outward. In these areas the winds were observed to blow in general away from the center of highest pressure. As the character- istics of these areas were in many respects the opposite of those observed in cyclones, they were given the name anti- cyclones. Between these two well defined types, there are innumerable forms par- taking more or less of the characteristics of one or flie other of the two principal types. In the northern and southern hemispheres, from latitude 30° to 70°, the upper portions of the atmosphere apparently flow in a constant stream in a direction approximately from west to east around the globe. The winds within the lower layers of the atmosphere in the middle lati- tudes do not always follow the course of the upper currents; the atmos- phere from the earth's surface to the level of the highest clouds is broken MARYLAND WEATHER SERVICE 315 up into areas in which the barometer is alternately low and high, into cyclones and anti-cyclones, as they are generally designated. These alternate areas of unsettled weather and tine weather are carried along as a whole in an easterly direction with the general drift of the upper atmosphere, constantly changing in form and intensity as they move, but retaining, in the main, their chief characteristics for thousands of miles ; the areas of high pressure are accompanied by comparatively little cloudiness, and temperatures below the seasonal average; while the areas of low pressure are attended by clouds and rain and a temperature above the seasonal average. These cyclonic and anti-cyclonic areas of the middle latitudes have a diameter varying from a few .hundred to a thousand, or even two thousand, miles and move eastward, as a whole, with a velocity averaging about 600 miles per day, across tlie United States, and hence occupy two or three days in passing a fixed point. Their passage east- ward explains the constant shifting in the direction of the winds of a given locality, the direction of the change in wind, as will be explained later, depending upon the position of the center of the anti-cyclones and cyclones with reference to the given locality. For instance, when a " low," or cyclonic area, approaches Baltimore from the west, if the center is north of the latitude of Baltimore, the wind becomes easterly ; as the center passes Baltimore, the wind shifts to the south, and then to the west or northwest. If the center of the storm passes to the south of the city, the changes in the wind direction are successively, east, north, and west, the reverse of those in the first case cited. If the center passes over Baltimore, the easterly wind is followed by a calm, or light variable wind, after which the wind will spring up abruptly from the west. These rules of weather sequences are applicable only to the well devel- oped and definitely formed areas of high and low pressure, and but im- perfectly apply to the more numerous moderately developed " highs " and " lows " whose eastward drift gives us our daily routine of weather changes. A clear conception of the character of cyclones and anti-cj'clones, as described above — of the distribution of pressure, the system of winds, the distribution of temperature, tlie state of the weather, and the move- 316 THE CLIMATE OF BALTi:\IORE ments of these areas, as a whole — is essential to a proper understanding of our daily weather changes, and especially of the more conspicuous weather types known as storms and cold waves. The essential features of cyclones and anti-cyclones may be most readily understood by a study of actual examples of the simpler well developed types. As an illustra- tion of a typical storm or cyclone, the weather chart of the morning of December 27, 1904, is reproduced in Figs. 85, 86. To those unfamiliar with the weather majD — and indeed to all excepting those who have de- voted much time to their study — the usual weather map is a confused tangle of lines and symbols requiring careful explanation and analysis before even the essential features of the weather conditions are under- stood. Some of these difficulties may be obviated by the use of a series of charts portraying the separate factors which go to make up the com- plex weather conditions, retaining in each chart, however, the controlling factor, namely, the system of lines representing the distribution of atmos- pheric pressure, or isobars, as they are called. AREAS OP UNSETTLED WEATHER (CYCLONES). It may be well at this point to call attention to the very frequent misuse of the terms cyclone and tornado. In scientific literature there is a clear distinction between the two, while in the popular mind, they are often synonymous. The confusion in the use of these terms is natural, and is largely due to a change in the meaning of the word " cyclone " as used in technical literature. A cyclone has at all times, since the days of Piddington (about 1850), been regarded as a severe and destructive storm. Later when the natiire of storms and weather changes was better understood, the meaning of the term was enlarged by the student of meteorology to include all atmospheric disturbances, whether large or small, intense or barely perceptible, in which the winds blow inward toward a central point, or area of low barometric pressure. Tlie general public has not yei adopted this amended definition, and all intense storms continue to be called cyclones, when the particular storm in mind may be a tornado, squall, an intense thunderstorm, or a hurricane. While the tornado, as a revolving storm, must be classed with cyclones, the ■;r service baouboiqsi gjisrio lariJBSw x^iBb .8 /J oJ aailqqe nobfinBlqxa sniwoIIoH 9riT Jb 9f)jBm snoitBvieado giroeaBlIumra no baaBd 91b eJiBrio sriT JioqsT siriJ ni .9mii flBibiiam riJ6Y ,.m .s 8 .(.irfB'? 8991395 ni) 9'iiiiBieqtnel Isupe lo 89nil sib asnil jlosia .(aerfoai ni ) eiuaaaiq oiierfqaomJB lBup9 5o 39nil 91b asnil b9fl Fir. ii5.— T-??i^ai^?«^»5?*)yft' iff g:M£ttiiaftiiJ 3lBm,'(8B9iB b9bBfl8 .bniw offt rfJiw xft awoiiB 9dT .niBi asiBoibni H .wona asJBOibni 8 21 3nib909iq s^J gniiuh rmoJaiebnnrii 6 lo 9oa9ni;90o 9rfJ 89:tB0lbfli T by the iv LEGEND. The following explanation applies to U. S. daily weather charts reproduced in this report. The charts are based on simultaneous observations made at 8 a. m., 75th meridian time. Black lines are lines of equal temperature (in degrees Fahr.). Red lines are lines of equal atmospheric pressure (in inches). Shaded areas mark the limits of overcast skies. The arrows fly with the wind. R indicates rain. S indicates snow. T indicates the occurrence of a thunderstorm during the preceding 12 hours. } clone ^' -,..,.: the '^ — ■1 rlcc;ff^,. ' ^enerfil MARYLAND WEATHER SERVICE 317 Fig. 85.— Typical Cyclone of December 27, 1904. Fio. 86.— Typical Cyclone of December 27, 1904. 318 THE CLIMATE OF BxVLTIMORE term is restricted to the intense and very destructive local storms of very small area occurring within certain limited portions of a larger storm; it is a cyclone within a cyclone. The accompanying charts, Figs. 85 and 86, show the distribution of pressure, wind direction, temperature, cloudiness, and precipitation, within the area of the typical cyclone which passed over the United States during December 27, 1904. Pressure and Winds. — The series of red curved lines show the dis- tribution of atmospheric pressure. The inner, nearly circular, line en- closing the word " low " in the chart for December 27 is drawn through localities in which the barometer read 29.40 inches, the lowest reported pressure. The remaining curves connect localities in which the barome- ter stood higher by successive intervals of two-tenths of an inch, until around the outer limits, at approximately a thousand miles from the center, the barometer read 30.40 inches, showing a difference in pressure of one inch. The atmosjihere is very responsive to local differences of pressure. With very slight differences, a small fraction of an inch, for example, there will be set up a movement of air from the locality having the higher toward that having the lower reading of the barometer until equilibrium is restored, for the same reason that water always has a tendency to flow from a higher to a lower level. Bearing in mind this general physical law of the flow of gases, we may understand the general drift of the atmosphere toward the central area of low pressure in a cyclone. This is clearly shown by the arrows which indicate the direction of the wind at so many stations of observation; the arrows fly with the wind and point in a general way toward the area of lowest pressure. As will be observed they only occasionally point directly toward the center; in most instances the direction taken by a particle of the atmosphere in its journey from the outer portion of a storm area toward the center is along a spiral course. This is due to the effect of the rotation of the earth about its axis, which always tends to urge a freely mo'ving particle toward the right of its initial direction, in the northern hemisphere. The topography of the region over which the storm passes, and more ]MARYLAND WEATHER SERVICE 319 important still, the distribution of the pressure about the central area as shown by the curved lines, or isobars, also greatly influence the direction of the wind. In general, and especially over land areas, the isobars vary widely from the circular form, and the actual wind directions fall roughly into two classes, easterly and westerly. Drawing a north and south line through the center of the cyclone, the winds to the east are observed to flow mostly from some point between northeast and southeast, while those to the west of the line blow mostly from some point between south- west and northwest. The winds blowdng directly from the south or from the north are not so frequent, or of as long duration as those from other directions. The effect of pressure distribution on the direction and force of the \\-'w.d will be brought out very clearly in later discussions of weather types. In general, it may be stated that the force of the wind is greater the greater the difference of pressure between two neighboring points ; that is, it is proportional to what is called " the gradient,'^ which is equivalent to difference of level in the flow of streams; the steeper the bed of the stream, the more rapid is the flow of water. Temperature and Wind Direction. — Especial attention is directed to the relation existing between wind direction and temperature in a cyclone. Localities reporting the same temperature at 8 a. m. are- joined by means of lines, or isotherms, as they are styled. The lines are drawn at intervals of 10° or 20° Fahr. It will be observed that the temperature of 70° above zero in lower Florida gradually diminishes, as we proceed northward, to 30° below zero in the extreme northern limits. This is the usual direction of decrease in temperature at all seasons of the year, though tlie rate of decrease is here much more rapid than under normal conditions. In the absence of a well defined atmospheric disturbance the isotherms run nearly parallel with the lines of latitude; a steady and fairly uniform decrease in temperature from south to north is the normal condition all over the northern hemisphere. The relative temperature of winds from different quarters may vary greatly in different localities, but in the main, a southerly wind is warmest and a northerly wind coldest, with intermediate degrees for the east and the west wind. Hence we may readily understand how a change in the direction of the wind 320 THE CLIMATE OF BALTIMORE may affect the temperature of a locality. On the approach of a cyclone, in most cases from the west, the winds to the east of the center become easterly, veering to southerly; the temperature rises in the eastern half of the storm, and particularly in the southeast quadrant where the winds are from southeast or south. In the southwest quadrant of the storm there is usualh- an abrupt change from the warm south or south- east wind to a much colder west or northwest wind. This condition of temperature distribution is strikingly exhibited in the charts. The isotherms are seen to bend northward far beyond their seasonal values in the southeast quadrant; on the other hand, in the southwest quadrant, the cold northwest winds are carried far beyond their normal limits to the southward. The result is a stronger contrast in temperature from the center of the storm westward than from south to north ; the isotherms run north and south, in the particular case cited, and not east and west as in normal weather conditions. These shifts in the direction of the wind during the passage of a cyclone are sufficient cause for the great majority of the temperature changes experienced in our latitudes. Distribution of Clouds and Precipitation. — The proportion of the area covered by clouds at 8 a. m. is indicated by the extent and intensity of the shading. It will be observed that in practically all of the region within the influence of the system of closed isobars, the skies were en- tirely overcast, and that over an area extending 500 to 600 miles in nearly all directions from the central point of lowest pressure, precipi- tation occurred at the hour of observation — rain to the south of the isotherm of 30° and snow north of this line. The area of precipitation in this particular storm was exceptional, but it serves to illustrate the general law of the distribution in well developed storms of large extent. In most cases the area of precipitation is more limited in extent and is surrounded by a band of overcast skies, beyond which is a partly clouded band merging into regions of clear skies. Not all cyclones, even when well developed, show the symmetry in the distribution of weather con- ditions indicated in the type selected; the variations are infinite, but there is a general conformity to the type when the storm is well developed. The center of the rain or snow area is usually to the east and south of MARYLAND WEATHER SERVICE 321 the center of low pressure. The details of the distribution and the character of the preciiDitation will be brought out more clearly in the later discussions of weather types of the season. The elements described separately in the preceding pages are finally brought together upon a single chart, the usual form of presenting in our daily weather charts the actual condition of the weather at a stated hour. Such charts, showing the actual state of the weather at 8 a. ra. throughout the United States and the British Provinces to the north, are issued daily about 11 a. m. by the United States Weather Bureau, based on telegraphic reports from about 175 stations. AREAS OF FAIR WEATHER (aNTI-CYCLONES) . As already stated in a preceding paragraph, the term anti-cyclone was first used to describe a weather type which shows characteristics just the opposite of those of the cyclones. The pressure is highest in the centre of the area, the winds blow in a general direction away from the center, the skies are mostly clear to partly clouded, with little or no precipitation, the temperatures are, in general, loiver at the center than on the eastern or western sides of the area ; that is, their isotherms curve southward toward the center of high pressure while those in the cyclone bend northward toward the central area of low pressure. A typical example of an anti-cyclonic system is shown in the weather' map of April 4, 1904, reproduced in Figs. 87, 88. It occupied approxi- mately the same position and covered the same area as did the cyclone of December 27, 1904, shown in Figs. 85, 86. Isobars and Winds. — The inner circle, or isobar of 30. GO inches, marks the central area of the anti-cyclonic system, from which there is a steady and uniform decrease in the height of the barometer outward in all directions, the successive isobars marking intervals of two-tenths of an inch in the height of the l)aromoter, as in the case of the cyclonic system described above. It will be observed that the gradient, or steepness of the successive steps between isobars, is less in the anti- cyclone than in the cyclone : the area covered by each system is approxi- mately the same, wliilo tlio total difference in pressure between tlie center 322 THE CLIMATE OF BALTIMORE Fig. 87. — Typical Anti-cyclone of April 4, 1904. LOW Fig. 88. — Typical Anti-cyclone of April 4, 1904. LEGEND. The following explanation applies to U. S. daily weather charts reproduced in this report. The charts are based on simultaneous observations made at 8 a. m., 75th meridian time. Black lines are lines of equal temperature (in degrees Fahr.). Red lines are lines of equal atmospheric pressure (in inches). Shaded areas mark the limits of overcast skies. The arrows fly with the wind. R Indicates rain. S indicates snow. T indicates the occurrence of a thunderstorm during the preceding 12 hours. IMORE .aviaoaj b^^ubo^qe^ siiBdo ladisevr ^liBb .8 .U oJ asilqqE noiiGaBlqxs gniv/ollol 9riT is 9bBm 8noiJBvi98do auosnBlIumia no baaBd sib sJibiId sriT Jioqai sidJ ni .sffliJ flBibiisra riJ6T ,.ai .b 8 .(.idB'K 899is9b fli) 9iuJjsi9qm9J lBup9 lo ssnil 91b asnil >lDBia .(asrioni ni) eiuaesiq oiisriqaomJB lBup9 5o 89ai{ 91b asnil b9fl i>'i- — i yp.ae!j(a' 16B3t9'7^ 5o atifhiT '^rf J jiTfim aB9'iB b9bBri8 .bniw 9ri} rfJiw yR bwoiib 9rIT .ni£i a9:tB9ibni H .wona 89iBDibai 8 SI snib909iq> sdi sniiub cmo:t2i9l)niJri} b Jo eoaa-nvooo 9riJ 89iA0ibat T .STUOfl Fig. 85.— Typi. clone of April 4, 1904. MARYLAND WEATHER SERVICE 333 and outer circumference in the case of the anti-cyclone is only half that in the cyclone, namely, one-half an inch. This is also shown by the number of the isobars, the cyclone having double the number shown in the anti-cyclone, while the successive steps of increase or decrease in pressure, and the entire areas covered by the two systems are the same. This is characteristic of the two types, though the proportions may vary greatly. The winds are observed to blow in a general direction away from the center. As the winds in an anti-cyclone are in general much lighter in force than in a cyclone, the actual directions recorded near the surface are influenced to a greater extent by local topography. This is especially marked near the center of the area where the winds are very light. The law of deviation of a particle of air to the right of the initial direction flowing from the center of the high area outward holds good, as noted in the discussion of the inward flow in cyclones; hence the arrows, indicating the direction of the wind in the anti-cyclone, are observed to point, not directly from the center but to the right of the radial path, the angular deviation depending upon the configuration of the isobars, the topography and other factors. The Winds and Distribution of Temperature. Especial attention is directed to the relation between wind direction and temperature. Here again, as in the case of the cyclone, the temperature is seen to be directly dependent upon wind direction. In those portions of the area where the winds are mostly from a warm southerly direction, particularly noticeable in the northwest quadrant in the illustration shown, the isotherms are bent far northward of their normal position for the season. In the northeast quadrant of the area, where the winds are mostly from the colder north, the isotherm of 30° is seen to dip far to the south. Along the Atlantic coast the cold of the northerly wind is considerably tempered by the presence of the ocean, over which the rate of change in temperature is much less marked than on land along a north and south line. In the center of a " high," another factor enters to lower the tem])era- ture below the normal for the season. Here the skies are clear and 324 THE CLIMATE OF BALTIMORE radiation from the surface of the earth is rapid. This is especially true during the night hours and^ in consequence, the effect is quite marked on a chart based on observations made at 8 a. m., before the heating effect of the rising sun becomes marked. As a result of the conditions described, the lowest temperatures in a well developed anti-cyclone are generally observed to be within the central isobar of a system, if meas- ured by departures from the normal seasonal values for a given locality. Distribution of Clouds. — The chart selected as an example of an anti- cyclone shows almost too well one of the characteristic features of this type of weather, iiamely, the absence of clouds. While freedom from precipitation and clouds is the most striking difference between a cyclone and an anti-cyclone, it is not usual to see a weather chart with a high area of so great an extent with skies practically free from clouds, as is here shown. Over an area embracing all the states and the Canadian Provinces, from the Mississippi Valley eastward to the Atlantic coast, an overcast sky was reported at 8 a. m. from only four or five observing stations. Usually, even in the well defined and well developed areas of high pressure there is a fair percentage of cloudiness, excepting within a radius of two hundrea ^r three hundred miles from the center. The presumption is that this freedom from clouds and precipitation in an anti-cyclone is due to a descending atmosphere, with attendant increase of temperature, due to compression and consequent decrease in the relative humidity; the opposite process, namel}^, a rising atmos- phere cooled by expansion, accompanied by an increasing relative humid- ity and by cloud formation, and later by rain or snow, marks the cyclone. THE EASTWARD DRIFT OF CYCLONES AND ANTI-CYCLONES. In describing the distribution of presfeure, winds, temperature, and clouds in typical c3^clones and anti-cyclones in the preceding paragraphs no reference was made to their movement as a whole. The systems are not stationary for any length of time. In addition to the internal circulation of the winds described, the entire systems are carried east- ward by the general drift of the upper atmosphere in the middle lati- tudes, retaining at the same time their chief characteristics as cyclones MARYLAND WEATHER SERVICE 325 and anti-eycloues for many hundreds, and sometimes thousands of miles. The small whirls formed in a rapidly flowing river and carried down stream with the general current, while at the same time maintaining their own gyratory motion, are often cited as illustrations of the drift of cyclones and anti-cyclones in the general eastward flow of the atmos- phere between the parallels of 30° and 70° north and south latitude. These " highs " and " lows " do not extend to a great altitude, but are Fig. 89. — Typical Cyclone and Anti-cyclone of March 3, 1904. formed apparently only in the lower portions of the atmosphere, the great majority of them being confined to the atmospheric strata below the highest mountain ranges. They are carried in an easterly direction in the middle latitudes with a varying velocity but averaging about 600 miles per day. wliile moving across the United States. There is a con- stant and rapid succession of these atmospheric whirls, or waves of high and low pressure, in the winter and spring season. In summer and early fall they arc loss frequent and not so well developed. There is 326 THE CLIMATE OF BALTIMORE an excellent example of a well developed " low " or cyclone, followed by a "high." or anti-cyclone in the chart for March 3, 1904. (Fig. 89.) The shifts in the direction of the wind experienced during the passage of these " lows " and "highs " may be illustrated upon this chart by noting the directions of the wind at points along a given parallel of latitude passing from east to west. The nature of the change of wind depends upon the position of the center of the cyclone or anti-cyclone with reference to the parallel selected. Taking the latitude of 40 '^ for example in the chart for March 3, 1904, we have, as we pass westward from the Atlantic seaboard, first a southeast wind followed by a small area of south wind, in AVest Virginia, followed by a northwest wind to the center of the succeeding " high " in western Nebraska. To the west of the center of the anti-cyclone we have again a southerly wind in Colorado and Utah. Selecting a parallel of latitude north of the center of the " low " and " high," we have a reversed order in the shift of the wind. In the " low " the change is from easterly to westerly by way of the north ; and in the " high " from westerly to easterly by way of the south. Similar shifts in the wind are experienced in a fixed locality, such as Baltimore, for example, as these cyclones and anti-cyclones approach and pass beyond the observing station. An easterly wind at Baltimore heralds the approach from the west or southwest of a more or less developed cyclone, or storm area, followed by increasing cloudiness and rain or snow, as the center of the storm approaches. After the wind veers to the south and then to the southwest the precipitation soon ceases, the solid cloud mass begins to break into patches of cloud and, as the wind gets into the west, the proportion of clear sky increases until the cyclonic system passes beyond the horizon in the east. This is the usual order of change. With the path of the storm center to the south of Baltimore, the wind backs from east to west by way of the north. The weather types described above are of unusual symmetry. The forms met with in our daily routine of weather conditions are infinite in variety. No two are exactly alike in all their details of pressure, temperature and cloud distribution, or in the paths pursued, but there are many easily recognizable types with marked family resemblances which are of great MARYLAND WEATHER SERVICE 327 assistance to the practical meteorologist engaged in weather forecasting. Some of these types will be discussed in detail in the following pages. WEATHER CHARTS OF THE NORTHERN HEMISPHERE. A remarkable series of daily weather charts covering a period of ten years was prepared and the results published under the auspices of the Fig. 90. — Pressure Distribution over the Northern Hemisphere, Dec. 4, 1886. United States Weather Bureau, then known as the Signal Service, from 1878 to 1887. The charts were based on reports received from co-operat- ing national weather services and covered the whole of the northern hemisphere between the latitudes of about 20° to 65°, excepting the 22 328 THE CLIMATE OF BALTIMORE Pacific Ocean. One of these cliarts, sliovving tlie actual distribution of pressure at noon, Greenwich time, for December 4, 1886, is reproduced in Fig. 90. The chart shows clearly the manner in which the lower atmosphere of the middle latitudes is segregated into successive areas of low and high pressure, or cyclones and anti-cyclones at a given hour. Weather of the Prixcipal Climatic Zones. It has been customary for convenience to divide the surface of the globe into three climatic zones, the tropical, the temperate, and the polar, separated by fixed parallels of latitude and based upon the altitude of the sun above the horizon. The climatic conditions experienced within these zones have no such definite boundaries. When we come to the question of daily weather conditions, it is even more difficult to assign any fixed limits to areas of characteristic weather types. Still, it is possible to designate a number of zones, in the central portions of which the weather conditions are sharply marked off from conditions in neigh- boring zones. the tropical zoxe. The climatic belt designated as the tropical zone has several sub-zones of characteristic types of weather. The entire zone is marked by a uniformly high temperature, but the moisture conditions and atmos- pheric movements vary greatly in neighboring regions. The temperature changes from day to day, or from season to season being very small, the seasons are marked by a varying frequency or quantity of rainfall, or by a change in the direction of the wind. One day is very much like another the year round, and the weather cycle is the daily cycle, offering a strong contrast with the rapid fluctuations experienced in more northern latitudes. The doldrums, or equatorial calms, are characterized by high tempera- ture and humidity, light winds or calms, much cloudiness and frequent and heavy rains, and almost daily thunderstorms — a combination caus- ing an oppressive and debilitating atmosphere. MARYLAXD WEATHEK SERVICE 329 To the north and south of tlie doldrums are the trade wind belts. Here the skies are mostly clear, while a fresh, dry, northeast or south^\"est wind, strongest over tlie ocean, l)lows steadily toward the equatorial belt of highest mean temperature. Beyond the northeast and southeast trades, there is another belt of light winds or calms, the so-called "horse" latitudes; these are areas of permanently high pressure, clear skies and warm dry air, resembling in many respects the summer anti-cyclone of the middle latitudes. Within this belt most of the great desert areas of the earth are formed, in the southern as well as the northern hemisphere. The moving cyclones and anti-cyclones, described in preceding pages as characteristic of temperate zone weather, are conspicuous by their ab- sence in most portions of the hot zone. In some portions, notably in the West Indies, the Philippines, and the Indian seas, cyclones of great inten- sity occur during the late summer and early fall, the well-known hurri- canes and typhoons, which are carried in a westerly direction by the general drift of the atmosphere in the equatorial regions ; but they are of infrequent occurrence when compared with the constant succession of temperate region cyclones. THE TEMPERATE ZOXES. In the middle latitudes, north and south of the equator and extending beyond the Arctic and Antarctic circles, the weather is completely domi- nated by the moving cyclones and anti-cyclones described in preceding paragraphs. Here the daily monotony of tropical weather is replaced by great variability of temperature conditions which mark the seasons, and by the more rapid fluctuations which accompany the passing of cyclones and anti-cyclones. Tropical heat succeeds polar cold and all the weathers of the globe are brought to our doors during the course of a year. These contrasts become more and more marked as we approach the central por- tions of the great continental areas of the northern hemisphere. In the extreme nortliwest of our own country and in the British northwest territory, the breeding ground of cyclones and anti-cyclones, the contrasts in temperature experienced at a single station within a few hours, or 330 THE CLIMATE OF BALTIMORE within very limited areas at the same hour, are sometimes marvelous. On the 10th of February in the year 1899, an anti-cyclone developed over Montana and the British territory just beyond. It spread rapidly over the United States as one of the most intense cold waves ever experienced in this country,, lowering the record of intense cold in many states in its progress southeastward to the Gulf of Mexico and the Atlantic coast. On the morning of the 10th a minimum temperature of 65° below zero was registered in the western part of Montana. Just west of the moun- tains in the neighboring state of Washington, the temperature at the same hour was 63° above zero, a difference of 128° between two points along the same parallel of latitude (50° north) less than 300 miles apart. The southeastward progress across the United States of some of the more marked cold waves is frequently attended by strong inversions in the normal distribution of temperature along the Atlantic coast. The front of the cold wave, with its cold northwest winds, may reach Florida from 12 to 24 hours before it is felt in New England and the British maritime provinces, which may be in the center of the well developed cyclone which frequently precedes the cold wave. Under such conditions, a strong southerly wind will blow along the north Atlantic coast and raise the temperature high above its normal seasonal value for these coasts, while the Gulf states are dominated by the intensely cold northwest wind of the anti-cyclone. In such cases, temperatures of 20° above zero or less are experienced in northern Florida while the warm southerly winds blowing over Newfoundland raise the temperature to 40° or 50° above, although the latter region is nearly 25° of latitude farther north than Florida. THE POLAR ZONES. While we are less familiar with the weather conditions of the extreme north and south portions of the globe than with those of the hot and temperate zones, we have abundant evidence that within the Arctic and Antarctic the passing cyclone controls conditions to the highest latitudes attained. The fluctuations in temperature are very great, with changes MARYLAND WEATHER SERVICE 331 in the direction of the wind, while the cold is usually intense. While there are apparently bright, clear and exhilarating days, the weather is mostly gloomy with a high humidity and frequent fogs, sleet and snow. The intense cold weakens the powers of resistance. With these disa- greeable weather conditions predominating there is the added gloom of long-continued darkness, the physiological effects of which are exceed- ingly distressing. THE SEASONS. In the middle latitudes, and particularly over the continental areas, the most conspicuous feature of the advance and retreat of the seasons is the marked rise or fall in the mean temperature from month to month. Take, for example, the annual rise and fall of the thermometer at Balti- more as shown by Fig. 25 on page 111 ; the curves b, c, and d are based on the mean monthly maximum, the normal monthly average and the mean monthly minimum temperatures respectively. The lowest temperatures occur in the months of January and February; from this portion of the curve there is a steady rise at a fairly uniform rate to the month of July, followed by an uninterrupted fall to midwinter. The smooth, simple curve represents a uniform increase and decrease in the power of the solar rays as the sun increases and decreases in altitude in the annual revolu- tion of the earth about the sun. This uniform increase and decrease throughout the year is still shown by constructing the annual curve from the mean temperatures of successive five-day periods. (See Table XVIII, page 89.) If, however, we represent the advance and retreat of the seasons by curves based on mean daily temperatures in place of mean monthly temperatures, we find a striking difference in the character of the two sets of curves, as is clearly shown by consulting Plate III. There is no such uniformity in the daily progress of temperature; the serrated appearance of the curve indicates clearly that the progress is marked by successive temperature waves having an average period of three or four days. The annual curve is broken up into a series of subsidiary curves of short but irregular periods, due to the constant succession of cyclones and anti-cyclones with their accompanying large fluctuations in tempera- 332 THE CLIMATE OF BALTIMORE ture. By substituting the actual temperatures experienced upon each day of an}'- given year, in j^lace of the meari daily temperatures for a long series of years, the irregularities of the annual curve become enor- mously increased. The extent to which the temperatures experienced upon a given day of the year have varied in past years is shown in curves A, C, and D in Plate IV. For example, on the 11th day of February, 1899, the luinimum temperature at Baltimore was 6° below zero; on the 11th of February in 1887, the maximum was 72° above zero, an extreme range of 78° for the 11th of February. Even in the summer months, when the variability in temperature conditions is least marked, the ex- treme ranges are about -10°. While differences in temperature constitute the most conspicuous fea- ture of the weather of successive seasons in most portions of the middle latitudes, the character and amount of precipitation, and the duration and the force of the wind are factors of great importance, and indeed these sometimes overshadow the temperature changes. The departures from the normal conditions of temperature, precipi- tation and wind experienced in a given season, must be referred back to the prevailing type of pressure distribution upon which they depend. A clear knowledge of the relative distribution of pressure over a widely extended area is essential to a proper understanding of the weather changes in any given locality; this knowledge should extend, not only to the rapidly moving cyclones and anti-cyclones, but also to the larger areas of high and low pressure known as permanent cyclones and anti-cyclones, which are a direct result of the general atmospheric circulation.^ As a result of the increasing cold of the winter months there is formed over the Xorth American continent a vast anti-cyclonic area. AVith the passing of the winter, this gradually disappears to give place to a baro- metric depression, or cyclone. The process is reversed over the neighbor- ing oceans; here, while the contrasts between the winter and sumiuer pressure distribution are not so marked, the pressure is lower in winter ^ See: Teisserence de Bort; Etude sur I'hiver de 1879-80. Ann. du Bureau Centr. Met'L. Paris, Vol. IV, 1881. MARYLAND WEATHER SERVICE 333 than in summer. The changes in the intensity and in the position of these great atmospheric systems have a direct influence upon the char- acter of the seasons over the eastern portions of the United States, and especially in the Atlantic coast states." The best developed and most conspicuous instance of this semi-annual transfer of vast quantities of air from the continent to ocean during the winter and from ocean to con- tinent during the summer, is seen in the winter and summer monsoons over India and the Indian Ocean. Bearing in mind what has been said concerning the influence of pres- sure distribution on the direction of winds, and hence on temperature and precipitation, we may realize how variations from the normal type of pressure distribution for a given month or season will affect the general character of the weather of the period in question. This influ- ence may be graphically shown by calculating the mean monthly distri- bution of atmospheric pressure over the North American Continent and adjacent oceans during an abnormally cold month and an abnormally warm month, and charting the results in connection with a map showing the normal distribution of pressure based on a long series of years of observation. This has been done in succeeding pages for the months of January, April, June, and October as types for the winter, spring, sum- mer, and autumn seasons respectively. In the succeeding pages some of the more conspicuous types of cy- clonic and anti-cyclonic control of the weather of the Middle Atlantic states will be considered in connection with a discussion of the seasons in which they most frequently occur. WINTER WEATHER. The winter season presents the most variable weather conditions of the year. Practically every type of weather may be experienced at one time or another during the course of the three months, and sometimes a great variety of types may be crowded into the short period of 24 to 48 * See: O. L. Fassig. Types of March Weather in the United States. Amer. Journ. Sci., New Haven, November, 1899, Vol. III. 334 THE CLIMATE OF BALTIMORE hours. A description of winter weather conditions which prevail in the vicinity of Baltimore would include all the types of the year, though some of them attain their greatest development in other seasons. An account of weather conditions from day to day in our latitudes is mostly confined to a consideration of the eastwardly moving procession of cyclonic and anti-cyclonic systems across the United States. While for purposes of convenience and clearness our descriptions are confined to well developed types, the fact must not be overlooked that the faintly developed systems are the most frequent and consequently in the long run determine the general character of the weather of a given locality. All of our weather types may be roughly separated into two fairly distinct classes — (a) areas of unsettled weather accompanying the pas- sage of cyclones, or areas of low pressure — (b) areas of fair weather asso- ciated with passing anti-cyclones, or areas of high pressure. While it is often difficult to distinguish these types clearly it will be found of great convenience to adhere to the classification in the following pages. Winter Cyclones. As stated in preceding paragraphs, the weather of our middle latitudes is characterized by an irregular succession of atmospheric waves passing from west to east; the areas in which the barometer reads high corres- ponding with the crest of the waves, while the areas of low barometric pressure may be compared with the troughs. When these crests and troughs are well developed and sharply defined the latter are known as cyclones, or simply as storms, while the crests are called anti-cyclones; in the winter season when these anti-cyclones develop to unusual inten- sity they constitute our cold waves. When they are well developed and move with average speed across the country these cylonic disturbances usually cover a period of two to three days in passing a given meridian. As they pass over a region they bring to it a fairly regular sequence of weather changes. The character of these successive changes is modified by various conditions. First in importance is the position of the region with reference to the center of the barometric depression, or storm. The path traversed by the center of the storm with reference to Baltimore depends largely upon the place of its origin. MARYLAND WEATHER SERVICE 335 In selecting a series of storms for illustration to show the different varieties of weather experienced in the vicinity of Baltimore during the course of the year, it will be found convenient to classify them, basing the classification upon the place of origin of the depression, or, perhaps better, the position of the center of the depression a day or two before its arrival over the region about Baltimore. Four types will be described in the order of their percentage of frequency across the horizon of Baltimore — the Lake storm, the South- west storm, the Gulf storm, and the Coast storm. These types imper- ceptibly merge into one another at times, but they have sufficient individuality to permit of ready separation into the classes named. All of these classes show their most intense development in the winter season, with perhaps the exception of the Coast storm; the latter is likely to be the northward extension of a West Indian hurricane, and hence shows a maximum frequency in the early autumn, or late summer. THE LAKE STORM. The Storm of December 24-26, 1902. The daily weather map of the United States Weather Bureau for 8 a. m., December 23, 1902, shows a distribution of pressure which caused a fairly normal condition of winter temperatures. A barometric pres- sure above the seasonal average prevailed over the eastern half of the country with a maximum over the Great Lakes, giving rise to northerly winds east of the Mississippi River. A depression, first shown on the map of the 22d over Puget Sound, had made its way eastward to Montana and North Dakota. West of the Mississippi this depression had already shown its influence in a drift of southerly winds towards the center of depression, but the isotherms had not as yet been greatly bent from their normal trend. Twenty-four hours later, at 8 a. m. of the 24th, the center of the depression had moved eastward a distance of about 600 miles, its center being over Lake Superior. The effect of 24 hours of southerly winds in advance of the center of the storm, coupled with the southward flow of winds from the colder northwest quadrant behind the storm center, changed the isotherms from their normal east-west 336 THE CLIMATE OF BALTIMORE Fig. 91.— The Lake Storm of December 24, 1902. 29 e Fig 92.— The Lake Storm of December 25, 1902. MARYLAND WEATHER SERVICE 337 trend to north-south lines. Temperatures in ach ance of the center were raised 15° to 20° in the southeast quadrant, while there was a fall of equal amount in the southwest quadrant. On its way eastward the area of precipitation grew in extent. The temperatures on all sides of the storm center being below freezing point the precij)itation was practically all in the form of snow. As is most frequently the case, the storm area was oval in shape, with its long axis extending north and south ; Fig. 93.— The Lake Storm of December 26, 1902. as a result the winds about the center blew from the southeast in advance of the long axis, and from the northwest in the rear of the advancing central line, on both sides l)lowing towards the trough of tlie lowest pres- sure. By Christmas morning the storm center had moved eastward to the Lower Lake region, a distance of about 500 miles, and before the close of the day the trough of lowest pressure had crossed the meridian of Baltimore. The eastern edge of the area of snowfall had reached the 338 THE CLIMATE OF BALTIMORE Atlantic coast by 8 a. m., from Massachusetts to Xorth Carolina. Dur- ing the preceding 12 hours snow had fallen in varying amounts over the entire area from Chicago eastward to the Atlantic coast, and from the Lakes southward to North Carolina. The area in advance of the storm over which the temperatures rose 15° to 20° now extended to the coast, while a cold wave (an area of high pressure) followed close behind the storm center, attended by northwesterly winds and clear skies. By 8 a. m. of the 26th of December, the center of the storm had reached the New England coast in its due eastward progress, covering another 600 miles in the preceding 24 hours. Here it remained nearly stationary for 24 hours before continuing its eastward course over the North Atlantic Ocean, By this time the area of high barometric pres- sure following the storm had spread over the entire region from the Eocky Mountains eastward to the Atlantic coast and from the Lakes to the Gulf coast, carrying freezing temperatures southward into Middle Florida. The weather map of December 27th shows a condition which frequently occurs in the winter months — a striking inversion of temperatures between Florida and Nova Scotia. Jacksonville, Fla., had a tempera- ture of 24° at 8 a. m., while Sydney, N. S., reported a temperature of 40° at the same hour. The reason for this apparent anomaly is readily found on examining the weather maps of the preceding days. The cold northwest winds flowing out of the area of high pressure in the rear of the advancing storm had reached the Gulf states while the warm southerly winds were still blowing over Nova Scotia in the southeast quadrant of the storm area. The path of this storm of December 24-26, 1902, from the Lake region eastward is the approximate path of nearly three-fourths of the baro- metric depressions which exert a direct influence upon the weather con- ditions in the vicinity of Baltimore. The path of the center lies well to the north of Baltimore. The successive changes in the elements of the weather experienced during the passage of this type of storm across the meridian of Baltimore are graphically illustrated in the accompanying diagram, a brief description of which will suffice to call attention to the most important factors. (See Figs. 91-94.) MARYLAND "WEATHER SERVICE 339 > )- \ \ 1 z s h Z UJ Ct 1 Q \ I / in \ 1 D I \ UJ > I Q UJ (r a / I "- \ Z •' 7 / .^- J \ / \ K " UJ f \ CO Q a + r~- < \ i ^ ^ /> UJ J u \ A " \ / 00 \ / I / IN 1 / U) y •/ CD ^ coy 0) u c\j w ^ /^ \ 1 / - i f 1 -« lo / f3 V \ / i — v — / -« tv / r. ^ \ / ' / rl J \ V / u \ \ 1 ^ V > - / \. / en 'I- o \ r~ / ' / - ► f- ► r^ * 2 C\J b U UJ Q < UJ J u 340 THE CLIMATE OF BALTIMORE Within the horizon of Baltimore the ajjproach of the storm from the west was announced by a steady fall in the height of the mercury in the barometer after 10 o'clock in the morning of the 24th of December. The day began with clear skies; soon after sunrise the clouds began to form, increasing in amount until the sky became overcast by 10 a. m. The winds blew from the north in the early morning. About 10 a. m. the direction changed to northeast, continuing to veer to east and then to soutlieast and south with the continued fall of the barometer. Co- incident with the changes in the wind from north to east and southeast the temperature rose steadily until nearly midnight; the usual diurnal fall in temperature after 3 p. m. being eliminated by the cyclonic rise. The barometer continued to fall until 8 a. m. of Christmas day, the time at which the trough of lowest pressure of the storm area crossed the meridian of Baltimore. At about the same hour the wind veered from southeast to southwest. The temperature continued to rise until 10 a. m. and then fell steadily with the persistent blowing of west to noi-thwest, winds. The humidity increased from 40 per cent at 10 a. m. of the 24th, when the winds changed to an easterly direction, to a maximum of 90 per cent at 8 a. m. of the 28th. With the change of wind from southerly to westerly the humidity fell rapidly to 45 per cent within a period of about two hours. A light dry snow began to fall betw^een 10 and 11 p. m. of the 24th, with a falling barometer and a southerly wind. The snow continued until 2 a. m. of the 25th, the total amount lieing about three inches. The sky remained overcast until 10 a. m. of the 25th, when the clouds began to break away soon after the shift of the wind from southeast to southwest. By 8 p. m. the sky was clear. The usual sharp rise in pres- sure after the storm was not experienced in the passage of this depres- sion, the barometer remaining comparatively low through the 25th and 26th. As a result there were no high winds in Baltimore during the progress of the storm : there was a slight increase in velocity, however, following the turn in the barometer and the change in direction of the wind from west to northwest. The series of maps showdng the progress of this storm across the United States well illustrates the normal winter conditions of a succes- MARYLAND WEATHER SERVICE 341 sion of areas of high and low pressure moving from west to east across the continent. On the map of December 2-ith we see an area of high pressure over Xew England, a depression over the Upper Lake region, another high area over Montana and the Canadian Xorthwest, and another depression appearing on the Pacific coast over Oregon and Washington. These systems move eastward at an average rate of about GOO miles per day with their centers mostly between the 40th and 50th parallels of latitude, bringing to localities over which they pass their characteristic changes in temperature, in wind direction and force, in clouds and sunshine, and in rainfall or snowfall. The Storm of January 7-S, 1903. The daily charts of January 6, 7, and 8, 1903, issued by the United States Weather Bureau, show the progress of another storm of the Lake region type. On the 5th a depression appeared upon the field of the map in the extreme northwest of the Canadian Provinces. By 8 a. m. of the 6th the depression had crossed the boundary line into Xorth Dakota as a well developed storm, its influence being felt over most of the area between the Eocky Mountains and the Mississippi Valley. In the succeeding 24 hours the center of the storm had traversed a distance of nearly a thousand miles, from Quaj^elle, Manitoba, to Central Michigan. The depression developed no precipitation area until the night of the 6th. By 8 a. m. of the ith snow had fallen over a sym- metrical oval area about the center, extending about a thousand miles from east to west and about six hundred miles from north to south. Passing eastward witli the same rapid rate of progress the center moved over Nova Scotia by 8 a. m. of the following day. Wliilc the area of snowfall attending this storm reached as far soutli as Tennessee and North Carolina, the precipitation was extremely light in Maryland and Virginia, and was of short duration. The rainfall or snowfall in Balti- more and vicinity is generally light and falls in tlic form of brief showers or snow flurries with storms of this type, unless their centers pass the meridian of Baltimore within a hundred miles, or less, to the north, when the precipitation may be heavy and of considerable duration. 343 THE CLIMATE OF BALTIMORE Fig. 95. — The Lake Storm of January 7, 1903. 30.0 Fig. 96.— The Lake Storm of January 8, 1903. MARYLAND WEATHER SERVICE 343 23 344 THE CLIMATE OF BALTIMORE In this storm the normal trend of the isotherms was not disturbed to the same extent as in the case of the storm of December 24-26, 1902, described above. The usual decided fall in temperature (20 degrees or more) followed in the path of the storm, but in this instance the cold wave R-as nearly 24 hours behind the center of the depression and did not reach the Atlantic coast. The local changes during the progress of the storm were not ver}' pronounced but they were representative of the type following a similar path. Early in the morning of the 7th Balti- more came within the area of influence of the Lake storm described above. The barometer began to fall about midnight of the 6th-7th, the wind changed at the same time from northwest to west and soon after to the south^^•est ; by 6 a. m. the wind had backed to the south and this direction prevailed until 4 p. m. Between 4 and 5 p. m. there was an abrupt change of the wind from south to west, accompanied by a rise in the barometer. The pressure rose slowly though steadily throughout January 8th as the center of the depression moved over the Atlantic oif the New England coast. The wind did not materially increase in force until 2 a. m,, about 10 hours after the beginning of the rise in the barometer, attaining a maximum velocity of 28 miles per hour before noon. The passage of a coast storm on the 5th-6th left a raw blustery Mdnd blowing from the west and northwest in the afternoon of the 6th, witfi clearing skies. Cloudiness increased during the morning of the 7th upon the approach of the Lake depression and the sky soon became overcast. There was a brief breaking away of the clouds between three and four p. m. Light snow fell between 8.30 and 10.30 a. m. of the 7th, between 11.50 a. m. and 1.45 p. m., between 4 p. m. and 5.30 p. m., and again between 9.15 and 9.40 p. m. The total fall of snow was not much over half an inch. During the 8th the sky was overcast with only occasional brief intervals of sunshine. There was a slow and steady fall in temperature and a steady rise in the barometer, with brisk westerly winds in the forenoon. Traces of snow fell between 11.20 and 11.51 a. m., 12.05 and 12.25 p. m., and between 7.55 and 8.25 p. m. The rise in temperature in advance of the storm was not well marked. MAKYLAXD WEATHER SERVICE 345 This may be readily accounted for by the brief duration of tlie south- erly winds in advance of the center, covering a period of less than 10 hours. (See Figs. 95-97.) The Storm of February 27 -March 1, 1903. These Lake storms sometimes develop into disturbances of great extent and intensity. The weather chart of 8 a. in., February 27, shows a depres- sion centered over Eastern Nebraska, formed apparently by the union of two distinct depressions; one of these had its origin in the Canadian Northwest Provinces, the other in the extreme southwest, over Arizona. By 8 a. m. of the 27th a very considerable rain area had already developed over the Central and Southern states, aided largely by the presence of a well developed area of high barometric pressure over the Atlantic Ocean oft the Middle Atlantic states. During the succeeding 24 hours the storm area grew to unusual proportions, while it moved eastward across the Lake region at a rate slightly above the normal rate of progress for such storms. By 8 a. m. of the 28th the precipitation area of the pre- ceding 12 hours embraced all of the country east of the Mississippi Elver. To the South and east of the storm center the areas in which southerly winds prevailed, temperatures rose from 15° to 40°, and the precipitation was in the form of rain; M^est of the center of the storm, in the area of northwest winds, there was a fall of 20° to 30° in 24 liours. The rain area was not only of unusual extent, but the eastward movement of the storm was marked by very heavy rains, measuring an inch and a half to two and a half inches in 24 hours at many stations in the South Atlantic and Gulf states. The passage of the trough of low pressure was also the occasion for the production of severe squalls and local storms. By the morning of March 1 the storm center had moved east- ward to the Gulf of St. Lawrence followed closely by a fall of 20° to 30° over a large area, embracing a dozen or more states. The local changes during the passage of this wide-spread storm across the meridian of Baltimore were exceptionally well marked and character- istic of the well developed storm of the type with a path across the Lake region and down the St. Lawrence Valley. (See Figs. 98 101.) THE CLIMATE OF BALTIMOItE Fig. 98.— The Lake Storm of February 27, 1903. Fig. 99.— The Lake Storm of February 28, 1903. MARYLAND WEATHER SERVICE 34.7 On the morning of the 26th an area of high barometric pressure rested over the Atlantic states, with its center over Maryland and Virginia. The winds were light and variahle in direction. The skies were clear, resulting in heavy frosts during the preceding night and in the early morning hours, throughout the state. With clear skies and light winds the temperature rose rapidly during the day, and the air became balmy and spring-like. Tliere were a few Fig. 100.— The Lake Storm of March 1, 1903. cirro-stratus clouds in the early morning, but they soon disappeared. At 10.40 a. m. the local Weather Bureau Office received the following telegram from the Central Olhce in Washington : " Southeast storm warnings ordered hoisted along the Atlantic coast from ]\Iiami, Fla., to Charleston, S. C. Storm over Texas is moving northeastward. Brisk to high easterly winds arc indicated this evening and tonight on the South Atlantic coast.*' 348 THE CLIMATE OF BALTIMORE MARYLAND WEATHER SERVICE 349 Clouds gathered during tJie night, and at dawn of the 37th the sk}^ was entirely overcast with a thin veil of stratus clouds. The atmosphere was humid and the clouds began to thicken. A solar corona was observed in the forenoon. The barometer fell steadily and rapidly throughout the day, the wind changed to southeast and east, while the temperature rose rapidly from a minimum of 33° at & a. m. to 52° at 4 p. m. A light rain fell from 2 p. m. to 2.10 p. m., began again at 3.35 p. m., continuing through the night. The amount of rainfall at midnight was 0.68 inch. At 11.45 a. m. southeast storm warnings were ordered up along the Atlantic coast as far north as Ft. Monroe, and later, southwest storm warnings were ordered from Baltimore to New York. The following day, February 28, was cloudy and warmer. The atmosphere was humid and oppressive in the forenoon, but became more pleasant in the afternoon. Light fog formed during the night; at dawn it was dense, but soon became lighter, disappearing by 11 a. m. About 1 p. m. there was a temporary break in the clouds, but in about an hour a heavy stratus mass arose and rapidly covered the sky. At 10.30 a. m. storm warnings were ordered changed to northwest from South Carolina to Virginia. The rain which began on the preceding day continued to 7 a. m., began again about 8.30 a. m. ; it was heavy for a few minutes after 10 a. m. and continued wdth brief interruptions until 3.15 p. m. The total fall from midnight was 0.46 inch. The winds were fresh to brisk between 10 and 11 a. m., increasing in the afternoon and evening to high ; the maximum velocity was 38 miles from the west at 2.45 p. m. The barometer continued to fall rapidly to a minimum of 29.27 inches at 2 p. m., while the temperature rose steadily from 52° at midnight to 71° at 2 p. m. At this hour the wind veered from south to southwest and then to the west, accompanied by a rapid rise in the barometer and a sharp fall in the temperature. The atmosphere became crisp and invigorating throughout the balance of the day, and the day following (March 1) the barometer rose and the temperature fell rapidly and steadily, while the wind continued from the west and northwest. During the passage of this storm the temperature rose 38° in advance of the center, from a minimum of 33° at 6 a. m. of February 27 to a 350 THE CLIMATE OF BALTIMORE maximum of 71° at 2 p. m. of the 28tli. Tlie barometer fell an inch during the same period. After the passing of the center of the storm across the meridian of Baltimore the temperature fell 38° in 30 hours, while the barometer rose over an inch during the same period. THE SOUTHWEST STORM. A much frequented path for storms has its origin in the Southwest and trends northeastward across the Lake region and down the St. Lawrence Valley, or across the New England states. Storms of this type may have their origin in the extreme northwest, or they may enter the United States from the Pacific Ocean off the coast of California, but they dip far to the south, their centers passing over Oklahoma or Texas, b.efore proceeding on their way eastward by way of the Lake region. In their journey southeastward these storms gather energy and moisture with increase in temperature. They are characterized by a sharp rise in temperature in advance of the center of the depression, as the warm moisture laden southerly winds from the Gulf and South Atlantic are drawn into the circulation for a relatively long period. As they move northward the temperature is not only lowered by rising currents in advance of the storm, but also by reason of their entrance into cooler latitudes. As a result of the lowering of temperature and their prox- imity to the main sources of water supply- — the Gulf and flie Atlantic Ocean — clouds and rain form rapidly over a very large area about their centers. While the paths of such storms may not pass in closer proximity to Baltimore than do the Northwest Lake storms, their rain areas extend farther southward and eastward from their centers and hence bring to Baltimore a longer period of unsettled weather and a heavier rainfall. The Storm of Fchniary 3-5, 1903. This storm entered the United States from the Pacific Ocean on February 1. At 8 a. m. of the 2d its center was over Arizona, and on the 3d over Texas. During the succeeding 24 hours the storm turned sharply to the northeast, increasing in energy and area, reaching Lake Michigan by 8 a. m. of the -ith. By this time the rain area had already MARYLAND WEATHER SERVICE 351 reached the Atlantic coast from Florida to Maine, while it extended westward to ^Tebraska. In its southeast quadrant the temperature rose 20° or more in 24 hours, while a marked cold wave closely followed the center of the depres- sion to the southwest, with a fall of 20° to 40° in 24 hours. In advance of the storm the precipitation occurred as rain, excepting in the north- east where the temperature fell below 32°. Here, and to the west of the Fig. 102.— The Southwest Storm of February 3, 1903. storm center, snow fell over a large area, in many places to a great depth. The barometric gi'adients in this storm were very steep, the difference between tlic pressure at the center and the outer edge of the storm being an inch or more. (See Figs. 102-10.").) Tiie following description of the conditions at Baltimore during the passage of this storm is taken from the records of the local ullice of tlie United States Weather lUireau : 352 THE CLIMATE OF BALTIMORE Fig. 103.— The Southwest Storm of February 4, 190:; Fig. 104.— The Southwest Storm of February 5, 1903. MARYLAND WEATHER SERVICE 353 February 3, 1903. A warm cloudy day. Cirrus clouds formed rapidly after 7 a. m., the sky becoming overcast by 10 a. m. The clouds increased in density. Light rain began at 10.55 p. m. and continued into the night. At midnight the amount of precipitation was 0.05 inch. The atmosphere was balmy and springlike. Fig. 105.— The Southwest Storm of February 3-6, 1903. February //, 1903. The day continued cloudy and warm with a sultry atmosphere in the morning. The temperature rose to 66° at 4 p. m., then fell sharply 7° just before 5 p. m., followed by a steady fall. The sky was over- cast until 1.50 p. ra. The strato-cumulus clouds changed to cumulus by 2.30 p. m.; these in turn disappearing by 4 p. m. Dense fog prevailed during the 354 THE CLIMATE OF BALTIMORE preceding night, became light in the early morning, and disappeared by 10.30 a. m. The light rain of the night before continued into the morning, be- coming heavy about 6.30 a. m., 0.20 inch falling in 5 minutes. This brief downpour was preceded by a single flash of lightning. The rain continued at intervals until 8.30 a. m. The total fall from midnight was 0.86 inch. A slight peal of thunder was heard at 10.21 a. m. At 4.50 p. m. there appeared an inky-black, closely compacted mass of strato-cumulus clouds, driven from the northwest, though the cloud mass showed a distinct northeastward move- ment. By 5 p. m. the entire mass had risen above the western horizon, covering about six-tenths of the sky. On the northern edge of the cloud mass several cumulo-nimbus of the " anvil " variety were seen. Rising above the western horizon were cumuli, small in size, and extending north and south for about 25°, with an overlying cirro-stratus layer. There were three air currents: The upper current was moving from the west; the middle current from the southwest; the surface wind was from the northwest. Though the cloud mass moved eastward in a body, the northeast end seemed fixed, and a general commotion was noticed in the base of the cloud strata in this portion; mammo-cumulus clouds appeared and disappeared for about ten minutes. At 5.15 p. m. breaks occurred in the mass, exposing snow-white cumulus peaks with the crowns growing in size, indicating ascending air currents. At 5.30 p. m. the mass was steadily being pushed southeastward and an alto-stratus layer set in from the northwest. The western edge of the cloud mass passed over the station at 5.40 p. m. From this time until 6.50 p. m. the mass was very dark in color, except on the extreme northeast edge, where several snow-white mountainous cumulo-nimbus prevailed. From and among these cumulo-nimbus broad flashes of lightning were seen from 5.50 p. m. until 6.50 p. m. Two successive peals of thunder were faintly heard in the eastern suburbs of the city at 5.30 p. m. Southwest storm warnings were ordered up along the coast from "Wilmington, N. C, to New York. The winds became brisk to high after 7 p. m., with a maximum velocity of 38 miles from the west at 7.15 p. m. Fehruary 5, J903. The day was partly cloudy and colder. The sky was overcast during the forenoon; clouds began to break away about noon, and by 3.30 p. m. they had disappeared. Brisk to high westerly winds con- tinued throughout the night and during the day, decreasing to fresh in the evening; the maximum velocities exceeded 40 miles an hour. Considerable damage was done by the wind to signs and awnings, and a few houses were partially unroofed. Snow flurries occurred between 8 a. m. and 11 a. m., but no snow remained on the ground; the total fall was less than a tenth of an inch. The Storm of December 26-28, 190Jf. This disturbance, like the storm described in the preceding paragraphs, had its origin over the Pacific Ocean. Its center appeared off the coast of Oregon on the 24th inst. Moving rapidly southeastward across the MARYLAND WEATHER ■ SERVICE 355 Eocky Mountains the storm reached Texas on the morning of the 26th; recurving sharply northeastward the center was over Central Illinois 24 hours later and over Toronto on the morning of the 28th. Following the usual course down the St. Lawrence Valley the storm passed eastward over Labrador to the Atlantic, crossing the continent from ocean to ocean in just five days along a path about -iOOO miles in length. Assuming a uniform rate of speed the average daily movement was 800 miles. The rate varied from 1000 miles in 24 hours from the Pacific Ocean across the Eocky Mountains, to 500 miles in crossing the Lake region. The storm was characterized by a precipitation area of unusual extent, and by heavy local rains and snows. The isotherms were bent from their normal east-west direction to a north-south trend near the center by the warm southerly winds in advance of, and the cold northwest winds in the rear of, the center of the storm. The rise in temperature in the southeast quadrant was 20° to 30°, while the subsequent fall in the southwest quadrant varied from 20° to 50° in 24 hours. The local changes in Baltimore during the progress of this storm were well marked and characteristic. While the center of the storm was over Texas, on the morning of the 26th, an area of high barometric pressure rested over the Xew England states. Tliis distribution of pressure caused north to northeast winds in the Middle Atlantic states. As the storm moved eastward and northward toward the Lake region the wind at Baltimore veered to east and southeast, and by noon of the 27th it had become south. During the night of the 27th-28th, while the center of the storm was over the Lake region, a secondary depression developed in the southeast quadrant of the main storm, over eastern Pennsylvania and Xew York, causing a sudden change of Avind to north at Baltimore; as the storm center moved eastward the winds settled to northwest witli rapidly increasing velocity. The barometer fell from 30.24 inches at 8 a. m. of the 26th to 29.10 inches at 4 a. m. of the 28th, and rose again to 30.00 inches by 10 a. m. of the 29tii. Co-incident witli tlie fall in pressure and the changes in the direction of the wind to the south, noted above, the temperature rose from 26° at 6 a. m. of the 26th to 55° at 4 a. m. of the 28th, then 356 THE CLIMATE OF BALTIMORE Fig. 106.— The Southwest Storm of December 26, 1904. Fig. 107.— The Southwest Storm of December 27, 1904. MARYLAND WEATHER SERVICE 357 fell with change of wind to the north and northwest, to 20° at 6 a. m. of the 29th. This cyclonic rise and fall in temperature totally obliterated the diurnal fluctuation usually noted in the daily temperature curve. The maximum temperature occurred at 4 a. m. of the 28th. There was a steady rise during the 27th from midnight to midnight, and a steady and regular fall throughout the following day. The rain was continuous but light. Beginning at 9 a. m. of the 26th as a light misting rain it continued as such without interruption until Fig. 108.— The Southwest Storm of December 28, 1904. 10.25 p. m., when it became heavier. About 6 a. m. of the 27th it again changed to a light mist which continued to the end of the precipitation period between 4 and 5 p. m. The total amount of rainfall (including some sleet) fdr the 32 hours was only 0.44 inch. (See Figs. 106-109.) The daily journal of the local Weather Bureau Office contains the following remarks concerning conditions on the 27th and 28th: December 27, 190 Jf. A cloudy day. Continuous fog. Light rain, continu- ing from midnight yesterday, turned to misting rain at 6.05 a. m.. and ended 358 THE CLIMATE OF BALTIMORE MARYLAND WEATHER SERVICE 359 at 4.30 p. m. Southwest storm warnings were ordered up by the Chief of Bureau at 9.45 a. m. from Jacksonville, Fla., to Fort Monroe, and southeast warnings at 11.15 a. m. from Baltimore to New York. A cold wave warning was received at 9.55 p. m. Dece7nber 2S, 190.'i. Partly cloudy until 9 a. m., followed by cloudy; clear after 1.30 p. m. A slow steady rise in temperature since Christmas morning culminated in a sharp rise to 55° at 4 a. m. From this hour the temperature fell steadily to 22° at midnight. Light rain fell between 1.15 a. m. and 2.10 a. m., amount 0.01 inch. The wind became brisk at 9.15 a. m. and continued so until nearly midnight. The velocity rose to a maximum of 40 miles per hour from the west at 12.45 p. m. The Storm of December 12-13, 1903. This storm first appeared within the field of view ou the lUth of December in the extreme Nortliwest. It crossed the Eoclcy Mountain range in Montana in a southeast course during the night of the lOth-llth, and its center was over Missouri and Arkansas at 8 a. m. of the 12th. Here it recurved to the uortlieast, taking the usual course across the Lake region, down the St. Lawrence Vallev and over Labrador, where it disappeared beyond the field of the weather map on the 14th inst. This storm resembled the southwest storm described above in most respects. There was an important dift'erence, however, in the form of the system of isobars surrounding the center as the storm crossed the meridian of Baltimore on December 13. The (jval shape of isobars, with the long axis extending approximately north and south is characteristic of many of this class of storms. The change from easterly to westerly winds, as the trough of low l)aro meter moves eastward, is very abrupt, and is frequently attended by severe squalls or thunderstorms. The isotherms extend nearly north and south and are close together in the vicinity of the trough of low pressure, or, in other words, the tempera- ture gradient is very steep and contrasts are great. In the case of this particular storm of the 13th there was a difference of 50° at 8 a. m. between Baltimore and Tiidinimpolis. on the same ])arallol of latitude. or a difference of 50° between the southerly winds prevailing in advance of the center and the northwest winds which blew out of the w^ell developed area of high pressure in the rear of the storm. The trough of low pressure passed over Baltimore almost at the exact time of the 8 a. m. observaiions of the Unilcd States Weather 24 360 THE CLIMATE OF BALTIMORE LOW LOW Fig. 110.— The Southwest Storm of December 12, 1903 LOW\ !^o\> Fig. 111.— The Southwest Storm of December 13, 1903. MARYLAND WEATHER SERVICE 361 362 THE CLIMATE OF BALTIMORE Bureau. A detailed presentation of the successive local changes during the 13th, as this storm traversed the horizon of Baltimore will Ije found in the accompanying diagram. It is also possible in this case to show the hourly changes in the relative humidity. At 8 a. m., with change of wind from south to southwest and then to northwest, there was a remarkably rapid change in the humidity, the decrease amounting to about 55 per cent in four hours. (See Figs. 110-112.) Fig. 113. — Paths and Rain Areas of Southwest Storms of January, 1898. In the reports of the local office of the United States Weather Bureau the loth is described as cloudy in the forenoon and clear in the afternoon. The temperature was very high in the morning, with a maximum of 52° about 8.30 o'clock. With a sudden change of wind at this hour from southwest, through the west, to northwest, a rapid fall in temperature took place (10° in the first hour), and it continued to grow colder to a minimum of 30° at midnight. The atmosphere was crisp and invigorat- MARYLAND WEATHER SERVICE 363 ing in the afternoon. A blustering wind prevailed in the forenoon. Light rain began during the night or early morning, and ended at 9.30 a. m. The total amount was 0.30 inch. The wind became brisk shortly after 9 a. m., changing to high westerly winds, and then to northAvest, with a maximum velocity of 41 miles per hour at 9.30 a. m. A cold wave warning was received at noon, announcing a probable change of 20° to 30° before the close of the following day. Southwest storm warnings had been ordered up along the coast from Savannah to New York on the 12th; tlVose were changed to northwest on the morning of the 13th. These southwest storms are usually accompanied by large rain areas and heavy local rains. At times a series of these storms will follow one another in close succession, all taking approximately the same path, from Texas across the Lake region and Xew England, or the St. Lawrence Valley out into the Atlantic. A remarkable series of tliis kind was experienced during the month of January, 1898. The .accompanying chart (Fig. 113) shows the paths of six storms of this type all occurring between the 8th and 26th of January, 1898, together with the total amount and distribution of precipitation recorded along the various paths. The rate of movement of the storms is shown by the circles along the lines illustrating the storm paths, the intervals representing periods of twelve hours. THE GULF STORM. Many of the storms which have their origin in the southwest or over the Pacific, and cross tlie country along the southwest path, continue their southeast course to the Gulf before recurving to the northeast. Some have their origin over the Gulf of Mexico and move northeastward to tlie Gulf of St. Lawrence. The path taken by these storms brings tlicir centers very close to Baltimore. Sometimes the center of the barometric depression passes just to the west of Baltimore, sometimes to the east, and occasionally immediately over the city. They are usually accompanied by heavy precipitation, and by high winds along the coast. Very frequently these storms develop over the Gulf of ]\Iexico while 364 THE CLIMATE OF BALTIMORE an area of high pressure prevails over the New England states. Under the influence of this distribution of pressure, northeast to east winds set in over the Middle Atlantic states and southeast winds over the South Atlantic states. The rain area spreads rapidly northward and eastward under these conditions and reaches Baltimore while the center of the depression is still in the Gulf states. The average winter temperature of Baltimore is close to the freezing point; hence slight changes in temperature will change the form of precipitation from rain to snow or from snow to rain, or to the disagree- able intermediate stage of sleet. As these Gulf storms are nearly always preceded by comparatively high temperatures and followed by tempera- tures below the freezing point, they are apt to cause much personal dis- comfort, with their rain, sleet, and snow, resulting in slushy or icy streets in the cities. Farther north the precipitation is mostly in the form of snow, and a short distance to the south it is all rain. The high winds which frequently accompany this type of storm not only increase the discomfort but add an element of danger. The Storm of February 1-3, 1902. {Center passes west of Baltimore.) The weather map of S a. m., February 1, 1902, shows the prevalence of two well developed areas of high barometric pressure, one in the north- east, with its center over the Gulf of St. Lawrence, the other in the extreme northwest, centered over Idaho and Montana. In the Gulf states and in the Southwest the barometer was low, and unsettled weather prevailed from the Lake region to the Gulf, and from the Atlantic coast westward nearly to the Rocky Mountains. In the Atlantic coast states the barometric depression was already well developed, and rain was fall- ing at 8 a. m. throughout the South Atlantic states, in Virginia, Mary- land, and Pennsylvania, and snow in New York and the New England states. As the storm moved rapidly northeastward it developed in intensity and in definiteness of outline, the rains became heavier and the area of precipitation increased. The high area over the Gulf of St. Lawrence MARYLAND WEATHER SERVICE 365 .30.4 Fio. 114— The Gulf Storm of February 1, 1902. Fig. 115.— The Gulf Storm of February 2, 1902. 366 THE CLIMATE OF BALTIMORE remained stationary while that in tiie extreme Northwest moved rapidly southeastward accompanied by a decided fall in temperature in the southwest quadrant of the storm area. At 8 a. m. of the 2d of February the area of lowest barometer was over Pennsylvania, the center of the storm having passed just to the west of Maryland during the preceding nisrht. The center of the western hi2:h area was over Kansas and Okla- homa. During the preceding 24 hours a fall of 15° to 30^ in tempera- FiG. 116.— The Gulf Storm of February 3, 1902. ture was experienced over a wide area from Iowa and Nebraska southward to the Gulf coast. High easterly winds prevailed during the night and early morning along the coast from the South Atlantic to the New England states. (See Figs. 114-117.) By the morning of the 3d the storm center had moved to tlie New England states, the cold wave had reached the Atlantic coast from Florida to North Carolina, and had overspread most of Virginia, Mary- land, and Pennsylvania. The local changes at Baltimore during the :makylaxd weather service 367 368 THE CLIMATE OF BALTIMORE passage of this storm are indicated in the accompanying diagram, and in the following extracts from the daily journal of the Weather Bureau : February 1, 1902. On February 1, while the center of the storm was over the Gulf States, the day was cloudy, the sky being continuously overcast. At 7.45 a. m. precipitation began in the form of sleet, turning in 10 minutes to a light misting rain. The winds were from northeast to north from noon to midnight, and very light, averaging but 3 to 4 miles per hour. Light rains continued at intervals until 8.40 p. m., the entire amount for the day being but 0.07 inch. The day was disagreeable; the sidewalks were icy. The maximum temperature of the day was 37° at 6 p. m. The barometer fell steadily throughout the day from 30.03 inches at 4 a. m. to 29.77 inches at midnight. The relative humidity was approximately 100 per cent all day. February 2, 1902. The day continued cloudy during the forenoon. The temperature rose slowly to 39° at 2 p. m., while the barometer fell to 29.25 inches. The wind changed from north to east at 8 a. m. and to west at 10 a. m., and continued light in force. Early in the afternoon the wind began to increase in force, reaching a maximum of 33 miles per hour from the west between 4 and 5 p. m., shortly after the barometer began to rise. From 2 p. m. the temperature fell steadily to 15° at midnight of the following day. After an interval of several hours light rain began again between 2 a. m. and 3 a. m. and continued without interruption until about noon, becoming heavy at 'times. From noon to 1.40 p. m. the precipitation was a mixture of rain and snow, the snow melting as it fell. The total fall of rain and snow com- bined was 0.44 inch. The day as a whole was extremely disagreeable. The sidewalks were icy and dangerous to pedestrians. In the forenoon the gutters and streets were filled with slush, which, as night approached, became frozen solid. Some damage was done to awnings, signboards, and chimneys by high winds. The wind, however, cleared the harbor of floating ice. Light fog prevailed during the preceding night and lifted at about 11 a. m. Northwest storm warnings were ordered up at 10 a. m. from Florida to Baltimore. A cold wave warning was received at 2.50 p. m., forecasting a fall to 15°, or below, in the interior of the State, and to 20° along the coast. The clouds disappeared rapidly after 3 p. m. and by 5 p. m. the sky was clear. February 3, 1902. The day was clear and much colder than the 2d. The temperature fell to 15° at 9 a. m. There "were no clouds excepting a few small cumuli in the afternoon. The wind continued brisk during the night, but diminished towards noon. Navigation was free on the western side of the Bay, but along the eastern shore the ice was piled up by the winds. All tributaries of the Chesapeake were frozen solid. The Storm of January 5-1 , 1905. (Center passes over Baltimore.) On the morning of January 5, 1905, a somewhat similar distribution of pressure obtained to tliat of February 1. 1902, described above. An MARYLAND WEATHER SERVICE 369 Fig. 118.— The Gulf Storm of Jauuary 5, 1905. Fio. 119.— The Gulf Storm of .lanuary 6, 1005. 370 THE CLIMATE OF BALTIMORE area of liigh barometer prevailed over the Atlantic coast states, and another over the Rock}- ]\[ountain Plateau. The pressure was low over the Mississippi Valley with a tendency to deepen over the Gulf states. By 8 a. m. of the following day the Atlantic coast area of high pressure had concentrated over the Xew England states, while the Eocky ^lountain high area had changed but little in intensity or outline. The center of the barometric depression had been transferred to Xorthern Floi'ida ^. % -36 Fig. 120.— The Gulf Storm of January 7, 1905. and Southern Georgia. This combination of pressure along the Atlantic coast always gives rise to northeasterly winds with a steady rain or snow. At 8 a. m. of the Gth rain was falling in the South Atlantic states and snow in the Middle Atlantic and Xew England states. The snow area also reached westward to the Ohio Valley and the Lake region in connection with the development of a secondary depression over Lake Michigan. (See Figs. 118-121.) MARYLAND WEATHER SERVICE 371 372 THE CLIMATE OF BALTIMORE After reaching the coast the storm took a sharp turn northward, increasing in intensity as it followed the coast line. The center passed directly over Baltimore at about 4 a. m. of the 7th with an abrupt change in the direction of the wind from south to northwest, and a fall in temperature. The winds were light to fresh during the progress of the storm over Baltimore, only exceeding 20 miles per hour for a short time between 3 p. m. and 4 p. m. By the morning of the 8th the center had passed northward to the Lower St. Lawrence River. The temperature rose rapidly 20° to 40° along the Atlantic coast in advance of the center of the storm, but fell more slowly after the center had passed, a^? the high area of the Eocky Mountain region was advancing but slowly eastward behind the storm. Heavy rains marked the spread of the storm in its eastern half all along the Atlantic coast; rains of one to two inches in 24 hours were reported from many of the Weather Bureau stations. The precipitation at Baltimore amounted to 2.34 inches during the 12 hours from noon to midnight of the 6th. Some details of the local conditions at Baltimore are shown in the following extracts from the daily journal of the local office of the United States Weather Bureau : January 5, 1905. A cold cloudy day. Light snow began at 8.30 a. m. and ended at 10 a. m. The winds were westerly in the forenoon and easterly in the afternoon. January 6, 1905. The day was cloudy and somewhat warmer than yester- day. Light rain began at 8.55 a. m., ended at 9.10 a. m. ; began again at 12.10 p. m. and continued to midnight. The rain was heavy from 7.40 p. m. to 7.49 p. m., 0.22 inch falling within the 9 minutes. The total precipitation for' the day was 2.34 inches. January 7, 1905. A cloudy day until 6.30 p. m.; the clouds broke away soon after and by 8 p. m. the sky was clear and remained so until midnight. The rain of the preceding night continued until 12.10 a. m. Rain began again at 6.40 a. m. and ended at 7.35 a. m.; began again at 8.50 a. m. and ended at 8.55 a. m. From 12.30 p. m. to 12.50 p. m. snow was mixed with rain. The total precipitation for the day was 0.02 inch. A continuous record of changes in the meteorological elements at Balti- more is shown in the accompanying diagram. (See Fig. 121.) MARYLAND WEATHER SERVICE 373 The storm of February 20-22, 1902. {Center passes east of Baltimore.) February, 1902, was remarkable for the number of Gulf storms experienced. In fact, these storms were a conspicuous feature of the entire winter of 1901-02. While the great majorit}' of our storms follow the northern route across the Lake region in a normal winter, Fig. 122.— The Gulf Storm of February 20, 1902. the storms of February, 1902, without exception, followed the southern path and crossed the horizon of Baltimore with remarkable regularity by way of the Gulf of Mexico. Occasionally there will occur a series of three or four storms in regular succession following this track. The area of cloudiness and rain accompanying a Gulf storm passes over a given locality in about two or three days; this is followed by four or live days of fair weather before the approach of another storm. During the winter of 1901-02 there was a remarkablv regular succession of these 374 THE CLIMATE OF BALTIMORE Fig. 123.— The Gulf Storm of February 21, 1902. Fig. 124.— The Gulf Storm of February 22, 1902. 25 376 THE CLIMATE OF BALTIMORE storms, the period of rain and succeeding fair weather covering seven days and causing the unusually long continued series of rainy Sundays so generally commented upon at the time. This is not an uncommon occurrence but the regularity of the succession was unusually well marked. (See Figs. 126 and 127; also Fig. 113.) Why storms take this southern course with such unusual frequency at times it is difficult to say. Perhaps all that can be said in explana- FiG. 126. — Normal Paths of Storms for February in Black. Average Path of Storms for February, 1902, in Red. tion is that it is due to a departure from the normal conditions in the general circulation of the atmosphere — some unusual movement of the large persistent areas of high and low pressure referred to in an earlier paragraph. One of the most notable of the series of Gulf storms referred to above passed over Baltimore on February 21 and 22, 1902 — a storm which will long be remembered by Baltimoreans on account of the intensely disagree- MARYLAND WEATHER SERVICE 377 able combination of rain, sleet, snow, and high winds experienced. The storm originated off the North Pacific coast on the 17th, moved rapidly southeastward, reaching the Western Gulf coast on the morning of the 19th. Here it lingered for a day, increasing in intensity and enlarging its rain area. Moving eastward along the Gulf coast to the Atlantic, the center followed the coast northward, passing to the east of Baltimore during the day of the 22d, then out to sea. The presence of an area of high pressure to the northeast of the storm assisted in producing a steady north to northeast wind during the 21st. The official records of the Weather Bureau describe the local conditions as follows: February 20, 1902. The day dawned clear and cold. A thin veil of cloud soon appeared, however, increasing in thickness as the day advanced, at SEPT. OCT NOV DEC. JAN. FEB. MCH 1 7 14 21 28 5 12 19 26 2 9 16 23 30 7 14 21 28 4 H 18 25 8 15 22 8 15 22 1 MOT. 6 P 1 1 1 1 1 1 6 A. 1 I 1 1 j i 1 ' ■ ..,' . . . 1 Fig. 127.— Diagram of Rainy Sundays of the Winter of 1901-2. times becoming dark and threatening. The winds throughout the day were light and varying in direction between west and north, and changing to south in the evening. The temperature rose with the advance of the day, but barely reached the melting point of ice even at mid-day. Sleet began to fall at 8 p. m., turning to rain during the night. The barometer fell slowly but steadily throughout the day and night. February .11, l'.i0.i. The rain of the preceding night froze as it fell, cover- ing everything with a thick coating of ice. On trees, telegraph wires, and all exposed objects, the ice collected to a thickness of an inch or more, the heavy weight causing considerable damage. The rain continued with scarcely any interruption until 7.15 p. m.; at times it fell in torrents. Travel upon the streets became difficult and dangerous. The heavy rains of the afternoon converted the ice upon the streets into a heavy slush. The temperature re- mained nearly constant, varying but little from the freezing point of water. The winds were northeast to north all day, and increasing in force, not attaining a storm velocity, however. The barometer continued to fall 378 THE CLIMATE OF BALTIMORE steadily and more rapidly toward night. The total precipitation for the day was 2.13 inches. Thq center of the depression was off the coast of North Carolina, the storm moving east of north and increasing in intensity. February 22, 1902. A rainy day, with very little range in temperature, the maximum being 36° and the minimum 34°. The barometer reached its lowest reading at 6 a. m., rising slowly but steadily from this hour. Fresh northerly winds prevailed. The heavy rain of the preceding day turned to a mist at 11 a. m. and was accompanied by light flurries of snow between 3 p. m. and 6 p. m. At 7.20 p. m. a light moist snow began to fall, continuing at 8 p. m. The total precipitation of the day was 0.40 inch. The ice remained upon the streets most of the day in spite of the heavy rains and was a source of great discomfort. The heavy accumulation of ice caused much damage to trees and to telegraph and electric wires. February 23, 1002. The day was clear and somewhat warmer than yester- day. The snow of the preceding night ended about midnight. The ice on the streets rapidly disappeared with the increased warmth. The day was pleasant and the atmosphere balmy. Altogether this storm was one of the most disagreeable experienced in Baltimore. (See Figs. 122-125.) THE BLIZZARD. When storms such as have been described in preceding paragraphs are accompanied by heavy snow, high winds, and a temperature well below the freezing point, they are popularly known as blizzards. This type of storm is fortunately of infrequent occurrence in the Middle Atlantic states. When they have occurred it has been in connection with a Gulf or Southwest storm. An invariable accompaniment of the blizzard is the presence of an excessively developed area of high baro- metric pressure following in the wake of the depression, causing a steep barometric gradient and feeding into the storm center with great energy the cold westerly winds of the anti-cyclone. Two storms of this type are especially worthy of consideration at some length owing to their exceptional severity all along the Atlantic coast — one is known as the blizzard of March, 1888, the other as the blizzard of February, 1899. The former, while occurring in March, is a marked instance of " winter lingering in the lap of spring." The Blizzard of March 11-13, 1888. The daily weather charts of the Weather Bureau for March 11, 1888, show the existence, in the morning, of an area of high pressure (anti- MARYLAND WEATHER SERVICE 379 Fig. 128.— The Blizzard of March 11, 1888. LOW ^. The day was rloar and much colder than that of the 8th. The maximum temperature of the day was the lowest maximum re- 388 THE CLIMATE OF BALTIMORE corded in Baltimore, namely 7°. There was a light fog in the morning. The winds were brisk to high, reaching a maximum velocity of 25 miles per hour. At 8 p. m. snow covered the ground to a depth of 11.3 inches. The ice in the harbor has increased in thickness to two inches. February 10, 1899. A clear day. Severe, cold weather. The maximum temperature was 3°, the minimum 7° below zero, the mean 2° below zero, the lowest in the official records for the maximum, minimum, and mean for a day. Much suffering resulted from the intense cold. Several mortormen were overcome, and were revived with difficulty. A number of persons were picked up out of the snow drifts benumbed and unconscious. The suffering among the poor was very great. A series of accidents followed the sudden thawing of water in the water pipes when fires were started in the morning. Ten inches of snow covered the ground at 8 p. m., and the ice in the harbor increased to six and eight inches in thickness. The two ice boats were busy all day in their attempts to keep the channel clear, but the ice formed almost as fast as it was broken. February 11, 1899. A clear day. The minimum temperature was 6° be- low zero. A light fog prevailed in the morning. Light snow began to fall at 5.35 p. m. and continued into the night. The snow on the ground at 8 p. m. was 9.7 inches in depth. There continues to be much suffering from cold, and one death from exposure is reported. February 12, 1899. A cloudy day, with slowly rising temperature. North- east storm warnings were ordered up in the forenoon. The following tele- gram was received from the Central Office in Washington: "Heavy snow is indicated this afternoon and to-night. Notify railroads and transportation companies." The snow which began yesterday at 5.35 p. m. became heavy at 8.15 p. m., then changed again to light snow during the night. It continued throughout the day. About five inches of snow fell during the day; the depth of snow on the ground at 8 p. m. was 14.5 inches. The weight of the snow had crushed a number of small sheds and a few wooden structures. To-day the President Street freight sheds gave way, owing to the accumula- tion of snow on the roof, and about 300 feet of the building fell; the damage amounted to about $20,000. The ice in the harbor is 6 to 8 inches thick. Navigation is practically suspended. Only heavy steel steamships are able to move. Trains are late and irregular. Much suffering continues among the poor. February 13, 1899. A cloudy day. Heavy snow fell all day; the 24-hour fall was 15.5 inches. The depth of snow on the ground at 8 p. m. was 30 inches, the greatest recorded in the official records. Brisk to high north to northwest winds attained a maximum velocity of 28 miles per hour. The continued high blustery winds and the increasing snowfall combined to pro- duce a typical blizzard. Railroad traffic was interrupted at an early hour. The street railways struggled to continue service, but the lines closed one by one, and none were in operation by nightfall. Much suffering continues. At least a score of people were overcome by the cold during the day. Birds are reported perishing in large numbers from cold and lack of food. VOLUME 2, PLATE XX. NOON MD" NOON MDT. MARVLANO WEATHER SERVICE. VOLUME 2, PLATE ) /-^\-^ ./- I l\\\l\\\l-\ ^W " ■* « 5 « " " '2 1° 5 7 6 6 5 9 s 10 10 s 10 e 1 3 5 3 4 t ■• 4 2 5 5 3 6 6 5 7 7 II 10 II (I l( II 2 10 12 1J 13 15 17 21 20 /a re 1( (( (0 II '6 ,8 13 979*«86222 lloimLY Observations at BAr/rrMORE Purinq the Buzzard and Cold Wave op Febboary 9-14, 1899. MARYLAND WEATHER SERVICE 389 February 14, 1899. A cold day with bright sunshine. Snow ended at 11.10 p. m. yesterday; one inch of snow was recorded this morning. Total snow depth at 8 p. m., 28 inches. The ice in the harbor is 10 inches thick. The city is practically snowbound. There was no mail delivery, no railroad move- ment, no street car service. Some vessels forced their way out of the harbor, but they were few in number. Much work is being done by the city and railroad authorities on the streets and lines of travel, but traflac was only partially restored. Much suffering continues. One man was found frozen to death this morning within six doors of his home. A milk famine is threatened. There has been a genei'al rise in the price of commodities, especially of country produce. The Merchants and Miners Transportation Company lost the steamship Texas this morning. The vessel was run ashore in an ineffectual attempt to force a way through the ice and sank. February 15, 1899. A clear day. Light fog in the morning. Street car service was resumed to-day in part. Trains are beginning to run on time. The ice in the harbor is one foot thick. The ice boats have succeeded in keeping a clear channel of 50 feet width. Four arrivals of vessels and one departure are reported for to-day. Snow on ground at 8 p. m., 26 inches. Areas of Fair Weather (Axti-cycloxes), In the preceding pages the cyclonic type — or unsettled weather — has been described in considerable detail. The characteristic conditions of this type are cloudiness, rainfall or snowfall, brisk to high winds, a relatively high temperature, and a low barometric pressure, with winds converging toward the central area of low pressure. In the Middle Atlantic states the cyclonic type dominates the weather conditions somewhat less than half the time, basing the calculation upon the number of days in the year during which some rain or snow falls. The annual number of days with precipitation at Baltimore has varied from 114 to 224 with a mean of 170. This implies that during some- what over half the year the anti-cyclonic, or fair weather type, prevails. The chief characteristics of anti-cyclonic areas are : Barometric pres- sure higher than that over surrounding areas; a system of comparatively light winds, diverging from the central portion of the area : comparatively clear skies; and relatively low temperature. (See Figs. 85 to 89.) High areas, or anti-cyclones, have already been described incidentally in the preceding discussion of storms. They are most numerous and more intensely developed during the winter season, when they move in rapid succession from the central continental areas, in the extreme Northwest, 390 THE CLIMATE OF BALTIMORE along the eastern slope of the Kocky Mountains, eastward or southeast- ward across the United States. When these areas grow to unusual proportions and develop a baro- metric pressure greatly in excess of the pressure in areas along their line of eastward progress, they constitute our " cold waves." There is no sharp line of separation, however, between the cold w-ave and the winter anti-cyclone — no more than there is between the storm and the barometric depression technically known as the cyclone. The anti-cyclone attains its greatest severity when a barometric depression develops in advance of it, causing an energetic inflow of cold northwest winds into the western portion of the depression. In area and in rate of movement the cyclone and anti-C3'clone resemble one another ; in the character of attendant weather conditions they are in most respects the exact opposite. The difference in the character of the weather prevailing over cyclonic and anti-cyclonic areas is strikingly exhibited in the weather chart for the 2d of February, 1902, reproduced in Fig. 115, showing the actual condition of the weather at 8 a. m. as reported by the observers of the United States Weather Bureau in all parts of the United States and Southern Canada. (See also Fig. 89, page 325.) A storm, or cyclone, of great extent and energy prevailed over the eastern portion of the country, with its central area of low barometric pressure over Pennsylvania and Maryland. The area of clouds extended from the Atlantic Ocean westward to the Mississippi Valley, and from the Great Lakes southward to the Gulf coast. The region over which rain or snow was falling at the time of observation, 8 a. m., w^hile more limited in extent still covered a considerable area, comprising practically all of the Kew England and Middle Atlantic states, Ohio, Kentucky, and the eastern half of the Lake region. In the Eastern Gulf states and the Atlantic coast states as far north as Virginia the rains of the preceding 24 hours were very heavy, in some localities exceeding an inch and a half. It will be observed also that the winds within the area just outlined blew in the main toward the central area of low pressure, and that the temperatures were markedly higher within the cyclonic area than in the anti-cyclonic area immediately to westward of the storm MARYLAND WEATHER SERVICE 391 area. The isotherms, or lines of equal temperature, bent far northward in advance of the storm center, where easterly to southerly winds pre- vailed ; to the west of the center the cold northwest winds reached far to the south. High atmospheric pressure prevailed over the area west of the Missis- sippi River, the highest barometer being over Kansas and Oklahoma. The skies were mostly free from clouds, the winds blew, in general, away from the central region of high pressure, the temperatures were compara- tively low, while the isotherms were bent southward, with a maximum dip near the center of the area. The intimate relationship existing between wind direction and the trend of the isotherms, as explained in preceding pages, is strikingly exhibited in this chart. The difference in temperature between the centers of cj'clone and anti-cyclone along the 40th parallel of latitude at 8 a. m. was fully 50°. The successive changes in weather conditions at Baltimore as these two systems — the cyclone and anti-cyclone — moved eastward are shown in the accompanying diagram. (See Fig. 117.) The weather conditions over the United States do not always exhibit these cyclonic and anti-cyclonic systems so well defined, but during the winter season their outlines are nearly always easily recognizable. In place of a definite succession of " highs " and " lows " such as are shown by this chart of February 2, 1902, there may be a number of ill-defined and scattered centers of high and low pressure, causing a period of unsettled weather conditions, COLD WAVES. When anti-cyclonic waves are accompanied by a fall of 20° or more (exclusive of the diurnal fluctuation) to a stated minimum, within a period of 24 hours, they are technically known as cold waves. Sudden changes in temperature such as are here described are of comparatively frequent occurrence in the northern tier of states, but do not reach as far south as Baltimore in any great numbers. In their progress eastward and southward these cold waves lose much of the severity shown when 20 393 THE CLIMATE OF BALTIMORE they first enter this country from the Canadian Northwest Territory. By the time they reach the Atlantic coast many of them have lost the dis- tinguishing marks of a genuine cold wave. In the official records of the United States Weather Bureau we find that out of a dozen or more anti- cyclones which enter this country every winter in the extreme Northwest as cold waves, but three, on an average, retain sufficient of their severity to be classed as cold waves as they pass over Baltimore. In some winters Baltimore has been entirely free from them. This was the case in the winters of 1873-4, 1885-6, and 1889-90. Some times as many as six have been experienced in one season, as in the winters of 1881-2, 1884-5, and 1903-4. From 1870 to 1904 the monthly distribu- tion of cold waves * in Baltimore has been as follows : November. December. January. February. March. Total. 8 19 23 25 17 93 The Cold Wave of December 13-15, 1901. The eastward and southward progress of cold waves from the north- west is well exemplified in the weather charts of the United States Weather Bureau for the 13th, 14th, and 15th of December, 1901. At 8 a. m. of the 13th there were two well defined and extensive areas of high pressure, or anti-cyclones, shown upon the chart: One covered the eastern section of the country, comprising all of the Atlantic coast states, with the maximum barometer over Nova Scotia; the other spread over most of the country west of the Mississippi Eiver, with the center to the north of Montana. Between these two vast anti-cyclones, there was a narrow trough of relatively low pressure, extending from the Upper Lake region to the Gulf of Mexico. In advance of this trough of low pressure, or elongated cyclonic depression, the temperatures rose rapidly under the influence of strong southerly winds. To westward of the trough the cold northwest winds blowing out of the well developed anti- cyclone brought freezing weather far down into the southern states. The western anti-cyclone moved southeastward as a great wave of cold air, closely following the cyclonic depression, causing a very steep gradient ^ See page 128 for details of cold waves. MARYLAND WEATHER SERVICE 393 Fig. 139.— Cold "Wave of December 13, 1901. / HIGH Fig. 140.— Cold Wave of December 14, 1901. 394 THE CLIMATE OF BALTIMORE in temperature from west to east along the wave front. In a straight line from Central Alabama northwestward to the Dakotas, from the center of the cyclone to the center of the anti-cyclone, there was a fall of 100° at 8 a. m. of the 14th; Montgomery, Ala., reported a temperature of 70° above zero, and Bismarck, N. Dak., a temperature, at the same hour, of 30° below zero. By 8 a. m. of the following day, the 15th, the isotherm of 20° extended along the West Gulf coast. The cold wave did Fig. 141.— Cold Wave of December 15, 1901. not reach Baltimore until the forenoon of the 15th, and the Atlantic coast on the morning of the 16th. When the cold wave takes a southeastward path, as it did in this in- stance, the freezing temperatures very frequently reach the Gulf coast states from 12 to 24 hours in advance of their occurrence in the Middle Atlantic and ISTew England states. (See Figs. 139-141.) MARYLAND WEATHER SERVICE 395 The Cold Wave of Fehruanj 10-13, 1S99. This, the most intense and wide-spread anti-c^'clone experienced in many years in this country, has already been referred to in preceding pages in the discussion of the '' Blizzard of Februar}^, 1899 " with which it was associated. (See Figs. 132 to 138, and PL XX.) Eecords of minimum temperatures long undisturbed were lowered in many states east of the Eocky Mountains during the eastward and south- ward progress of this cold wave. It extended even to the West India Islands. At Havana, Cuba, the temperature fell to 54° on the 13th. At Washington, D. C, a minimum of 15° below zero was reported on the 11th. The lowest temperature in the official record at Baltimore was 7° below zero, but temperatures considerably lower were reported from the suburban districts. In passing across the Gulf states zero temperatures were experienced on the 13th as far south as the coast. Cold waves probably differ from anti-cyclones in general only in the intensity of their development and in the circumstance of being preceded by a cyclonic depression. The conditions most favorable for the develop- ment of anti-cyclones are the clear skies and dry quiet atmosphere of the extreme Xorthwest — conditions which favor rapid terrestial radiation. The general eastward drift of the air in latitudes between 30° and 60° north latitude carries the anti-cyclones, when once formed, along in the general current. With the eastward movement there is a southward ten- dency of these anti-cyclones, due probably to the centrifugal force of the earth's axial revolution. These anti-cyclones, like the cyclones, move across the continent at irregular intervals, occupying four to five days in travelling from the Pacific to tlie Atlantic coast, according to their extent and path. In most cases the center of the anti-cyclone, or high area, passes eastward to the north of Baltimore, but frequently the center passes directly over Maryland, and occasionally to the south. Those passing along the northern route are most likely to bring very low temperatures to our lati- tudes, but this is not always true. In fact, in the case of the cold wave of December 13th to 15th, 1901, described above, the center passed directly over Baltimore. At such times the temperatures duo in the cold 396 THE CLIMATE OF BALTIMORE winds from the north or northwest, which accompany all cold waves, are further lowered by the rapid terrestial radiation which takes place in the calm clear centers of the anti-cycloues, especially during the night hours. The low temperatures associated with cold waves do not as a rule con- tinue more than three or four days. By the fourth or fifth day the normal minimum for the month is again reached. (See pages 117 to 133 for additional details of cold days.) The Origin of Cold Waves. The cold waves of the United States have been explained in various ways by those who have studied their history. One hypothesis accounts for them by attributing them to the upper westerly winds which are drawn across the Eocky Mountain range into the cyclonic depressions to the east of the mountains. The clear dry air which descends along the eastern slope cools by radiation into space during the long winter nights to such an extent as to greatly overbalance the warming effect due to compression and insolation as the air descends from the higher to the lower levels. Another theory attril)utes them to the southward movement of detached masses of the dry cold atmosphere which form rapidly during tJie winter months in the region to the west of Hudson's Bay. A third hypothesis refers them to slowly descending currents of the anti-trades which flow from the equator toward the Arctic region in the higher levels of the atmosphere, the loss of heat during the long journey over the continent being sufficient, by the time they reach the surface in latitudes between 60° and 70°, to account for the low temperatures observed in our cold waves. These explanations appear plausible, and it may be that the excessively low temperatures observed in our severest cold waves — temperatures of 50° to 60° below zero — must be attributed to a combination of two or all of the causes mentioned. Eecent investigations into the conditions at very great elevations over the equator have shown the existence of low temperatures never experi- MARYLAND WEATHER SERVICE 397 enced at the earth's surface, e\en within the Arctic Circle. In the sum- mer of 1906 M. Teisserenc de Bort and Mr. Rotch by means of pilot bal- loons succeeded in obtaining records of a temperature of 123° below zero from an elevation of about nine miles above the earth within the Tropics. THE COLD VS^INTER OF 1903-4. One of the coldest winters on record in the history of Baltimore was that of 1903-4. In some respects it was more severe than the winter of 1855-6, generally regarded as the hardest winter experienced in the Middle Atlantic states. There were very few excessively cold days, the low average temperature for the three winter months being due to long continued moderate cold. The average temperature for the entire season was 6.3° per day below the normal. The winter most nearly approach- ing this in continued cold was that of 1892-3, with a daily departure of 5.2° below the normal for the season. Next in order comes the famous winter of 1855-6 with an average daily departure of 4.6° below the sea- sonal average. With a minimum of 2° above zero on January 5, the winter was not at all remarkable for low temperatures. The trying combination of intense cold, high wind, and snow, occasionally experienced in Baltimore winters, was entirely lacking. The season was characterized by an almost unbroken period of moderately cold weather, and an unusual frequency of snowfall, rather than an excessive quantity. The ice was very heavy; an abundant crop was cut before the first of January, an unusual event for the vicinity of Baltimore. The ice in the Bay and harbor impeded navigation to a greater extent than for many years. For a time during the second decade of January, and again in the middle of February, ice covered the entire Bay from the Susquehanna to the Patuxent, and even the larger steamers were obliged to remain in port. There was but a single " cold wave " in the technical sense of the term, that is, a fall of 20° within 24 hours to a minimum of 20°, neglect- ing the usual diurnal variation in temperature ; this occurred on the 26th of December with a fall of 22° and a minimum of 11°. 398 THE CLIMATE OF BALTIMORE There was less than the average amount of precipitation for the winter season, including snow and rain. The deficiency for the three montlis aggregated over four inches. The frequency of days witli snow was particularly significant. The normal number of sno^vs in the winter season at Baltimore is 12; in 1903-4 there were 22. It is a well recog- nized fact that a snow cover lowers the temperature materially. Hourly observations of temperature at Baltimore during a period of ten years show that on days when the ground was covered with snow the mean temperature was 10° lower than on the normal winter day. The tem- perature is lowered during the night by reason of the intense radiation from the snow surface; during the day much of the heat which would otherwise go to W' arm the atmosphere is employed in melting and vaporiz- ing the snow. Some of the most significant departures from normal winter condi- tions are indicated in the following comparison : THE WINTERS OF 1903-4 AND 1889-90 CONTRASTED. TO , Cold Warm w-^Tf Winter Winter W inter. i9n3_4. 1889-90. Number of days with a mean temperature below 32°. . 34 66 11 Number of days with a mean temperature above 40°. .20 8 63 Number of days with a min. temperature below 32°. . 59 78 28 Number of days with a max. temperature below 32°. .14 21 5 Number of days with a temperature of 20° or less. ... 18 39 6 Number of days with a measurable amount of snow. .12 22 7 Total depth of snowfall in inches 24 26 5 Total precipitation (rain and melted snow) in inches. 10 6 7 Number of days with snow on ground 22 40 — ' First killing frost occurred Nov. 7 Nov. 7 Nov. 6 Last killing frost occurred Apr. 4 Apr. 17 Apr. 2 ^ No record. The following notes on ice conditions during the winter of 1903-4 are taken from the daily journal of the United States Weather Bureau : 1903. Nov. 7. First ice. Dec. 16. Ice 4 to 5 inches thick. 1904. Jan. 4. Ice 6 to 7 inches on Druid Hill Lake. Bay frozen over from Sharp's Island to Ft. Carroll. Ice off Sharp's Island 8 inches thick. MAKYLAXD WEATHER SERVICE 399 Ice on Druid Hill Lake, S inches. Heavy ice 50 to 60 miles down the Bay. All sailing vessels and small steamers tied up. Heavy drift ice in Bay — flows 5 feet thick, making navigation dangerous. From Sandy Point up the harbor ice jams delay strongest steamers. Ice on Druid Hill Lake, 12 inches. Ice fields from Susquehanna to Patuxent. Passable only by means of ice boats. Whole Bay solidly frozen over down to Cove Point. Ice 3 to 18 inches. Navigation suspended. Fields of heavy drift ice reported as far south as the Potomac. Conditions the worst of the winter. Some ice floes have an area of 400 to 500 acres. Navigation hazardous. Navigation practically suspended. Ice in harbor again becoming a serious obstacle to navigation in spite of the good work of the ice boats. Ice conditions nearly as bad as at any time of the season. The steamer Alabama from Old Point ran into drift ice extending from shore to shore only 30 miles above Old Point. The largest steamers are remaining in port on account of ice. The ice in the Bay is causing no more serious trouble. THE WARM WINTER OF 1889-90, The winter of 1889-90 was quite as remarkable for its mildness as that of 1903-4 was for its severity. The excess in temperature was even greater than the deficiency of the winter in 1903-1, being nearly 8° above normal per day, as compared with 6° below in 1903-1. The winter passed without a single cold wave. The seasonal snowfall (5 inches) was the lightest since 1871, or since the establishment of the local office of the Weather Bureau. The number of days upon which snow fell was but 7 as compared with an average number of 12 and as compared with 22 in 1903-4. There were but six days of the season on which the temperature fell as low as 20° ; in 1903-4 there were 39 days with a temperature of 20° or below. On 12 days of the winter of 1889-90 the tempernture rose to points never reached upon those days in 30 years, and upon two days the highest temperature recorded in December and January occurred, namely, 73° on December 26, 1889 and January 13, 1890. 1904. Jan. 6. Jan. 8. Jan. 12. Jan. 13. Jan. 18. Jan. 19. Jan. 20. Jan. 29. Jan. 31. Feb. 2. Feb. 15. Feb. 18. Feb. 19. Feb. 21. Mar. 2. 400 THE CLIMATE OF BALTIMORE THE DISTRIBUTION OF ATMOSPHERIC PRESSURE DURING THE COLD WINTER OF 1903-4 AND THE WARM WINTER OF 1889-90. (PI. XXIY.) The character of the weather ol a given locality depends, so far as temperature and winds are concerned, upon the general distribution of the atmospheric pressure over a large area surrounding the locality. The pressure determines the wind direction directly, and the winds, in turn, modify the temperature. If we examine carefully a chart showing the normal winter atmospheric pressure over the North American continent, we find that the barometer is relatively low over Labrador and surround- ing regions, and high over the interior of the continent and over the central and southern states. Such a distribution of pressure, as shown in the discussion of cyclones and anti-cyclones in preceding pages, gives to Baltimore and vicinity a prevailing west to northwest wind. If this normal winter distribution of pressure is disturbed, a change in the prevailing wind directions will result, with attendant changes in tempera- ture, and other factors. If we examine the distribution of the mean seasonal pressure over the country during the winter of 1903-i, we find a distribution differing greatly from the normal. The centers of the areas of high pressure during December, January, and February were to the west and northwest of, or over Baltimore, but the areas were much more extensive and intensely developed, causing a steady and strong flow of cold northwest winds during the entire season. A winter season may be severe by reason of a considerable percentage of exceptionally cold days, or by reason of long continued moderate cold. The season of 1903-4 belonged to the latter class. If we examine, on the other hand, the distribution of seasonal pressure during the winter of 1889-90, we find a totally different condition — a wide departure from the normal type. The barometer was persistently high over the southeastern por- tion of the country — an extension westward of the permanent area of high pressure over the Atlantic Ocean. This distribution of pressure gave to Baltimore and vicinity a prevailing wind direction from the south and southwest during December and January, and an easterly direction during February. As opposed to the prevailing northwest winds of an average winter, these directions all give a relatively high temperature. MAEYLAND WEATHER SERVICE 401 In addition, the storm tracks of the winter 1889-90 were, without excep- tion, far to the north of Baltimore, diifering widely from the usual dis- tribution, and causing an unusual percentage of southerly winds. The influence of the mean distribution of atmospheric pressure upon the general character of the weather will be further considered in con- nection with the discussion of weather conditions in other seasons. THE VARIABILITY OF WINTER WEATHER. In the middle latitudes winter is a season of great contrasts. This is particularly characteristic of regions lying near the paths of the storm centers. It is not difficult to find a reason for the great and sudden fluctuations experienced when we bear in mind the conditions described in the preceding pages. In our latitudes there is a continuous succes- sion of atmospheric waves moving from west to east across the continent. Within the troughs of these waves as they move eastward we find, in any well developed type, warm southerly winds which raise the temperature of the localities over which they blow approximately 10° above the normal for the time and place. As the crest passes over the locality it brings with it a cold northwest wind, carrying the temperature an equal amount below the normal. The crests usually follow t!ie troughs within 24 to 36 hours; they may, if they are associated with a rapidly moving storm, follow one another at intervals of 13 hours, or even less. In the winter season conditions are favorable in the United States for bringing about great fluctuations in temperature in short periods of time. Just to the south we have the tropical regions where temperatures are high throughout the year, and atmospheric moisture abundant. Beyond our nortliern boundary line are the regions of long winter nights and a clear dry atmospliere, factors favoring a rapid lowering of tempera- ture by radiation. Warm and cold air are alternately brought to the regions midway between these reservoirs of heat and cold, as areas of low and high atmospheric pressure, or cyclones and anti-cyclones, follow one another in rapid succession from west to east across tlie continent. This material transfer of warm air by southerly winds and cold air by northwest winds is responsible for fluctuations of great magnitude. The 402 THE CLI]\rATE OF BALTIMORE Fig. 142.— Cold February 11, 1899. Fig. 143. — Warm February 11, 1887. MARYLAND WEATHER SERVICE 403 contrast is, in all well developed cold waves, intensified by conditions attending all cyclones and anti-cyclones. The warm, moist southerly winds are attended with cloudy skies which cut off rapid terrestrial radia- tion, thus preventing loss of heat. The cold northwest winds are dry, while the sky is generally clear, permitting rapid loss of heat by terres- trial radiation. These processes, attending all storms, are sufficient to account for the greatest fluctuations observed in terrestrial temperatures, making it entirely unnecessary to call in the aid of exti-a-terrestrial forces to explain any unusually high or low temperatures observed. The great fluctuations in temperature experienced in past years in Baltimore are discussed at considerable length in the preceding pages of this report (see especially Plate lY, following page 82) ; the conditions under which they occurred, however, are not described. Curve D, Plate IV, shows the extreme range of temperature for each day of the year in a period of 30 years. The great ranges are shown to be most frequent in the winter, months, although the month of March is not far behind in this respect. February 11 shows the greatest range — on the 11th of February, 1887, a maximum of 72° was recorded in Baltimore; on February 11, 1899, a minimum of 6° below zero, making a total range for the 11th of February of 78°. Even in the month of least variability, August, the difference between the observed maximum and minimum is 31°. The general weather conditions which prevailed upon the two days of February, showing such great contrasts in temperature are represented in the accompanying charts. (See Figs. 142 and 143.) On the 11th of February, 1887, a well developed cyclone was passing eastward with its center almost over Baltimore. Warm southerly winds had been blowiEg over the city for a considerable period, and with some force. By the afternoon of the 11th the temperature had risen to 50°. The depression was followed closely by an anti-cyclone, and on the fol- lowing days the temperature fell to a minimum of 23° as the center of the anti-cyclone approached. On the 11th of February, 1899, Baltimore was within one of tlie most extensive and intense anti-cyclonic areas recorded in local weather 404 THE CLIiLA.TE OF BALTIMORE chronolog}', two days before the great blizzard of this month. The cold northwest winds, aided by the intense terrestial radiation permitted by the clear skies and dry atmosphere, lowered the early morning tempera- ture to 6° below zero, within a degree of the greatest cold recorded in Baltimore in 30 years. Even more striking are the fluctuations in temperature attending the passage of single storms. On the 24th of February, 1900, Baltimore was within the area of influence of a cyclone, followed closely by an energetic anti-cyclone. By 7 p. m., with strong southerly winds, the temperature rose to 55°. About 8 p. m. the wind suddenly changed to a strong northwest wind and the temperature fell rapidly to 8° by mid- night, a fall of 47° in five hours. HOURLY CHANGES ON FEBRUARY 24, 1900. P. M. Mid- Xoon. 1 2 3 4 5 6 7 8 9 10 11 night. Pressure Unches) 29.43 .36 .31 .23 .19 .19 .18 .19 .18 .18 .19 .20 29.21 Temperature (F.) 43 45 46 46 49 61 62 55 37 33 28 18 8 Wind Direction SE E SE SE SE 8 SW SW NW W NW W NW Wind Velocity (miles per hour) 8 8 7 12 8 7 10 12 12 6 8 7 8 On the 31st of December, 1898, as the center of a depression passed over Baltimore and was followed by the advancing front of an anti- cyclone, the temperature fell from 57° at 7 a. m. to 18° at midnight. The usual diurnal rise in temperature to 3 p. m. was totally obliterated. The temperature continued to fall steadily to a minimum of 5° by 8 a. ra. of January 2, 1899. HOURLY TEMPERATURE CHANGES OF DECEMBER 31, 1898. A. M. P. M. Mid- 7 8 9 10 n Noon. 1 3 3 4 5 6 7 8 9 10 11 night. Temperature (F.).. 57 57 55 47 41 38 38 38 37 35 33 31 29 37 25 22 20 18 WindDirection SW SW NW NW N N N N N N N N N N N N N N The Weather of Christmas Day {December i;i5). In order to illustrate the variability of wanter weather in Baltimore, special days have been selected and the general conditions on that day for a long series of years graphically represented upon a single page. MARYLAND WEATHER SERVICE 405 As the popular interest in the weather conditions upon public holidays is always great, one or two days have been selected in each season. For the winter season we have chosen Christmas Day and Washington's Birthday. Charting the weather conditions in this manner the contrasts are more readily perceived than b}' the use of words and figures. The factors represented are: The maximum, minimum, and mean temperatures for the day, the mean atmospheric pressure, the average amount of cloudiness during the course of the day, the prevailing wind direction, and the amount of precipitation. Take for example the weather conditions experienced upon the 25th of December in each year from ISTl, the date of establishment of the local office of the United States Weather Bureau in Baltimore, to the year 1906. The mean temperature for the month of December is 37° ; for the 25th of December it is 35.5°. The maximum temperature on the 25th was on two occasions (in 1889 and in 1893) as high as 67°; in 1872 the minimum temperature of the day was 8°, a range of 59°. The fluctuations from year to year are very irregular; sometimes they are abrupt, as the change from 1871 (44°) to 1872 (14°) ; at other periods they are gradual, as from 1881 to 1884, a slight and steady fall from 40° to 28° in the mean temperature of the day. There were 9 clear days, 13 fair or partly cloudy days, and 14 cloudy days. The winds were prevailingly east to northeast and but once from the south. The day has been remarkably free from precipitation of measurable quantities, and this has been mostly in the form of rain. In not a single year since 1871 has snow fallen to the depth of one inch upon Christmas Day. Snow has been on the ground to a greater depth but in such instances it fell on the day preceding and remained on the ground. (See Fig. 144.) :§ ^ O ts. O g o C o fa c C — C • c 5 . U ; 1 \ A^ ■ ■ a~i ; ,^;; '9 ^ ^i^^' ' 5 i o o ti \ 1 < T . >i i £ N; ^> >> \ z ► ^ 1 i "-J : . en ; —1 ; k '/ / 2: CO 5' i| ^ si "CC oz f : UJ La Si : 2 m }< / ! ,^^ : 4 ~Q in <*?- ;^ H £ :\ ^V) GO 4^ < ; vl!/ : ^ ^ ^^ 5»i_ S ^ J ^ •^ -> 1 f > / o 05 r' % ^ ■ ffi ' ^ f JD - i^ ■"^ '^' f X \ V ^■ ^x ^ 1 1 i rs [i I S : \/ \: :. ': \ ^ k I /: : 1 cc t ^ / V .' fe 'A \ y // / ■ ; i w '■ ^ ' 1 >. n -*\ I i /'^ ^ o 1 \^' s N \ i ! i ^ r1 / ■ ; V J. ' i "^ ^ vJ i or ! ; uj '0 ^^ ^ ; '^ y i:-iiX-Li Irt : ^ y^ ?^ f si. = ^i ac i — ^ ^ > : s; '■■ r \ ^ ^ ■^^ ,/ t ^^ -^ ^ ^ : ii • i I ' *. 1 tr u. ; 'i 1 ; \ i V; k 15 ; 1 o ^ 50 — M o in o «rt c> {vi — — o MAKTLAXD WEATHER SERVICE 407 THE CHARACTER OF THE WEATHER ON CHRISTMAS DAY IN BALTIMORE FROM 1871-1906. Year. Max. Temp. Min. Temp. Mean Temp. Character of Day. Wind Direction. Daily Wind Movement. Precip- itation.i (De i-rees Fah r.) (Miles) (Inches) 1871 48 40 44 Pt. cldy NE 45 1872 19 8 14 Cloudy NE 159 1873 41 30 36 Pt. cldy SE 98 1874 39 30 34 Clear W 245 1875 54 40 47 Pt. cldy E&S 52 0.09 1876 27 18 22 Cloudy N&NE 85 0.06 1877 47 43 45 Cloudy E 58 0.03 1878 25 15 20 Clear W 268 1879 52 32 42 Cloudy N 95 0.38 1880 38 25 32 Cloudy NE 111 0.35 1881 49 32 40 Clear W 72 1882 47 36 42 Pt. cldy w 108 1883 40 32 36 Pt. cldy NE&SW 70 0.39 1884 32 25 28 Cloudy N 191 1885 39 30 34 Pt. cldy NE 144 1886 45 28 36 Clear N&NW 198 1887 38 31 35 Cloudy N 106 1888 53 29 41 Clear Calm 17 1889 67 44 56 Pt. cldy SW 101 0.04 1890 32 25 28 Cloudy NE&NW 114 1891 48 47 48 Cloudy SE 127 0.02 1892 29 14 22 Cloudy NW 225 0.02 1893 67 40 54 Clear SW 116 1894 48 38 43 Pt. cldy NW 171 0.15 1895 52 43 48 Cloudy E 122 0.01 1896 28 18 23 Clear W 106 1897 34 16 25 Clear S 62 1898 38 32 35 Pt. cldy NE 76 1899 35 23 29 Pt. cldy W 134 1900 50 35 42 Pt. cldy W 115 1901 48 36 42 Cloudy NE 85 1902 32 19 26 Pt. cldy W 165 0.17 1903 43 34 38 Cloudy NW 83 0.19 1904 28 23 26 Cloudy NE 194 0.14 1905 38 26 32 Pt. cldy NW 162 1906 33 13 22 Clear NW 299 Means 42 30 36 Pt. cldy 142 0.14 • Amounts loss than .01 inch not considered. 27 c ^^ O c o ^6 o' rt CvJ — « (?> o w — — .^ ■'";■■■■ ■ : ; ! \ } ' :/+\ i : oo : : ) I h- . \ \: ;i i V^^ :5= ^ ! /! >- \-^ • < I ■• • ■ ■ ■ i, 1^ /ffij • 7! 1 ^ >- a: N yH o \J cc 1 1^ C3 5- [ ii TEMPERATURE ki-^ ^ / 4/ i. •\ ^ UJ u. ; '' K K ? \ ! £ ; /| \: I ; J^ ; , j Y ^ ^ s. ^ /X : , i i > ^ ^ / ; i \\ M 1 '■ M : / v\ s^ : • r^ ■ 1 s %l / (T^^ *■ : ^ 7; i ^ ■ '^■ ; \ ,_ \ ^^ v \ : * ^ =^ < / n\ f "^ ^ V (): y / : \ ^ '^U i\ 1 \ ; y i\ h /■\ 1 TEMPERATURE" / 1 :| ; ^ \[ ^ \ : ^=^- ■ •-Tn ■ \\ f^ viks / ;o^ /' \ ^\ f 1 c= 1 =3 _v-/ o V UJ •*• - :i:-fc» o /"^ ^- 1 i: \ \ \ i / Q: /I: 1 2 — a:" <: .> ^ J::^ ■^ \ i: "^ ^ ^: -^ St J: 1 7 ./ § V r- \ \^ i^ y ^ J i T-^ : ^__ : CQ MARYLAND WEATHER SERVICE 409 The Weather of Washington's Birthday (February 22). The weather of the month of February presents more contrasts than that of any other portion of the year. The 22d is no exception ; with an average of 38° the temperature was as high as 74° in 1874, and as low as 13° in 1896, a range of 61°. From 1878 there was a steady fall in the mean temperature of the day to the year 1885, though there is apparently no regular period discernible in the annual fluctuation. The grouping of the days with precipitation is somewhat striking. From 1871 to 1890 there were but two occasions upon which rain or snow fell to any considerable depth ; namel}-, in 1876 with a rainfall of three- tenths of an inch, and IS 78 with a heavy rainfall measuring over an inch and seven-tenths. Snow fell in 1879, 1882, 1883, and in 1889, but the fall was extremely light in all cases. From 1891 to 1907 rain or snow fell in 1891, 1893, 1894, 1897, 1900, 1901, 1902, and in 1904, though the amounts were small in 1892, 1898, 1899, 1903, and 1907. Fig. 145. THE WEATHER OF FEBRUARY 22. Year. Max. Temp Min. Temp. Mean Temp. Character of Daj'. Wind Direction. Daily Wind Movement. Precip- itation. (Degrees Fab r.) (Miles) (Inches) 1871 35 20 28 Clear W . . . . . . 1S72 35 28 32 Clear NW 170 1873 35 24 29 Pt. cldy • NW 199 1874 74 51 62 Pt. cldy SW 173 1875 48 29 38 Pt. cldy E 97 1876 46 36 41 Clear W 166 0.34 1877 57 30 44 Pt. cldy SE 66 1878 63 50 56 Cloudy W 196 1.71 1879 36 21 28 Cloudy S 167 0.05 1880 50 31 40 Clear S 110 1881 47 32 40 Clear SE 143 1882 40 34 37 Pt. cldy NW 279 1883 43 31 37 Cloudy SW 95 0.01 1884 49 36 42 Pt. cldy SE 90 1885 27 14 20 Clear NW 183 1886 43 32 38 Clear S 168 1887 54 35 44 Pt. cldy N 158 1888 56 35 46 Clear N 74 1889 36 30 33 Cloudy SW 81 0.09 1890 44 25 34 Clear w 127 410 THE CLIMATE OF BALTIMORE Year. 1891 Max. Min. Mean Temp. Temp. Temp. (Degrees Fahr.) 47 38 42 Character of Day. Clear Wind Direction. NW Daily Wind Movement. (Miles) 203 Precip- itation. (Inches) 0.54 1892 49 39 44 Pt. cldy NE 260 1893 32 24 28 Clear W 524 0.21 1894 35 27 31 Clear W 109 0.32 1895 32 29 30 Pt. cldy NW 300 1896 42 13 28 Clear sw 180 1897 43 36 40 Cloudy sw 201 0.71 1898 43 34 38 Cloudy w 111 1899 60 42 51 Cloudy w 122 1900 57 41 49 Pt. cldy w 164 0.54 1901 34 21 28 Pt. cldy w 120 0.01 1902 36 34 35 Cloudy N 224 0.40 1903 34 25 30 Clear NW 227 1904 47 34 40 Pt. cldy NW 240 0.74 1905 39 31 35 Cloudy NB 240 1906 54 39 47 Clear N 253 1907 26 16 21 Clear NW 249 Means 45 31 Pt. cldy 179 0.44 SPKING WEATHEE. The spring season is a transition period between winter and summer weather conditions. Normally the season has three months, but May is practically a summer month, while March frequently has more of the characteristics of winter than spring. The season, so far as injurious weather conditions are concerned, is very brief. The cyclones and anti- cyclones of early spring belong to the winter type; they are quite as energetic, and follow much the same paths. With the steady approach of the sun the increased heat becomes more apparent, however, and the contrasts in temperature between the winds preceding and following the travelling cyclones become more marked. One of the first harbingers of spring is the appearance of an area of high atmospheric pressure off the coast of the South Atlantic states. This high area may have advanced from the northwest and, after crossing the continent, settled off the coast, but it is more likely to be an inde- MARYLAND WEATHER SERVICE 411 pendent formation in place, or the westward extension of the permanent area of high pressure normally found over the Atlantic Ocean between the Azores and the South Atlantic states. The presence of an anti- cyclone in the southeast gives to the Middle Atlantic states a steady flow of warm southeast to southwest winds. March weather is extremely variable. The month has given us some of our severest winter weather, such as the blizzard of 1888, described in preceding pages; it may also be excessively cold and raw, as in 1906. On the other hand, there may be an abundance of fine warm days forcing all vegetable growth four or five weeks in advance of the average season, as in March, 1898. The explanation for these striking contrasts may be found in the general distribution of atmospheric pressure over the North American continent and adjacent oceans during the early spring season. Under normal conditions in the month of March there is a well devek)ped area of high pressure over the central portion of the continent — the British Northwest Territory; another area of high pressure prevails over the Atlantic Ocean with its axis along the parallel of 30° north latitude, extending westward nearly to the Atlantic coast. These areas are not the same as the travelling anti-cyclones described in the preceding pages. They remain nearly stationary for long periods of time, but are sub- ject to more or less shifting about a central point from time to time. The travelling anti-cyclones are probably detached portions of the larger areas. The character of the weather within these so-called permanent areas of high pressure is the same, however, as is found within the smaller travelling anti-cyclones. Normally, the Middle Atlantic states lie in a belt between these two high areas, and alternately fall within the influence of first one, then the other. One brings us cold weather, the other warm. Occasionally the continental high area will extend to, or move southward and eastward of, its normal limits and bring within its influence the whole of the Middle Atlantic states. Such was the case in the years 1883,' 1885, 1888, and 1801. Tlie month of :March ^8ee: 0. L. Fassig. Types of March Weather in the U. S. Amer. Jour. Soi., Nov., 1899. 412 THE CLIMATE OF BALTIMORE in these years was decidedly below the normal in temperature throughout the Middle i\.tlantic states. Again there may be a weak development of the continental high area, in conjunction with a strong development of the Atlantic Ocean high area, or its extension westward beyond its usual limits. The Middle Atlantic states would then be brought within its influence and produce strong southeast to southwest winds or light variable winds and high temperatures. During the month of March in 1878, 1882, 1894, and 1898 the distribution of the monthly mean pressure was such as is indicated, and in all of these months the tempera- ture was well above the normal value. (See Plate XXIV.) March Winds and Storms. March is proverbally a windy month. The wind movement for this month is the largest of the year, exceeding even that of the winter months : AVERAGE DAILY WIND MOVEMENT AT BALTIMORE. (Average of 30 years.) Jan. Feb. Mar. Aiir. May. June. July. Aug. Sept. Oct. Nov. Dec. Year. Miles 143 163 175 166 119 143 134 122 129 137 143 143 145 This comparatively high wind velocity is probably due to the great contrasts in temperature, characteristic of the month. It is the time of the year when the temperature rises most rapidly and the inter-diurnal changes in temperature are greatest: MEAN DAILY CHANGE IN TEMPERATURE AT BALTIMORE. (Based on 30 years of observations.) Jan. Feb. Mar. Apr. May. June. July. Aug. Sept. Oct. |Nov. Dec. Av. daily change± 1.0° 1.0° 1.2° 0.8° 1.0° 0.6° 0.6° 0.4° 0.9° 0.:° 1.0° 0.8° F. March follows February in the frequency and duration of storm winds : AVERAGE DURATION AND FREQUENCY OP STORM WINDS AT BALTIMORE. (Average of 5 years' observations.) Average monthly f requencj- ) of winds exceeding25 miles > 6.3 9.3 7.6 per hour. ) Average duration in hours ' o i-n . and minutes. \ ^■''" * Longest periods of continu- I ,„ ous storm winds (hoursj. i' fe < a >^ ic o O o ® 9.3 7.6 5.0 4.4 2.0 2.6 1.1 3.6 4.3 ■4.8 4.4 1.40 3.40 3.30 1.30 0.13 0.30 3.00 1.25 3.40 2.00 3.00 46 23 16 7 0.5 0.5 IS 7 9 6 19 MARYLAND WEATHER SERVICE. VOLUME 2, PLATE XXM. EFFECTS OF THE ICE-STORM OF MARCH, 1906 IN THE BLUE RIDGE MOUNTAINS. MARYLAND WEATHER SERVICE 413 ICE STORilS. While the afternoon temijeratures during March are well above the freezing point, the early morning temperatures generally dip below the frost line. Hence the isotherm of 32° is frequently crossed during the passage of the storms of this month. Precipitation, beginning as rain, changes to sleet or snow, or the rain as it falls upon the cold ground or trees and shrubs freezes, covering all exposed objects with a coating of ice. These ice storms are of frequent occurrence in the early spring, but generally occur in March. They frequently cause great damage to property by overloading trees, telegraph lines, etc., and are a source of considerable danger to pedestrians on the city streets. The ice upon trees and telegraph lines sometimes collects in great quantity. Under favor- able conditions in the presence of a fog or, in the mountain districts in a driving low cloud, the frozen particles of fog or cloud collect upon the windward side of exposed objects to a thickness of two and even three inches. (See Pis. XXI and XXII.) THE SQUALL OF MARCH 1, 1907. A type of storm which occurs with increasing frequency with the advance of spring is shown on the weather map of March 1, 1907. The eastward advance of the storm across the country is marked by the occur- rence of severe squalls and thunderstorms, accompanied by heavy rains, within a restricted area of the general storm, or cyclone. The isobars enclosing the storm area form long narrow troughs which are likely to be particularly well marked as the cyclone passes across the Mississippi Valley. Drawing an approximately north and south line through the center of the storm the winds to the east blow from thu southeast, while those to the west blow from the northwest. In the vicinity of this line of conflict between the southeast and northwest winds, from the center of the storm southward, we have numerous squalls, thunderstorms, and heavy rains. These local storms occur almost entirely within the south- ern quadrant of the general cyclone of the type here described. The contrasts in temperature between the southeast winds to the east of the "squall line," as the north-south line above described is sometimes 414 THE CLIMATE OF BALTIMORE calltjd, and the northwest winds to the west, are very pronounced; these contrasts are especially well marked in passing from the center of the storm in a northwesterly direction. The map of March 1, 1907, shows a difference of 50° at 8 a. m. between St. Louis, Mo., and Huron, S. Dak., a distance of about 500 miles. The chart also shows the distribu- tion of thunderstorms, occurring within the 12-hour period preceding 8 Fig. 146.— The Squall of March 1, 1907. a. m. of this date. The rains of the Lower Mississippi Valley were heavy, and in some localities, excessive: Mobile, Ala., reported a fall of 6.42 inches in the preceding 24 hours; Montgomery, Ala., 1.32 inches; Anniston, Ala., 1.16 inches; Meridian, Miss., 2.64 inches; Little Eock, Ark., 1.42 inches; and Memphis, Tenn., 1.30 inches. (See Fig. 146.) Later in the season, with the increasing heat of spring, the most intense variety of local storm — the tornado — is frequently developed within the thunderstorm and squall area of this type of general storm. As the MARYLAND WEATHER SERVICE 415 storm moves farther eastward the squalls and thunderstorms continue to develop, but the tornado becomes of less frequent occurrence, disappear- ing almost entirely by the time the center of the general storm reaches the coast. In Maryland, for example, a real tornado is of very rare occurrence. The storm described above changed its shape materially during the succeeding 24 hours, becoming by 8 a. m. of March 2 more circular in form, with its center over the Lower Lake region. In changing to the more common cyclonic type the change in wind direction during the passage of the center of the storm became less abrupt and the squally character of the shift in the wind disappeared to a great extent. EQUINOCTIAL STORMS. There is a widespread popular belief in the occurrence of severe storms at the equinoctial periods in March and September. Just why there should be any unusual atmospheric disturbance when the sun " crosses the line " has never been made clear. The belief is an old one, and is especially firmly fixed in the minds of sailors. Like many of the old weather " saws,'' it will not stand the test of rigid comparison with re- corded observations. As a rule, people are not very critical in their defini- tions of storms, or in verifying the time of their occurrence. In the ab- sence of a severe storm a very mild disturbance will satisfy them. Or if the storm should occur three or four days preceding or following the equinoxial day it is an equinoxial storm ahead of time or a little delayed in its arrival. With such elastic restrictions it is not difiicult to realize an equinoctial storm in any month of March. But under these condi- tions a storm is just as likely to occur upon any other day in the month. If a disturbance be required to show a wind velocity exceeding 25 miles per hour in order to be classed as a storm, there is, according to the Baltimore records, a storm wind every fourth day. If uniformly dis- tributed through the month any clay might be selected for a storm and an error of more than two days in time could not be made. Moreover, the records show that the wind velocities on the 21st and 22d of Marcli are not greater than on the days preceding and following. If we con- 416 THE CLIMATE OF BALTIMORE sider the occurrence of gales, or winds exceeding 40 miles per hour, we find that during a period of 30 years there were 42, distributed through the year as indicated below : FREQUENCY OF GALES NEAR BALTIMORE. Jan. Feb. Mar. Apr. Maj\ June. July. Au^. Sept. Oct. Nov. Dec. Year. 3 10 3 2 2 3 4 3 3 5 6 42 September, the month of the autumnal equinoctial storm, is the only month in the year without a gale to its credit in 30 years, while the month of March has less than the average monthly frequency. If we regard rainfall as one of the essential features of a storm, statis- tics are no more favorable to the equinoctial theory than they are in the case of winds. On March 21 the rainfall frequency, based on 31 years of observations, is slightly less than 50 per cent; that is, in 31 years, rain occurred 15 times; on the 22d the percentage of frequency is 40 per cent. On these days rain has occurred less than half the time. Eainfall frequency is somewhat greater on the 19th and 20th, having occurred 16 times in 31 years. (See Table XLIV, page 181.) When considering amount of rainfall instead of frequency, statistics are even more unfavorable. The average amount of rainfall recorded in Baltimore during 30 years on the 21st of March is 0.21 inch. The amounts for March 19 and 20 are decidedly greater, namely, 0.45 inch and 0.37 inch respectively. The daily average for the entire month is 0.31 inch; hence the amount recorded on the 21st is below the average for the month. Similar statistics may be shown for the September equinoctial day : On September 21, the average rainfall is 0.16 inch; for the 19th it is 0.93 inch, nearly six times as much; for the 20th it is 0.23 inch; for the 22d, 0.30 inch. The average daily amount for the entire month of September is 0.30 inch. For rainfall frequency in September we have: September 19 26 per cent. 20 30 21 26 22 32 MARYLAND WEATHER SERVICE 417 The daily average for the entire montli is 30 per cent. Eain occurs on the 21st day of March on the average in approximately alternate years; on the 21st day of September every third year. The equinoctial storm, as a destructive storm, or as a storm confined to a single day in the vicinity of Baltimore, is a myth. Hail Storms. True hail is of infrequent occurrence in the winter months. While it is often reported in the cold season a careful observer would in nearly every instance report sleet. There is a radical difference between sleet and hail, both in the manner of formation and in physical character- istics. Sleet is an intermediate stage between rain and snow, or a mix- ture of the two forms, and occurs during the passage of a storm when the temperature crosses the freezing point. Hail, on the other hand, occurs almost entirely during the warm season, when the surface tem- peratures are far above the freezing point of water. THE FREQUENCY OF OCCURRENCE OF HAIL IN BALTIMORE. (Total number in 28 years.) Jan. Feb. Mar. Apr. May. .Tune. .Tuly. Aug-. Sept. Oct. Nov. Dec. Year 4 3 3 3 II 13 9 6 1 1 1 55 It will be noted by the above figures that over half of the hail storms reported in Baltimore in 28 years occurred in the months of May, June, and July. Further details as to frequency and time of occurrence may be found on pages 284-287 of this report. The most favorable period of hail formation appears to be the latter part of the spring season, and the early summer. The hail storm, like the thunderstorm, is intimately associated with the general cyclone. It occurs in the southern quadrant of the general storm, along the " squall line," described in a preceding paragraph. Hail storms occur almost exclusively in connection with thunderstorms. The reverse of this state- ment is, however, not true, while but 55 hail storms are recorded in the local annals of the Weather Bureau in a period of 28 years, there is a record of 678 thunderstorms during the same period. 418 THE CLIMATE OF BALTIMORE THE HAIL STORM OF MAY 19, 1904. On Ma}^ 17 and IS, 1904, a condition of unsettled weather prevailed over the country east of the Mississippi Eiver. Cloudy and rainy weather accompanied the slow eastward movement of a shallow barometric depres- sion from the Middle Mississippi Valley to the Middle and South Atlantic states. On the morning of the 18th the center of depression was over North Carolina. From the morning of the 18th to the morning of the 19th the storm became more limited in area and moved rapidly northward. At 8 a. m. the center was over Lake Huron, with an exten- sion southeastward, forming a secondary depression. There Avas a well defined line of separation between southeast and westerly winds, extend- ing from the center of the storm southeastward across Central New York, Eastern Pennsylvania, and Southern New Jersey. During the 19th the storm moved slowly eastward, accompanied by severe local storms and heavy rains in the Middle Atlantic and New England states. In Maryland the thunderstorm was accompanied by hail between two and three o'clock in the afternoon. (See Fig. 147.) The local conditions prevailing at Baltimore during the progress of this storm are shown in detail in the accompanying diagram (Fig. 148). THE HAIL STORM OP APRIL 27, 1890. A hail storm of unusual severity passed over the city on the 27th of April, 1890. A detailed account of local changes appears in the daily journal of the office of the Weather Bureau, from which we quote the following : Dense fog prevailed in the morning, gradually disappearing during the forenoon. The temperature rose rapidly from 48° at 8 a. m. to 71° in the afternoon. Cautionary southeast storm signals were displayed at 10.45 a. m. The most destructive hail storm on record at Baltimore visited the city this afternoon between 3.45 and 4 o'clock. It came from a point between west and northwest, and travelled in a direction a little south of east. A half-hour before the arrival of the storm, a dense black cloud-mass, tinged with purple and green, was observed extending from the western horizon across the sky to the northeast, and rapidly approaching. There were at this time two or three subdued peals of thunder, following some vivid flashes of lightning in the west. The front of the bank was in great commo- tion and. as it approached, appeared to be preceded by a thin misty veil of MARYLAND AVEATHEE SERVICE 419 o H,HIGH HIGH \ A< LOV \ .3 /^ 50° X. \^^ GH \ // U'f LOVVl V l7~^ ^-""^ ^/J ^ ->' — — - -/: - ^'^ 7 0- Fig. 147.— The Hail Storm of May 19, 1904. MDT. MAY 1 9 1 904 2 9': 52 WIND DIRECTION V ■/// ^W 10 10 IS fe 4 ;{ II S 8 8 t) §^ WIND MOVEMENT ta Tr Tr. Fin. 148.— The Hail Storm of May 19. 1904. 420 THE CLIMATE OF BALTIMORE cloud. There was a sound like the roll of musketry, and the storm burst sud- denly upon the city with an almost deafening roar as the great hail stones rained down upon the tin roofs and crashed into the windows, not a building in the path of the storm escaping damage. Several persons were knocked down by the stones, and many, including a number of children, were cut and bruised. Horses, pelted and cut until the blood streamed from them, could not be controlled, and many ran away, damaging the vehicles and injuring the occupants. Rain fell in torrents with the hail (0.80 inch falling between 3.45 p. m. and 4 p. m.), poured through the shattered skylights and windows, and flooded houses. The streets were like rivers, and in many places the street-car tracks were covered to a depth of 6 inches by the soil washed down from adjacent hills. To add to the general disaster the wind blew with great violence, unroofing buildings, breaking in the remains of windows, uprooting trees, and giving the hail stones a dangerous slant to the eastward. For 15 minutes the city was in a state of complete panic, and then the storm passed away almost as quickly as it had come. A half-hour after the storm had left the city, the rain had nearly ceased, the wind was again light, and a rainbow appeared in the east. The hail stones M'ere as large as hen's eggs. Several measured more than 2 inches in diameter. Three weighed together 12 ounces. Some as large as a man's fist were reported by reliable parties as having fallen in West Baltimore, where the damage by hail was greatest. The hail stones were of various formations. One large stone was covered, on the side unbroken by its fall, with a number of sharp-pointed prisms, and there were many others like it. A large number were oval in form, and these on examination, ex- hibited a lamellated structure, being composed of alternate layers of trans- parent and opaque ice, commencing at the center with an opaque nucleus. Others were spheroidal in form and were similar in structure to the oval ones, and like them, very hard. The large prism-covered stones were com- posed, in the center, of a mass of sponge ice, and were generally crushed upon striking the pavement; no lamellated structure was observed in these. Although the rain commenced falling at 3.40 p. m. and ended at 5 p. m., it was excessive only between 3.45 p. m. and 4 p. m. when the 0.80 inch referred to above fell. There was a marked fall in temperature during the progress of the storm. The maximum of the day, 71°, occurred some time before the storm's arrival. Between 3.30 p. m. and 4 p. m. the temperature fell from 67° to 52°, rising again to 60° after the storm had passed. The wind, which had been very light all day, upon the approach of the storm suddenly veered from the southeast to west-northwest and blew with rapidly increasing force, reaching a velocity of 30 miles per hour at 5 minutes before 4 o'clock. After 4 o'clock it decreased in force and again became light. From information received from suburban points, it is estimated that the hail band began about 10 miles west-northwest of the city and terminated 5 miles east of the city, and extended in breadth from the southern limits to MARYLAND WEATHER SERVICE 421 5 miles to the north. This would make the baud about 25 miles long and about 10 miles broad. The amount of property destroyed is estimated at from $60,000 to $100,000, the damage for the most part being confined to windows of western ex- posure — a great many thousands of which were broken — and to skylights and greenhouses. A few windows with a northern exposure were also broken. The damage from the wind was great, a number of houses losing their roofs, while many trees in all parts of the city were blown down. There was no loss of life. The physical structure of hail stones is well known. There is a central nucleus of opaque snow or ice, consisting of snowflakes and ice crystals mixed with air bubbles. This central nucleus is surrounded by a series of thin alternating layers of clear ice and opaque snowy material. There may be as many as 10 or 12 of these layers. The diam- eters of the stones vary from a few tenths of an inch to three and even four inches. The variation in the shape of the stones is also very great. Although the theory of hail formation has received a great deal of attention it is still in a very unsatisfactory state. The explanation of the method of formation of the successive layers of packed snow and clear ice presents great difficulties as we are obliged to rely almost wholly upon speculation as to the physical processes which go on within the heavy cloud mass which constitutes the laboratory of the hail stone. The outward form of the tall '" thunder head " identified with hail storms is being carefully studied, and we may hope soon by means of instruments carried aloft by kites and balloons to gather some valuable facts as to the physical processes going on within, w^hich will eventually lead to a better understanding of hail formation. Spring Frosts. Marked falls in temperature have a special significance in April and the early part of May in most sections of the Middle Atlantic states. During an average season spring fruits have passed the critical period of injurious frosts by the middle of April in the vicinity of Baltimore, as the average occurrence of the last killing frost falls within the first week of this month. Frequently, however, a killing frost will occur in the latter part of April, and on rare occasions in the first decade of May. 422 THE CLIMATE OF BALTIMORE During April light to heavy frosts are generally looked for when a pronounced area of high pressure (an anti-cyclone) passes directly over the Middle Atlantic states from the west or northwest. In addition to the fall in temperature occasioned by the actual transfer of masses of cold air from the northwest or north into the Middle Atlantic states, there is a still further reduction in temperature, owing to the rapid loss of heat Fig. 149.— The Frost of May 9, 1906. during the night within the anti-cyclone ; the dry air and cloudless skies accompanying these areas facilitating rapid radiation from the ground. When, as frequently happens, the anti-cyclone is preceded by a baro- metric depression accompanied by rain, the probability of the occurrence of frost is greatly increased, as the atmosphere is then charged with moisture on the approach of the fall in temperature within the anti- cyclonic area. If at the time of the 8 p. m. observation the temperature is between 40° and 50° at Baltimore, and the arrival of a pronounced ilAEYLAND WEATHER SERVICE 423 anti-cjclone is expected during the following night from the west or northwest, frosts are very likely to occur in the early morning hours. The injury resulting from a frost depends, not only on the fall in tempera- ture, but also upon the state of vegetation, the amount of moisture in the atmosphere, and upon the wind movement. Clear, quiet nights greatly facilitate the production of frost in the lower places, allowing the colder, heavier air to settle near the ground. A clouded sky will prevent rapid radiation from the ground. A moderate -wind movement, by thoroughly mixing the air, will prevent any great difference in temperature between the layers near the ground and the air at higher levels. A frost occurring shortly after the appearance of tender plants is likely to do more damage than a heavier frost later on when the plant has become more vigorous. For details concerning the occurrence of spring frosts see pages 129 to 135. The general weather conditions on the morning of May 9, 1906, show a situation from which frost may be expected in the Middle Atlantic states during the following night. Temperatures ranging between 28° and 35° occurred in nearly all counties of Maryland on the 10th and 11th. This anti-cyclone was the occasion of one of the severest spells of cold weather experienced in Maryland so late in the season. All stations in the mountainous portion of the state, and many stations elsewhere, experienced temperatures below freezing. Even the southern portion of the Eastern Shore Avas not exempt, Salisbury reporting 29° and Princess Anne 31°. At a number of stations the temperature did not fall below 40°. The minimum in Baltimore was 38° on the 10th. While the frost was quite severe, fruits and vegetables were too far advanced to suffer any very great amount of injury. (See Fig. 149.) ICE WITHOUT FROST. The weather map of 8 a. m., April 17, 1905, shows freezing conditions throughout Maryland, but no frosts were reported. A well developed high area with its crest extending from Montana southeastward to Florida was associated with a deep cyclone centered over the Lower St. 28 424 THE CLIMATE OF BALTIMORE Lawrence V'alley. Strong, steady west winds prevailed as a result of this distribution of pressure over the Middle Atlantic states. The dry air aided by considerable movement prevented the formation of frost, though ice formed in a number of places. (See Fig. 150.) Periods of Unsettled Weather. In the eastward drift of cyclones in our latitudes, the rain area occupies from one to two days in passing a given point. In 76 per cent of all Fig. 150. — Ice without Frost, April 17, 1905. occurrences of precipitation the rain or snow falls within a forty-eight hour limit in Baltimore. In 13 per cent of instances the precipitation covers all or a portion of three days. This leaves very little margin for extended periods of consecutive days with rain or snow. (See page 213.) Long periods of unsettled weather with rain or snow are of most frequent occurrence in the spring season. MARYLAND WEATHER SERVICE 425 PERIODS OF UNSETTLED WEATHER. (With 6 or more consecutive days of rain or snow.) D. J. F. M. A. M. J. J. A. S. O. N. Y. Total frequency ill 34 years 13 II 15 20 14 23 13 12 IT 9 7 11 161 Maximum duration (days) 22 19 IT 22 13 23 18 15 19 12 10 10 23 Number of intervening daj'S with- out rain 644214324001 4 Seasonal frequency 38 57 42 27 164 These periods of unsettled weather may be due to a great variety of causes. There is not a regular and periodic succession of well developed cyclones and anti-cyclones, such as have been described in preceding pages. The well developed types have been selected for illustrative pur- poses, as they are simple in outline and more readily interpreted. In studying the actual weather map from day to day we find there are no two weather conditions exactly alike ; there is an infinite variety in the outline of isobars, the trend of isotherms, the shape of rain, areas, etc. An unusual succession of rainy days may be due to the very slow move- ment of a cyclonic area, to a rapid succession of storms, to a recurving of the storm upon its path, or to a combination of these causes. It may also be due to the persistence of a so-called " flat map," a map without well defined cyclones or anti-cyclones. THE RAINY PERIOD OF APRIL 19-25, 1901. On the morning of April 16, 1901, a depression entered the United States in the extreme Southwest; at 8 a. m. its center was over Arizona. This depression travelled slowly eastward, causing moderate to heavy rains over Texas and the West Gulf states during the ITth and 18th. By the morning of the 18th the center was over Alabama. In connec- tion with another depression centered over Ohio a long trougli of low pressure was formed extending from the Lower Lake region to the Gulf of Mexico. By the morning of the 19th the two depressions had merged into a single storm with its center over Georgia. Under the influence of these two storm centers, aided by an area of high pressure in the extreme Northeast, east to northeast winds set in at Baltimore, and rain began to fall on tlie 19th. The storm turned sharply up the coast on the 19th, ac- companied by a very heavy rainfall. At 8 a. m. of the 20th the center was 426 THE CLIMATE OF BALTIMORE over North Carolina and Southern Virginia. From the morning of the 20th to the evening of the 21st the storm center had travelled only from Southern Virginia to Western Maryland. On the morning of the 22d the center was found over the Ohio Valley, having been deflected westward — a rare occurrence. The winds at Baltimore continued to blow from an east- erly direction. This depression began to fill up and two secondary centers of low pressure developed along the coast, one over Eastern Maryland and the other over the South Atlantic states. By the morning of the 24th these two secondary depressions had merged into a single storm center off the coast of Delaware and New Jersey. This storm moved slowly northeastward just off the coast, disappearing by the morning of tho 26th. This very slow movement and peculiar path of the original storm, and the subsequent formation and sluggish progress of the secondary depressions in the neighborhood of Maryland kept Baltimore within their rain areas for six successive days. The amounts recorded upon each day of this period were not large, but the six days' total exceeded two inches. DAILY RAINFALL APRIL 19 TO 25, 1901. April 19, 1901 0.06 inch. 20, " 0.56 21, " 0.54 " 22, " Trace 23, " 0.11 " 24, " 0.71 " 25, " 0.05 Total 2.03 inches. The rainfall period lasted 162 hours, but the precipitation was not continuous, scattered showers occurring on the 21st, 22d, 23d, and 25th. THE RAINY PERIOD OF MAY lG-26, 1894. There is a general impression that rainy periods of much greater length than that recorded above are of frequent occurrence; a close inspection of records, however, will reveal the fact that there are inter- vening days without a trace of rain. MARYLAND WEATHER SERVICE 427 One of the longest periods noted in the Baltimore records of unsettled weather with daily rainfall was that of the 16th to the 26th of May, 1894. This period was connected with the passage of a Lake storm of great extent and energy. Here again the path of the storm after leaving the Lake region was peculiar, while the progress was very slow — in fact the center was nearly stationary in the vicinity of Maryland for the greater part of three days. The storm had its origin over the Northern Eocky Mountain slope on the 15th. It moved slowly to the Lower Lake region until the 18th, then dipped abruptly southward and remained over Virginia, Maryland, and West Virginia until the depression gradually filled up on the 21st. In the meantime a second depression developed over Georgia and Xorth Carolina and moved slowly up the coast, disappearing off the coast of New England on the 26th. From the 16th to the 26th Baltimore was within the rain areas of these two storms and much of the time very near their centers. The rainfall recorded during this period was as follows: DAILY RAINFALL AT BALTIMORE FROM MAY 16-26, 1894. May 16, 1894 0.13 inch May 22, 1894 0.01 inch 17, " 0.16 18, " 0.51 19, " 0.09 20, " 1.07 21, " 0.19 23, " 1.34 24, " 0.16 25, " 0.05 26, " 0.14 Total 4.45 While the entire period covered by the rainfall was a little over 10 days, there were but 63 hours of actual rainfall. (See pages 174 and 219 et seq. for frequency and duration of wet spells.) The most notable instances of a long continued rainfall occurring since hourly records were begun by the United States Weather Bureau in 1893 at Baltimore, was that of April 27 to May 1, 1895. The records show that rain fell for 102 consecutive hours. Though the rain was reduced to a light mist at times, it never entirely ceased during this period. The total amount of rainfall was 3.69 inches. There was no 428 THE CLIMATE OF BALTIMORE well defined storm area in the vicinity of Baltimore. The barometer was high over the New England states, and a shallow though ill defined depression covered the Gulf of Mexico; the depression moved slowly northward and eastward some distance off the coast, causing a steady northeast wind at Baltimore — a mild " northeaster." There is a myth associated with the occurrence of rainfall on St. Swithin's day which seldom fails to receive the attention of the press on the 15th of July: " St. Swithin's Day, if ye do rain. For forty days it will remain; St. Swithin's Day, an ye be fair, For forty days 'twill rain nae mair." The nearest approach to a fulfillment of this prophecy, according to the Baltimore rainfall records of the past 36 years, is a period of 9 con- secutive days with rain in the month of July. (See page 213.) The Variability or Weather in Spring. As an illustration of the changeableness of weather conditions in the spring of the 3'ear, we may present the irregular fluctuations from day to day, or we may show the changes which have taken place upon the same calendar day of each year for a long series of years. As regards the degi'ee of variability in temperature the month of March ranks with the winter months. Throughout the early spring rapid fluctuations and strong contrasts in the conditions of successive days are of common occurrence. Owing to the general interest in the character of the weather on the 4th of March, at least once in four years, this day is selected as a type of March weather. Whatever reason there may be from a historical point of view in favor of continuing to inaugurate our presidents on the fourth of March there is little to recommend the day in past experience of weather conditions and in the small chances for a favorable day. The day falls within a period of rapid warming up in the northern hemi- sphere, but the advancing sun has not yet carried the day beyond the realm of freezing weather. The early morning temperatures are nearly always below 32°, while the frequent incursions of cold air from the MARYLAND WEATHER SERVICE 429 north and west bring even the average heat of the day close to the freez- ing point on most occasions. Added to the discomforts of a raw cold atmosphere we have the pro- verbial March wiads, not infrequently combined with sleet or rain or snow, or a combination of all of these disagreeable elements. The probability for a fine day is so small that it is surprising that the efforts to change the date of inauguration to the latter part of the fol- lowing month have not yet been successful. THE WEATHER OF MARCH 4. The condition of the weather on the 4th of March is graphically repre- sented in the accompanying diagram for each year since 1871. It might have added interest to carry the diagram back to an earlier date but the period of 37 years represents practically all the chief combinations likely to have occurred in the longer period. While the conditions charted represent Baltimore weather, the close proximity of Washington makes it improbable that there were at any time any material differences in the weather conditions of the two cities. There is most certainly no difference in the variability of the elements. The average daily temperature on the 4th of March is not far from the point of frost formation. With the usual daily range the fluctuations will be above and below the frost line. When precipitation takes place it is apt to change from rain to snow or from snow to rain, with the intermediate stage of sleet. In 1873 the temperature fell to 5° above zero in the early morning of the 4th; on the following 4th of March, 1874, the afternoon temperature registered 68° above. In 1880 the temperature rose as high as 74°. Between these widely separated limits the temperature has kept up a continual see-saw about the middle point of 38°. In the past 37 years rain, snow, or sleet has fallen on 16 occasions, just 46 per cent of the days. The average amount of precipitation is about half an inch and the average duration about 9 hours. The sky has been overcast 10 times, partly clouded 15 times, and clear 11 times. The prevailing winds have been from the west and northwest. o o o o !>. to cO t- ^ 1 o ^ i.? i - C > ^ - 5 O lO - — do ^ i ; ^rh' ■ '-^ ! I '- 1 i /. ^ ^ u ii:; X.: n i ; I o 'ii 4 ^ / r; ■^ ^ ^ : > ; 1 £ A v\ / ^ N v- X : 1 f ; UJ or \i i s ^ j ^: /I / or. J^^ : -5 CC < 1 ai' V \ \ \ o t9i a> y ^' \s ^IV;^ en i V \ : X^ />; ^A- / : I S«- <;;. ^S \ ^i'' i it^ -<'■' >y :^^' ,•* : N sN, \ Ai' 11 '■ © // i V : ^ ^ i/S; \ i i ^ \ ' ^ H i9i irTi' N \ / i^^ ^ ■ } / ■rih' ^ i / 1 1 Ia^ o / k/' J :0 V 1 £ i <;■- "^ ^ •n^ ' f ■ ■ >/^ / ^n'^ ^ ■ ; X ]/' ; ^S^' ;»■ • V sX, \ ^ ■ ^ •2^ ' ! ? .-^ ^ =^'' 5^ / ' ^rh^ ■ i *^ ^ ^> >^^ / 1 |Ti^ ! f // V N.^ jT: 1> '■ LlJ ct- a. 5 Zl 5 ( : ;9;' fl : i s s;- s ;Y/ / ^ ; 1 ^ ^ o o r^ ■ 2S ^ ^y ^"i i 1^ T^ ^^ ? LlJ j^ : s « ^ i .•-' ^^ / s iX^ / :i^ / :0: ) : I/O ■ r X mm - ! i i ;V; 1 : 2 ^ 1 N — — MARYLAND WEATHER SERVICE 431 THE WEATHER OF MARCH 4. Year. Max. Min. Mean Character Wind Daily Wind Precip- Temp. Temp. Temp. of Day. Direction. Movement. itation. (De grees Fah r.) (Miles) (Inches) 1871 44 39 41 Pt. cldy N 0.51 1872 44 IS 31 Clear SW & NW 106 1873 21 5 13 Pt. cldy NW 362 1874 68 42 55 Pt. cldy NW 188 0.21 1875 35 27 31 Clear NW 197 1876 42 27 34 Pt. cldy SE 138 1877 57 33 45 Pt. cldy NW 137 0.09 1878 53 35 44 Pt. cldy NW 206 1879 45 26 36 Cloudy NE 76 0.02 1880 74 48 62 Pt. cldy W 223 0.01 1881 38 32 35 Pt. cldy w 331 1.13 1882 57 40 48 Clear NW 141 1883 43 28 36 Pt. cldy NW 218 1884 32 18 25 Clear NW 214 1885 53 24 44 Cloudy E 95 1886 42 29 36 Pt. cldy NW 236 1887 36 27 32 Cloudy NE 117 0.45 1888 35 24 30 Clear NW 198 1889 44 36 40 Cloudy NE&NW 309 2.71 1890 48 27 38 Clear NE 133 1891 42 30 36 Pt. cldy NW 206 6.18 1892 55 35 45 Cloudy NW 181 1893 31 24 28 Pt. cldy NW 438 0.18 1894 56 31 44 Clear SE 110 1895 55 36 46 Pt. cldy SW 306 0.02 1896 42 23 32 Clear N 458 1897 47 35 41 Clear E 117 1898 40 35 38 Cloudy N 205 0.13 1899 44 36 40 Cloudy E 171 0.07 1900 53 31 42 Cloudy SW 65 1901 50 35 42 Cloudy NW 128 0.21 1902 43 32 38 Cloudy W 108 0.03 1903 54 35 44 Pt. cldy SE 84 1904 35 24 30 Clear NW 244 1905 47 27 37 Pt. cldy N 250 0.01 1906 54 38 46 Pt. cldy NW 340 1907 39 24 32 Clear NW 185 Means 45 31 38 Pt. cldy 175 0.53 432 THE CLIMATE OF BALTIMORE THE WEATHER OF MAY 1. The closing days of April and the first days of May are among the pleasantest of the year in many respects. The temperature is well above the frost line — the mean temperature for the first of May is 59°. The early morning temperatures are more constant than in March or April, the minimum averaging about 50°. The maximum has gone as high as 85°, but it has generally been below 70°. The winds are light and largely from a southerly or easterly direction. (See Fig. 152.) The rainy days are less frequent than earlier in the season; there are few days in the year with a smaller rainfall probability (see Plate IX). The duration of rainfall is about 7 hours as compared with 9 hours in March and 10 to 12 hours in the winter months. The average amount of rainfall is also quite small compared with that of other days. The chances for fine weather — a moderate temperature and without rain — are better for the 30th of April and the 1st of May than for any period of the year, excepting the first week in September and the middle of October. The weather conditions on the 4th of March and on the first of May represent with a fair degree of accuracy the conditions of the early and late spring respectively. They represent sharp contrasts — the former showing all the characteristics of our variable winter climate, while the latter resembles closely the more uniform and settled conditions of our summers. THE WEATHER OF EASTER SUNDAY. The weather of this day has been prevailingly cloudy to partly cloudy, with a moderate westerly wind. Eain has occurred on 14 of the 37 anniversaries from 1871 to 1907, and was light in amount on all but one occasion. With an average temperature of 52°, the extremes have ranged between 84° in 1887 and 28° in 1874. 434 THE CLIMATE OF BALTIMORE THE WEATHER OF MAY 1. Year. 1871 Max. Min. Temp. Temp. (Degrees Fahr.) 67 58 Mean Temp. 62 Character of Day. Cloudy Wind Direction. E Daily AVind Movement. (Miles) Precip- itation. (Inches) 1872 73 59 66 Cloudy SE 175 1873 59 47 53 Cloudy S 127 0.24 1874 70 47 58 Cloudy W 184 1875 61 47 54 Cloudy E 214 0.01 1876 61 34 48 Clear NW 324 1877 53 44 48 Cloudy NW 109 1878 80 52 66 Pt. cldy SB-SW-W 82 1879 65 45 55 Pt. cldy NW 249 1880 61 38 50 Clear NW 245 1881 67 46 56 Clear SE 175 1882 68 47 58 Clear W&NW 131 1883 60 45 52 Pt. cldy SE 125 1884 74 57 66 Pt. cldy SE&S 155 1885 62 48 55 Cloudy NE 97 0.23 1886 58 46 52 Cloudy NE 163 1887 70 52 61 Pt. cldy SE 110 1888 76 54 65 Pt. cldy NW 151 0.04 1889 54 48 51 Pt. cldy N&SW 100 0.23 1890 87 58 72 Pt. cldy S&NE 144 0.30 1891 80 64 72 Clear W 161 1892 77 49 63 Clear SE&S 293 1893 64 , 49 56 Cloudy E 194 0.02 1894 80 50 65 Clear SW 124 1895 57 54 56 Pt. cldy NE 353 0.12 1896 54 50 52 Cloudy SE 225 1897 69 57 63 Cloudy E 287 0.94 1898 79 50 64 Pt. cldy SE&NW 73 1899 84 58 71 Clear SE 101 1900 77 . 53 65 Pt. cldy W 81 ... 1901 78 51 64 Pt. cldy E 171 1902 77 56 66 Clear NW 187 1903 69 44 56 Clear NW 306 1904 72 49 60 Pt. cldy NW 136 1905 62 48 55 Clear NW 198 1906 75 59 67 Cloudy N 140 ... 1907 Means 63 68 52 50 58 59 Cloudy Pt. cldy N 189 174 0.04 0.22 MARYLAND WEATHER SERVICE 435 THE WEATHER OF EASTER SUNDAYS. Year. Date. Max. Temp. Min. Temp. Mean Temp. Character ot Day. Wind Direc- tion. Daily Pre- Wind cipita- Movement. tion. (Degrees Fahr.) (Miles) (Inches) 1871 April 9 82 73 78 Pt. cldy W 1872 March 31 64 44 54 Cloudy W 183 0.32 1873 April 13 60 42 51 Pt. cldy NW 244 0.01 1874 April 5 42 28 35 Cloudy SE 174 0.10 1875 March 28 54 35 44 Pt. cldy S 137 1876 April 16 65 49 57 Cloudy SW 212 1877 April 1 59 42 50 Cloudy SE 116 0.02 1878 April 21 79 60 70 Clear NW 156 1879 April 13 58 35 46 Pt. cldy SE 157 1880 March 28 59 38 48 Cloudy NW 99 0.06 1881 April 17 64 44 54 Pt. cldy W 227 1882 April 9 56 49 53 Cloudy S 97 0.38 1883 March 25 46 30 38 Cloudy SE 22 1884 April 13 55 47 51 Pt. cldy SW 89 0.15 1885 April 5 60 35 48 Pt. cldy SW 241 1886 April 25 80 54 67 Pt. cldy NE 142 0.01 1887 April 10 84 46 Go Clear NW 102 1888 April 1 58 43 50 Pt. cldy SE 160 1889 April 21 80 60 70 Clear NW 172 1890 April 6 62 36 49 Clear SW 115 1891 March 29 50 38 44 Clear N 201 1892 April 17 58 42 50 Pt. cldy NE 158 1893 April 2 62 47 54 Clear NW 266 1894 March 25 45 39 42 Cloudy N&NW 137 O.Ol 1895 April 14 59 43 51 Clear N 178 1896 April 5 55 33 44 Clear NW 209 1897 April 18 61 41 51 Pt. cldy W 103 1898 April 10 63 47 55 Pt. cldy E 86 0.01 1899 April 2 43 31 O 1 Pt. cldy W 146 1900 April 15 65 40 52 Clear SW 75 1901 April 7 56 46 51 Cloudy W 154 0.02 1902 March 30 62 45 54 Pt. cldy W 118 0.24 1903 April 12 55 48 52 Cloudy E 175 1904 April 3 46 33 40 Cloudy NW 295 1905 April 23 64 45 54 Clear N 172 1906 April 15 67 51 59 Cloudy NW 183 1.07 1907 March 31 56 42 49 Cloudy N 168 0.03 Means 60 43 52 Pt. cldy 157 0.24 436 THE CLIMATE OF BALTIMORE SUMMER WEATHER. As the spring advances, atmospheric movements on a large scale become more sluggish. Well defined cyclones and anti-cyclones are of less fre- quent occurrence and less intense in their development. This is due, doubtless, to the decreasing contrasts in temperature between north and south, and between the oceans and the continents. Attention has already been called to the relatively great differences in temperature between Florida, for instance, and Montana, in the winter months, compared with the differences in the summer months, an average difference of about 75° in January increasing to 100° at times, as compared with about 30° in July. With the northward movement of the sun the whole atmosphere of the northern hemisphere rapidly rises in temperature during the day. At the same time the days become longer, and the nights shorter; the loss of heat during the long winter nights over the continental masses becomes steadily less. With the increasing heat of the summer the mass of the air over the continents becomes specifically lighter than that over the oceans. The general surface circulation of the air between continents and oceans is reversed. In the winter time the general drift at the surface is from continents to oceans, in the summer time from the oceans to the conti- nents. As the winter area of high pressure over the northern-central portion of the JSTorth American continent diminishes in strength, the Atlantic high area increases in extent and intensity. With its center usually over the Azores, it extends westward across the Atlantic Ocean to the South Atlantic states in the summer time. With this change in the distribution of atmospheric pressure from winter to summer there is a change in the prevailing wind direction. Maryland, in common with all of the Middle Atlantic states, has a prevailing west to northwest wind in the" winter months, the air blowing out of the continental high area; in the summer months the prevailing direction is southeast or southwest, coming from the Atlantic high area to the southeast of the Middle Atlantic states. MARYLAND WEATHER SERVICE 437 In the summer season the paths of the centers of cyclones are con- fined mostly to the northern tier of states, the Lake region and the St. Lawrence Valley. Hence marked cyclonic changes in temperature are infrequent in the states farther south. The distinguishing feature of the temperature changes is the diurnal variation, the difference between the early morning and the afternoon readings of the thermometer. In the winter and early spring months the irregular cyclonic changes are far greater than the diurnal change. This conspicuous prominence of the diurnal fluctuation, which is characteristic of tropical climates, is not confined to temperature; it also appears in the wind direction, the rainfall and in local storm frequency. The increasing magnitude of the diurnal period in the summer months at Baltimore has already been discussed in considerable detail in the first part of the report. Further attention will be directed to character- istic weather types of the summer season in the following pages, especially to conditions which give rise to local storms, such as thunderstorms, squalls, tornadoes, and to hot spells. Summer Storms. A glance at Fig. 77 and Fig. 78, on page 217, will at once reveal the existence of an intimate connection between high temperatures and the occurrence of thunderstorms. In our latitudes these turbulent atmos- pheric disturbances become more and more frequent with the northward movement of the sun, increasing steadily from December to July, and then rapidly decreasing to December. The following figures show the annual average frequency for tlie vicinity of Baltimore : THUNDERSTORMS RECORDED AT BALTIMORE (187G-1904). Jan. Feb. Mar. Apr. Maj-. June. July. Aug. Sept. Oct. Nov. Dec. Year. 13 11 20 33 107 156 17D 111 43 7 R 2 678 Over 80 per cent of the total annual number of these storms occur dur- ing the hot season — May to August — and including September and April, the total frequency for the summer half year amounts to 92 per cent, leaving but 8 per cent for the winter half year. The same intimate connection with change in temperature is shown in the diurnal period of thunderstorm occurrence. (See page 276.) 438 THE CLIMATE OF BALTIMORE HOURLY FREQUENCY OF THUNDERSTORMS. Hoursonding 1 3 3 4 5 6 7 8 9 10 11 13 A. M 7 1 3 7 3 6 6 8 9 15 P. M 3;.' 46 76 78 61 65 69 38 34 31 16 13 Total 610 More than two-thirds of all thunderstorms recorded at Baltimore in 28 years began during the seven hours from noon to seven p. m. Many of the storms occurring later than 7 p. m. had their origin in the early afternoon hours and were carried eastward several hours before being dissipated. When the thunderstorm does occur in the cold season it is in connec- tion with a relatively warm inflow of air towards the center of a cyclone. During the summer months many thunderstorms occur in the absence of any well defined general cyclonic depression. During a period of abnormally high temperatures due to bright sunshine and a sluggish wind circulation, the lower layers of the atmosphere are excessively heated, resulting in a marked disturbance in the normal rate of decrease of heat with elevation. A brisk vertical circulation is set up, which, in the presence of a high humidity, results in rapid cooling of the ascending air, and formation of cumulus clouds. As this circulation increases in energy the cloud soon developes into a " thunder head " with its accom- paniment of heavy rain, lightning, and thunder. Hence dynamic cool- ing of warm moist convection currents is the chief cause of thunder- storms of the summer season. When these storms occur in connection with a general cyclone the mass of warm moist air which produces the thunder cloud is mechanically forced upward in addition to rising as a convection current. The thunderstorms occurring in connection with a general cyclone are likely to be of wider extent than those due to con- vection currents alone, and the cool air following the storm is apt to be more lasting. The fall in temperature following the local heat thunder- storm is usually of brief duration. THE THUNDERSTORM OF JULY 20, 1902. On July 20, 1902, a thunderstorm of unusual severity passed over Baltimore. Twelve lives were lost, while several hundred houses were un- roofed or otherwise seriously damaged, involving a loss of over $200,000. MARYLAND WEATHER SERVICE 439 The wind attained a velocity seldom equalled in the annals of Baltimore weather. On tlie morning of the 20th the weather chart of the United States Weather Bureau showed an area of moderately low pressure over the Lower Lake region, the Middle Atlantic and Southern New England states, with the center of the depression over Lake Erie and Southern Michigan at 8 a. m. Cloudy and rainy weather prevailed over a wide area about the center of the oval depression. A well defined area of high pressure covered the country between the Mississippi Eiver and the Eocky Mountains. By 8 p. m. the barometric depression had moved eastward with its center over Pennsylvania, the isobars meanwhile becom- ing nearly circular. The weather from the Mississippi Eiver to the Atlantic coast and from the Lake region to the Gulf of Mexico was in an unsettled condition, thunderstorms occurring during the day at nearly every reporting station of the Weather Bureau in the Middle Atlantic, the South Atlantic and Gulf states, and the Lake region, accompanied in most cases by light rains. The chart for 8 p. m. shows the general weather conditions within an hour or two after the occurrence of many of the local storms in the Middle Atlantic states. (See Fig. 153.) The progressive changes in the elements as recorded at Baltimore during the day were particularly interesting and instructive. The baro- meter slowly and steadily fell during the forenoon; the wind blew from the southwest with a velocity of only 4 to 8 miles until 8 a. m., then increased steadily to 15 or 16 miles per hour by 10 a. m. The tempera- ture rose rapidly from a minimum of 72° at 6 a. m. to a maximum of 94° at 1 p. m. The sky was overcast during the night, but there was considerable sunshine from 7 a. m. until 1 p. m., when a sheet of stratus clouds appeared in the west, intensely dark and advancing rapidly towards the zenith. At 1.25 p. m. small torn cumulus clouds passed rapidly southwest to northeast in advance of the cloud of dust. This was followed by a stratus layer which by 1.30 p. m. covered the entire sky. At 1.45 p. m. the stratus clouds, moving southwest to northeast, began to break away. Between the open spaces strato-cunuilus clouds were visible above moving from west to east. (See Fig. 154.) 29 440 THE CLIMATE OF BALTIMORE The first peals of thunder were heard at 1.23 p. m., coming from the southwest. The electrical display was brilliant during the height of the storm. The last thunder was heard at 2.25 p. m. Light rain began at 1.27 p. m., changing to a heavy shower at 1.29 p. m. ; the rain moved in dense sheets from southwest to northeast. In the meantime the wind was increasing in velocity; at 1 p. m. it registered 17 miles per hour, Fig. 153.— The Thunderstorm of July 20, 1902. remaining at this velocity until 1.20 p. m. In the next five minutes the velocity suddenly increased to 46 miles, and at 1.31 p. m. it blew at the excessive rate of 75 miles per hour from the west. Coincident with the sudden increase in the wind velocity the pressure rose nearly a tenth of an inch in the course of a few minutes, while the temperature fell as rapidly from 94° to 69° and the rain fell in torrents for a few minutes. The wind rapidly fell and by 2 p. m. had regained its early morning velocity of 5 to 6 miles per hour. The pressure regained in half an MARYLAND WEATHER SERVICE 441 hour the height at which it stood before the sudden rise, and then continued to fall slowly until about 8 p. m. The temperature rose rapidly after the sudden fall, recording 84° at 3.30 p. m., maintaining MDT. 6 A. NOON 6 P. MDT. Fig. 154.— The Thunderstorm of July 20, 1902. this reading approximately until G p. m., when it fell rapidly to 76° shortly after 8 p. m. An hour or so before the sudden fluctuations in wind velocity, tem- perature, and pressure recorded above, the wind veered from southwest to 443 THE CLIMATE OE BALTIMORE MARYLAND WEATHER SERVICE 443 west. During the afternoon and evening the direction alternated fre- quently between the west and southwest. The storm was of short duration; the interval between the first and last thunder heard being about an hour. Its greatest intensity was reached within 20 minutes after the first peal of thunder was heard. The rainfall began at 1.27 p. m. and ended at 1.55 p. m., with a total precipitation of a little over half an inch. During the morning, cirrus, cirro-stratus, and alto-cumulus clouds were observed passing across the sky from west to east with considerable rapidity. A cloud of dust preceded the thunderstorm, carrying leaves, paper and other light objects high up into the air. Just preceding and during the storm the humidity was fully 40 per cent higher than at the 8 a. m. observation, ^o portion of the city was free from damage caused by the storm, although north Baltimore seemed to have sufi'ered less severely than other sections. The storm described above caused considerable damage in all parts of Maryland, though most of the loss of life and property occurred in the vicinity of Baltimore. A special effort was made at the time to trace the path of the storm across the state. Co-operative observers in Maryland, Virginia, West Virginia, and Delaware were requested to report accurately ihe time of day when the first thunder was heard in their respective localities. Eeplies were received from about 150 observers, making it possible by charting the recorded times upon a map and joining, by a line, localities over which the storm passed at about the same hour, to follow the storm from West Virginia eastward to the Atlantic coast. The accompanying chart shows the hourly rate at which the storm travelled. In Central West Virginia the storm began at 10 a. m. From West Virginia it passed into Maryland by way of Washington county between noon and 1 p. m. It then advanced with an irregular wave front eastward to the Chesapeake Bay by 2 p. m. (See Fig. 155.) The storm front then moved in a southeast direction, passing beyond the limits of Maryland in Worcester County between 6 p. m. and 7 p. m. The total distance traversed by the storm from 10 a. m. to 6 p. m. was about 200 miles, or at the rate of 25 miles per hour. The 444 THE CLIMATE OF BALTIMORE irregular form of the storm front and its var3dng rate of progress across the state are clearh^ shown in the chart (Fig. 153). THE THUNDERSTORM OF JULY 3, 1902. At 8 p. m. of July 3, 1902, a moderate barometric depression was centered over Lake Ontario, while the western edge of an extensive area of high pressure covered the South Atlantic states and extended far out Fig. 156.— The Thunderstorm of July 3, 1902. over the Atlantic Ocean. The depression was attended by moderate rains in New England, the Lake region, and the Middle Atlantic states. In the vicinity of Baltimore, as shown by official records, the day was partly cloudy and oppressive; a film of cirro-stratus covered the sky from early morning, through which the sun shone with great intensity. At 4 p. m. stratus clouds were observed moving rapidly from the north and northeast. During the morning and early afternoon a light to fresh MARYLAND WEATHER SERVICE 445 breeze blew from the southwest; at 3.15 p. m. the direction of the wind changed to west, and its force increased to brisk for a brief time. B}- 4 p. m. the velocity of the wind began to increase rapidW, the clouds MDT, 6 A. NOON 6 P. MDT Fig. 1.37.— The Thunderstorm of July 3, 1902. maintaining their original direction. At 4.25 p. m. the wind shifted to the northwest, blowing with increasing velocity, and attaining a maxi- mum rate of 34 miles per hour at 4.29 p. m. At the same time great numbers of cumulus clouds were rapidly carried across the sky from the 446 THE CLIMATE OF BALTIMORE north-northwest. The storm front moved with great rapidity to the southeast, the usual dust cloud marking the advance. The squall-wind car- ried light objects high into the air. A number of lives were lost during the squall, while considerable property was damaged, and many trees were uprooted. Eain began at 4.35 p. m. and continued until 4.50 p. m., the amount being 0.04 inch. On the arrival of the stormfront, marked changes in the barometer and thermometer were noted. (See accompanying diagram.) The barometer fell rapidly throughout the day until shortly after 4 p. m., while the thermometer rose from 66° at 5 a. m. to 96° at 4 p. m. With the sudden change of wind at 4.30 p. m. from southwest to west and northwest, there was an abrupt rise of nearly a tenth of an inch in the barometer. The temperature fell as abruptly from 96° to 74°, while the wind rose from 12 miles to 32 miles per hour. The change was accompanied by a sharp shower of rain of a few minutes' duration. The barometer lost a part of the sudden rise during the following hour and then continued to rise slowly during the balance of the day. The temperature remained low after the storm. No thunder and light- ning were noted in connection with this storm while it passed over Balti- more. Electrical displays were, however, reported from many parts of the state on this da3^ The progressive changes were all characteristic of a well defined thunderstorm. THE THUNDERSTORM OF JULY 12, 1904. There is a type of pressure distribution which invariably gives rise to numerous and severe thunderstorms and squalls. It is represented in the accompanying chart showing the general weather conditions at 8 p. m. of July 12, 1904. It is the V-shaped depression referred to in a preced- ing paragraph in connection with a discussion of storm types. In this instance the " squall line " is particularly well defined by a long narrow band of thunderstorms and rain at or near 8 p. m., extending from the St. Lawrence Valley southward through New York, New Jersey, Eastern Pennsylvania, Eastern Maryland, Delaware, and along the coast south- ward to Florida. (See Fig. 158.) MARYLAND WEATHER SERVICE 447 Fig. 158.— The Thunderstorm of July 12, 1904. THE TORNADO OF JULY 12, 1903. The Middle Atlantic states are rarely visited by tornadoes. There are descriptions of such storms on record in the annals of Baltimore weather, but the storms were of a mild type of tornado so far as can be judged by local descriptions. Tornadoes occur under general conditions similar to those which give rise to thunderstorms and squalls. They differ, however, from the latter in the character of the atmospheric circulation within the storm, in their greater destructiveness, and in the fact that tliey are more restricted in the area of their activity. The air within a thunderstorm moves about a horizontal axis, while within a tornado the circulation is about a vertical axis. The thunderstorm moves eastward with the general cyclone with a long wave front many miles in length, the tornado moves along with the general storm in the form of a vertical column of limited extent, generally less than half a mile in diameter. 448 THE CLIMATE OF BALTIMORE HIGH Fig. 159.— The Tornado of July 12, 1903 (8 a. m.). Pig. 160.— The Tornado of July 12. 1903 (8 p. m.). MARYLAND WEATHER SERVICE 449 usually recognizable as a downward extension of the cloud mass which generally reaches to the ground, but sometimes dangles in mid-air like the loose end of a suspended rope. The storm of July 13, 1903, as it passed over Baltimore, had, from the best information obtainable from eye witnesses, many of the traits of the real tornado, although it is frequently difficult to distinguish between a mild type of tornado and an intensely developed thunderstorm. The general weather conditions were favorable for the production of local storms over a large portion of the Atlantic and Gulf Coast states and the Ohio Valley. Cloudy and unsettled weather prevailed in the sections named at 8 a. m. of the 12th. Thunderstorms were reported from many stations for the preceding twelve hours. There was an area of high pressure in the northwest, and a barometric depression over the Gulf of St. Lawrence, with a secondary depression forming over the Lower Mississippi Valley. The temperature conditions were nearly nor- mal, but the humidity was high. The prevailing wind direction at stations in the Atlantic Coast states was from the southwest and light in force, excepting in the South Atlantic states, where they were fresh to brisk. During the succeeding 24 hours the secondary depression had developed and moved rapidly northeastward over Maryland, the center being over Massachusetts at 8 a. m. of the 13th, accompanied by heavy rains and severe local storms in the South Atlantic states, and near the coast in the Middle Atlantic and New England states. The following heavy rainfalls were reported for the preceding 24 hours at 8 a. m. of the 13th: Baltimore, Md., 3.98 inches; Washington, D. C, 3.02 inches; Atlantic City. N". J., 1.74 inch. (See Figs. 159 and 160.) The records of the local office of the United States Weather Bureau contain the following account of the storm as recorded by the official and other observers: On July 12 thunderstorms and heavy rainfall were general throughout the section. At Baltimore the storm at its height developed destructive features over a limited area. A funnel-shaped cloud, peculiar to the tornado was clearly in evidence a few minutes after noon, and the narrow path pursued by this cloud was also the path of devastation. The cloud moved from west to east, descending to the house-tops at two points within the city, leaving a 450 THE CLIMATE OF BALTIMORE wide gap of comparatively slight loss between; it evidently struck the ground again a short distance beyond the city proper, judging from the local damage there, and then disappeared, as far as surface traces were concerned. Re- ports from parts of Kent County, however, would indicate that the storm crossed the Bay and moved over the Eastern Shore, for in that county a nar- row area was visited by destructive winds. The following is the special report of Mr. James S. Harris, the co-operative observer at Coleman, regarding this visitation: "About 1 p. m. an angry black cloud came suddenly over, causing a darkness as of twilight, accom- panied by a cyclone and hail. Wheat in shock and stack was blown about, trees blown down, and houses wrecked." In his use of the word " cyclone " the writer doubtless intended to describe a tornado, a confusion of terms so frequently met with in popular accounts of storms of this class. Baltimore and Coleman seem to have marked the extreme limits of tornado winds, although the thunderstorms were more or less severe at many other points on the same day. In Baltimore the first area visited embraced much of the 1700 blocks of Fulton Avenue, Mount Street, and Calhoun Street; here a funnel-shaped cloud was distinctly observed by a number of the residents, but no definite account of its manner of formation was obtained beyond this. In the second district, which extended from Eager Street and Broadway eastward for six blocks, with a width varying from two blocks to less than a block, the damage was greater and the information obtained was more explicit. A heavy storm cloud approached from the northwest and another from the southwest; they apparently merged at Eager Street and Broadway, where the destruction abruptly began. The funnel-shaped cloud was seen by many, and a loud roaring sound was followed by almost complete darkness as the storm burst. The upper cloud mass was distinguishable, however, with its narrowing extension downwards, the latter appearing to lag slightly behind the mass above in its movement eastward. The whole travelled with almost incredible velocity, only a few seconds elapsing between the time the cloud descended to the house-tops at Eager Street and Broadway and the time when it rose into the air again six blocks to the eastward. In both districts the nature of the destruction pointed clearly to the fact that the city had been visited by a tornado. In some of the wrecked houses the walls were blown outward, as though by sudden expansion of confined air within, although fully as many fell inward. In one case the four walls had bulged outward, and the roof lay within, about half-way down to the fioor of the second story, while not far off roofs had been lifted high into the air and carried a block and a half away before being deposited in an alley to the rear. In all several hundred houses were unroofed or otherwise badly wrecked. The money loss was estimated at $200,000; happily there was no loss of life, although one man was seriously hurt by falling walls, and nu- merous narrow escapes from injury were reported. At the Weather Bureau Office, about a mile and a half away, no damage occurred. The self-registering instruments, while presenting interesting records, do not adequately portray the conditions as they existed at the MARYLAND WEATHER SERVICE 451 centers of severe damage. The rainfall at the station was unusually heavy; 2.87 inches fell in 33 minutes, from 12.04 p. m. to 12.37 p. m. The following maximum falls were tabulated: Greatest amount ount in 5 minutes, O.SO inc " 10 1.35 ' " 15 1.92 ' " 20 2.22 ' " 25 2.46 ' " 30 2.75 ' " 35 2.87 ' Further details of rainfall in connection with this storm are given on pages 212 and 213. The rate of rainfall in the districts of greatest storm loss must have been much heavier. The streets were running streams of water, and cellars were entirely filled within a few minutes. At the Baltimore station the wind was comparatively high from 12.04 p. m. to 12.15 p. m., and brisk to light thereafter. The maximum velocity was 46 miles per hour at about 12.05 p. m. The wind direction veered through nearly all of the points of the compass during the storm, as shown by the following record: Noon to 12.05 p. m Southwest. 12.05 • ' 12.15 12.15 ' ' 12.20 12.20 ' ' 12.25 12.25 ' ' 12.45 12.45 ' ' 1.00 .West (mostly). .Northwest. .North. .Northeast. .East (mostly). There was a sharp fall of about 15° in temperature at the beginning of the storm, but at the office of the Weather Bureau the variation in atmos- pheric pressure was very slight. The only noteworthy feature of the pressure curve was a small but sudden rise of about 0.05 inch, characteristic of severe thunderstorms accompanied by hail. The general weather conditions on the day of the storm are recorded as follows in the local office of the United States Weather Bureau : Cloudy day. Not so warm. Atmosphere very oppressive in forenoon; pleasant afterwards. Maximum, 85° at 11 a. m.; temperature then fell to 70° at noon, rose to 74° at 4 p. m., fell to 69° at 8 p. m., and remained stationary until midnight. Sky partly covered at dawn, became overcast by 9 a. m. with alto-stratus clouds. At 11 a. m. a dark low-lying cloud mass appeared on the northern and western horizons, moving slowly. Shortly before noon, the movement of the cloud mass increased very rapidly, and the sky became covered in a few minutes, continuing so the rest of the day. This movement of the clouds was followed by a terrific thunderstorm, thunder being first heard at 11.48 a. m., continuing all the afternoon at intervals, being last heard at 6.45 p. m., becoming recognizable as a second storm at 452 THE CLIMATE OF BALTIMORE about 6 p. m. The first storm moved from west to east, the second passed from south to north. A trace of rain fell in the early morning. The periods of rainfall during the day were as follows: 12 noon to 1.10 p. m. 1.45 p. m. to 2.10 p. m. 2.50 p. m. to 3.20 p. m. 4.45 p. m. to 5.05 p. m. 5.40 p. m. to 7.50 p. m. 9.25 p.m. to 9.40 p.m. The rainfall was excessive from 12.07 p. m. to 12.42 p. m. (2.87 inches), and heavy from 6.20 p. m. to 6.40 p. m. (0.72 inch) ; the total amount for the day was 3.90 inches. A light southeast wind before noon shifted suddenly to west at noon with increased force, being brisk to high from 12.02 p. m. until 12.27 p. m., with a maximum velocity, at the station, of 46 miles from the west at 12.07 p. m. The winds were light and variable the rest of the day, mostly from the north. WATERSPOUTS. Waterspouts are in their mode of formation and in their external appearance similar to tornadoes. In extent, however, they are much more restricted, while they do not compare with the tornado in destructive power. They are of comparatively infrequent occurrence and it is not often that they are observed at close range by an intelligent observer, hence the following description is of special interest: Early in April, 1902, Captain Fergus Ferguson of the British S. S. Hestia left Baltimore for one of the Cuban ports. On April 4, towards sunset, while off Hatteras, the Captain observed several waterspouts in process of forma- tion at a distance of 300 to 400 yards to windward. The largest of these, and the only one completely formed, seemed to be headed directly for the Hestia. The Captain at first attempted to change his course sufficient to avoid running into it, but soon discovered that this could not be done. Giving orders for all on deck to go below, he remained until the spout was close upon his ship, and then hastily sought a place of safety. In a moment he heard a deafening roar which was quickly followed by strong gusts of wind and a sudden shock as the spout struck amidships and passed over the deck towards the stern. The Captain reappeared upon deck in time to see two tarpaulins, which had covered the hatches, and a plank 8 feet long by 10 inches wide, high in the air, while his log line with log attached extended straight up into the air to a distance of about 40 feet. Beyond the loss of the lighter movable objects on deck and a temporary feeling of apprehension no harm was done. MARYLAND WEATHER SERVICE 453 When first seen, the waterspout was incomplete. A portion of the cloud dipped down from the general cloud level of about 2000 feet, while at the same time a column of water was apparently rising from the surface of the ocean just below. At an elevation of between 200 and 300 feet the ascending water column and the descending cloud column met. The diameter of the spout was approximately the width of the Hestia, or between 40 and 50 feet. Within the column there was a dark core, almost black, with a diameter of about 2 feet. Captain Ferguson did not clearly recall evidences of a whirling motion, but a strong upward movement is clearly indicated by the facts noted above. No reference was made to any considerable quantity of water being shipped as the waterspout passed over the vessel, a fact which would indicate that the lower portion of the column was composed mostly of spray carried up by the strong wind from the surface of the ocean. At the time of occtirrence of the waterspout the Hestia was near the center of a shallow but well defined barometric depression just off the coast of North Carolina. The general storm was moving slowly up the coast. A ridge of high pressure extended from the St. Lawrence Valley southwestward to the West Gulf states. The winds along the coast from New England to North Carolina were northerly. Summer Hot Spells. One of the most characteristic features of our summer season is the frequent recurrence of a longer or shorter series of excessively warm days. No summer season is entirely free from them, although at times they are not frequent enough or intense enough to cause comment. These periods vary greatly in length and in the frequency of their occurrence. When the atmosphere is comparatively dry, high temperatures may be endured without great personal discomfort. A high humidity combined with even moderately high temperatures is the cause of most of the unfavorable comment upon the summer weather of the Middle Atlantic coast states. In the usual course of summer events cyclones and anti-cyclones, though not as frequent as in winter or spring, and not so intense, are 3'et sufficiently frequent in their passage across the northern portion of the country to maintain a fairly well mixed atmosphere and thus prevent the accumulation of excessively heated air near the surface of the earth. At times, however, we have a comparatively stationary system of cyclones 454 THE CLIMATE OF BALTIMORE and anti-cyclones with a small gradient, which may persist with very little change in position for many days. Such a system is of frequent occurrence in the summer season in the United States. An area of high pressure settles over the South Atlantic states, or over the Atlantic Ocean with an extension covering the South Atlantic states, while a barometric depression rests over the Missouri Valley or the eastern slope of the Rocky Mountains. While this distribution of pressure continues there is a steady flow of warm dry southeast to southwest winds over the Middle Atlantic states. If in addition the gradient is small, or the South Atlantic high area moves over the Middle Atlantic states, the winds become very light while the clear skies permit uninterrupted sunshine. The sluggish movement of the atmosphere together with the unobstructed insolation permits the accumulation of excessively heated layers of atmos- phere at the surface of the earth. Sometimes these conditions will per- sist for two or three weeks before the cyclonic and anti-cyclonic systems begin to move eastward in their accumstomed paths and bring about a change. THE SUMMER OF 1900. The summer of 1900 was probably the warmest in the annals of Middle Atlantic states weather. The temperatures at the beginning of the season were about normal, June averaging but 0.2° per day below the average of 30 years. Beginning with July the average monthly tepera- tures remained far above their normal seasonal values until the close of November, the departures from the normal increasing steadily from July to September and then decreasing slowly to November. July -f2.0° August +4.7° September -}-5.1° October +4.5° November +3.5° July, 1900. During the first two days in July northerly winds pre- vailed in Maryland, accompanied by a cool morning temperature of about 60° in the central and eastern portions of the state. On the Allegany plateau the night temperatures were as low as 40°. The maximum' after- MARYLAND WEATHER SERVICE 455 noon temperatures were about 85°. On the whole, these days were several degrees cooler than the normal for the season. On the 3d the temperature began to rise rapidly. At Baltimore the maximum was 92°, and, with the exception of four or five days during which the maximum registered in the eighties, the afternoon temperatures remained well above 90° until the 21st of the month. From the 22d to the 31st the maximum readings ranged between 80° and 91°. The hot spell culminated in temperatures of 100° on the 16th and 17th. These temperatures, occurring at Balti- more, fairly represent the conditions that prevailed in the central, eastern, and southern portions of the state. In the valleys of Washington and Allegany counties the figures are somewhat higher. Thus, at Hagers- town, a reading of 105° was recorded on the 16th; at Hancock, 105° on the 15th, 16th, and 17th; at Green Spring Furnace, 106° on the 17th, the highest in the state. Within a very restricted area Mar3dand offers a great variety of climatic conditions. On the Allegany plateau, in Garrett Count}^ the thermometer did not register above 92° during the entire month, and then only on one or two days. The temperatures here indicated are all shade temperatures, that is, they were registered by thermometers placed in standard shelters which protect the instruments from the direct rays of the sun, or reflected rays from neighboring objects, but are so constructed as to permit of free circu- lation of the air. Thermometers exposed to the direct rays of the sun at Chase, in Baltimore County, and at Chewsville, in Washington County, gave an average maximum of 104° on 13 days, with an absolute maximum of 110°. Such temperatures, are, however, not unusual with thermome- ters so exposed. The average number of days with a maximum tempera- ture of 90° or above in July at Baltimore, based on the 30 years of care- fully kept records of the U. S. Weather Bureau, is 9 days. Their fre- quency has varied from a total absence in 1891, to 18 in 1876. During July, 1900, there were 15 such days at Baltimore, 17 at Washington, 18 at Hagerstown, 19 at Laurel, 21 at Taneytown, and 27 at Hancock. Frostburg had but 5, Grantsville and Deer Park 2 each, while at Sunn)^- side, Garrett County, there was but one day. The average daily maximum temperature at Baltimore during these 15 days was 95°; the normal 30 456 THE CLIMATE OF BALTniORE average for the same period is 86°, showing a daily excess of 9°. These excessive temperatures caused the average daily temperatures for the entire month in Maryland and Delaware to be 2.5° to 8° above the normal value for the season. The weather conditions whicli usually accompany hot spells were pres- ent in a marked degree during July, 1900. The skies were remarkably clear ; the winds were prevailingly southwest, and generally light in force ; the rainfall was deficient in quantity and frequency. The records from over 50 stations in Maryland, Delaware, and the District of Columbia show an average of 17 clear, 11 partly cloudy, and 3 cloudy days. The average conditions at Baltimore, derived from 30 years of observations, are 10 clear, 13 partly cloudy, and 8 cloudy days. The winds were almost constantly from the south or southwest while the high temperatures pre- vailed. At Baltimore they were from the southeast, south, or southwest during 20 days out of the 31. The average hourly velocity was but 4.6 miles, approximately the lowest in 25 years, during July, while the high- est velocity for the month was only 18 miles, the smallest maximum recorded at Baltimore. Scattered showers fell from the 3d to the 9th; on the 12th and 30th rainfall was general throughout the states of Mary- land and Delaware; during the period from the 17th to 26th local showers were frequent. With but few exceptions the total rainfall for the month Avas decidedly below the average. Baltimore had but 1.31 inch and Washington, D. C, but 1.25 inch, whereas the average rainfall for July in this vicinity is about 4.50 inches. The relative humidity during the period of intense heat was somewhat below the average for the month, a circumstance affording some cause for thankfulness. While suffering the discomforts of an intense spell of warm weather, we are apt to overestimate its severity as compared with those experienced in the past. Statistics, however, support the assertion that this July hot spell was one of the most trying on record in our vicinity. It is always difficult to make just comparisons in dealing with weather conditions. We feel hot and uncomfortable and look for the cause in high tempera- tures alone, but do not always find them as high as expected. The ele- ment of personal discomfort is due to certain combinations of tempera- MARYLAND WEATHER SERVICE 457 ture, humidity, and air movement, and we have no single set of values to express this element, ^^'e can and do measure accurately the tempera- ture, the humidity, and the wind direction and velocity, each separately. Upon these figures we must base our judgment of the severity of any disa- greeable period of weather. Since 1871, the date of the establishment of the Weather station at Baltimore, the number of days in July with a max- imum temperature of 90° or above has exceeded 15 but twice. In 1878 there were 16 such days with an absolute maximum of 98° ; the average of the maximum temperatures was 92.5° as compared with 95° in 1900. The average relative humidity was the same in both instances, namely 63 per cent. The average daily wind movement was greater in 1878 than in 1900, 128 miles in the former and 117 miles in the latter period. In 1876 there were 18 consecutive days with an average maximum of 93°, and an absolute maximum of 99°; the average relative humidity during this period was 63 per cent; while the average wind movement was 125 miles per day. As a result of this comparison with the two most con- spicuous rivals for notoriety, we find that the hot spell of July, 1900, was but little shorter in duration ; that the humidity was as high ; that the average temperature was fully 2° higher; and that the wind velocity, a powerful element of relief on a muggy da}^, was less. August, 1900. According to statistics of the Baltimore Health Officer there were 30 deaths during August due directly to sunstroke, and 32 in addition due to excessive heat as a secondary cause. When we come to examine the record of weather conditions during this period, and compare it with the hot spells of the past, we find nothing to equal it in intensity since the establishment of the Weather Bureau Station in Baltimore in 1871. Baltimore has on an average five days in August with a temperature of 90° or above, with a maximum in the past of 98°. In August, 1900, there were 17 such days, with a maximum of 100°, while this maximum was practically maintained for six consecutive days. Temperatures were even higher, and hot days more frequent at other points in Maryland and Delaware. Thus, in Washington County there were 20 days with a maximum temperature of 90° or above, with an absolute maximum of 458 THE CLIMATE OF BALTIMORE 103° at Hancock. The highest temperature recorded within the two states was 10-i° at Millsboro, Delaware, on the 14th. The hot wave began on the 6th, with a maximum temperature at Baltimore of 97°; from the 7th to the 12th inclusive the afternoon heat reached 99° or 100° each day; from the 13th to the 19th the daily maxi- mum ranged between 90° and 94°. Fortunately the relative humidity was comparatively low, averaging but 65 per cent, the normal value being 70 per cent. A comparatively cool period of four days followed, with heavy showers. The temperature rose again on the 24th to 87°, and ranged between 88° and 96° to the close of the month. While the tem- perature averaged 6° less daily during the latter period than from the 6t]i to the 19th, the relative humidity rose from 65 per cent to 81 per cent. To add to the discomfort of heat and humidity, the air movement was extremely light. The total wind movement over Baltimore during the month averaged but 108 miles per day; this is equivalent to an aver- age of 4.5 miles per hour. Such conditions following closely upon the long-continued hot weather of July and the first half of AugTist brought intense suffering to man and beast. Comparing the hot period of this month with earlier notable hot spells since 1871 we have the following: Length of Period, Averag-e ^^aximum. August, 1872 12 days 93° August, 1888 10 days 92° August, 1896 10 days 94° August, 1900 17 days 95° A particularly uncomfortable feature of the hot spell was ihe high night temperature. During four successive nights the minimum tem- perature ranged from 80° to 82°. At no other time in the preceding 30 years has the night minimum exceeded 78°. The normal temperature for the month of August at Baltimore is 75°. During August, 1900, the mean temperature was 80°; this value was equalled but once, namely, in 1872. The abnormally warm weather of August was not confined to narrow limits. During the first week the temperature was above normal from the Eocky Mountains eastward to the Lower Lakes and the Appalachian MAKYLAXD WEATHER SERVICE 459 Mountains. In South Dakota the daily excess was 13° above the normal value. During the second week the warm area extended eastward to the Atlantic coast, and the areas of maximum excess were transferred east- ward to Michigan and to the region including Philadelphia, Baltimore, and Washington, D. C. The temperature continued abnormally high dur- ing the third and fourth weeks, but the maximum daily excess fell from 12° to 9°. The high temperatures have frequently been attributed in the daily press to a greater solar activity as shown by the increasing number of spots upon the sun's disk. A less remote and more plausible explanation may be found in the unusual distribution of atmospheric pressure during the hot spell. There is a type of pressure distribution which always brings warm weather to the Middle Atlantic states. When the barometer is high over the South Atlantic states, or just off the coast, while it is relatively low over an extensive area to westward and northward, the winds over the Middle Atlantic states are generally from a southerly direction, and light in force, while the skies are clear. Near the center of high pressure, moreover, the air descends from higher levels and is warmed by compression in descending. These conditions, all favorable to the production of high temperatures, were present in a marked degree during the period of hot weather in July and August. Clear skies favored the rapid warming up of the surface of the earth and the adjacent layers of air during the day; and the frequent calms and the prevailing light winds — the average for the entire period of the hot spell being but 4.5 miles per hour at the Baltimore station — prevented the rapid exchange of temperatures between adjacent regions, or between upper and lower layers of the atmosphere. As a result the air near the surface of the earth was excessively heated. At the high level stations of Western Maryland the temperatures were comparatively moderate. The maxi- mum for the month of August was but 89° at Deer Park, and 91° at Orantsville. General Weather Conditions During the Hot Spell of 1900. The distribution of pressure and general weather conditions at the beginning of the August hot spell are shown in the cliart for August 460 THE CLIMATE OF BALTIMORE 6, 1900. An extensive area of high pressure which had drifted slowly across the Lake region moved southward, the center being over the Middle Atlantic states on the 5th. The center of the high area remained for nearly two weeks in approximately the same position. Clear skies and light southerly winds were the prevailing conditions in the Middle Atlantic states. Occasionallv the center of the high area would be a Fig. 161.— Chart of August 6, 1900 (during Hot Spell). but the atmosphere was drawn from the same source to the south, and little further to the southwest, causing a northwest wind at Baltimore, change in local direction would bring about no change in temperature. Over the Eastern Eocky Mountain slope the pressure remained compara- tively low throughout the heated term. On the 12th of August a trough of low pressure developed between the southern high area and an area of high pressure over the Canadian Provinces, causing cloudiness and thunderstorms in the Lake region; this condition developed a MAllTLAXD WEATHER SERVICE 461 depression over the Lower Lake region on the 13th, attended by showers and thunderstorms as far south as Maryland and Xorthern Virginia. But the relief brought about by these showers was only temporary. Local showers in connection with thunderstorms also afforded some relief in the Middle Atlantic states, the Ohio and the Missouri valleys on the 16th, but the temperatures soon regained their intensity. On the 19th an area of high pressure developed to the north of the Lake region while the South Atlantic states high area drifted to the southwest and gradually dissipated. In the meantime a trough of low pressure developed between the two high areas in the Middle Atlantic states, bringing clouds and rain and breaking up the general conditions of pressure which caused the hot spell. The high temperatures of a hot spell are generally first experienced in the Missouri and Mississippi valleys; the distribution of pressure as shown in the chart for August 5 indicates very favorable conditions for a strong drift of warm southerly winds into these valleys. The area of excessive temperatures then moves eastward toward the Atlantic sea- board. This is clearly shown in the accompanying charts which outline the areas over which the temperatures were in excess or deficiency of their normal values for each week from July 23, 1900, to September 2-1, 1900. (See Plate XXIII.) Beginning with the week ending July 30, we find the line of no departure from the normal temperature for the week passing through Baltimore, and that over practically the entire central portion of the country the temperatures were from 1° to 3° below their normal values. In tlie accompanying charts, while the line of zero change again passes through Baltimore, the " hot wave "' had already been well established in the Upper Missouri Valley where the daily average temperatures were 9° to 12° above their seasonal values. By the close of the following week the area of greatest excess of temperature above the normal was transferred to ^Maryland, Pennsylvania, and Vir- ginia with a departure of 12^ per day above the normal for the season, while the temperatures had somewhat abated in Missouri and Mississippi valleys. The following charts show that the unseasonable temperatures continued without interruption over practically all of the country east of 462 THE CLIMATE OF BALTIMORE the Kocky Mountains until the middle of September, the week ending September 24 showing the first appearance of temperatures below the seasonal averages in the northern half of the country, while they still continued high south of the Ohio Eiver. It is interesting to note in these charts that the area of the hot wave embraced practically all of the country east of the Eocky Mountains, and that west of the mountain range the temperatures were below their seasonal averages. This is generally true of our hot waves, the Kocky Mountains forming a natural boundary between areas of excessive and deficient temperature. THE SUMMER OF 1901. During the latter part of June and the first week of July, 1901, a heated term of even greater intensity than that of August of the pre- ceding year occurred, although it was fortunately of shorter duration. Afternoon temperatures exceeding 90° at Baltimore began on June 26 with an area of high pressure centered over the Middle Atlantic state? and a barometric depression over the Upper Missouri Valley. The temperature rose steadily until the first of July, reaching a maximum of 103° on the 1st and 2d; from the 3d there was a steady fall in the mean temperature of the day to a normal condition on the 7th, when a thunderstorm accompanied by heavy rain brought on an abrupt fall of 30° in the temperature between 4 o'clock and 4.15 p. m. The excessive heat began about 10 days earlier in the Central West. During the weeic ending June 17 the temperature rose to 6° above the normal seasonal value in the Middle Mississippi Valley. In the follow- ing week the area of excessive heat embraced all the district between the Eocky Mountains and the Alleghanys, with a maximum departure from the normal still in the Middle Mississippi Valley. By July 1 the area had extended to the Atlantic coast, while the heat was steadily increasing in intensity in the Mississippi Valley and the Lake region to a daily maxi- mum excess of 12° above the normal temperatures. During the follow- ing week the center of the heated area was transferred eastward to the Middle Atlantic states with a maximum dailv excess of 12° within the VOLUME 2, PLATE XXIII. *»-« • / Fig. 4.— Week ending August 20, 1900. r' Fig. 8.— Week ending September 17, 1900. i FIGURES 1-9 SHOW TEMPERATURE DEPARTURES DURING THE HOT SPELL OF 1900. Black lines show temperature departures below normal during hot spell of IQOO. Red lines show temperature departures above normal during the hot spell of 1900. )0. I MARYLAND WEATHER SERVICE. VOLUME 2, PLATE XXIII. Fig. 1.— Week ending July 30, 1900. Fio. 2. — Week ending August 6, 1900. FiQ. 3. — Week ending August 13, 1900. Fio. 4.— Week ending August 20, 1900. FiQ. B.— Week ending August 27, 1900. Fio. G. — Week ending September 3, 1900. Fio. 7. — Week ending September 10, 1900. Fio. 8. — Week ending September 17, 1900. FIGURES 1-9 SHOW TEMPERATURE DEPARTURES DURING THE HOT SPELL OF 1900. Black lines sliow temperature departures below normal during hot spell of 1000. Red lines show temperature departures above normal during the hot spell of 1900. Fio. 9.— Week ending September 24, 1900. Fm. in. — Maximum temperatures of .luly, 1901. Fro. 11. — Maximum temperatures of August, 1900. MAEYLAND WEATHER SERVICE 463 area embracing Baltimore, Philadelphia, and New York, while the heat had somewhat moderated in the Middle Mississippi A'alle}-. In the 2d week in July the heat was again on the increase in the Middle Mississippi Valley, with moderating temperatures in the Middle Atlantic states and the Ohio Valley. The New England states experienced but little of the excessive heat of this period. In the Middle West during the second week of July maximum temperatures of 102° to 104° were of frequent occur- rence, establishing new records for excessive heat in a number of localities. In Baltimore the hot spell continued about 10 days, while the highest tem- perature, 103° on the 1st and 2d of July, was within 1° of the highest ever recorded at Baltimore. THE HOT PERIODS OF AUGUST^ 1900, AND JULY, 1901, CO.MPARED. The two periods of excessive heat described above were the most intense noted in the official records of the Weather Bureau since the establishment of the Baltimore office of the National Bureau. While there were many characteristics in common, the two periods showed a marked difference in their effects upon the residents of Baltimore. The death rate is always increased during a well marked hot spell in the large cities of the country. It is a difficult matter, however, to determine the immediate cause of the increased rate. It can not in general be attributed alone to increase in temperature of these hot spells, though this is probably the dominant factor. The humidity doubtless plays an important part in increasing the number of deaths. Perhaps, also, the weather condi- tions of the preceding weeks must be taken into account. The hot spell of August, 1900, covered a slightly longer period than that of June- July, 1901; the temperatures also averaged somewhat higher-; the wind movement was approximately the same. There was an astonishing dif- ference, however, in the number of deaths reported by the Baltimore Health Department as due directly to heat. During the 1900 period there were 32 deaths due to heat prostration; in the June and July period of 1901 there were twice as many — namely, 64. The only marked difl'^rence between weather conditions noted was the difference in humidity — in 1901 the average daily relative humidity 464 THE CLIMATE OF BALTIMORE was 66 per cent of saturation, while in 1900 it was 57 per cent. The August, 1900, period was preceded by excessive temperatures in May and June, though of short duration, and by an exceptionally long and 1 AUG 1900 5 }0 15 20 JUNE JULY 1 901 25 30 5 lO o' / - \ / / > \ / / \ 1 1 1 M AXI MUM ^^ 1 X. \ J \ M V,X mum; / -^ / \ h S*"*" 1 / \ a-. \ /- \ / \ M IN Ml JM / ^ villi \ MINIMUM \ 1 /r\ \ \ / \ V- y \ / \ N \ 1 n° ^ / \ / N ORMA L 57° \ 1 1 1 1 NORMAL. 67- \ V ^ 1 \ X. ■^ / \ 6 o° ^ V'' ^^~^ // / / \ ^=r-^ // ^ f«-%J " ■ — -^ / ^ ^ -\ ^ Fig. 1G2.— Temperature during Hot Spells of 1900 and 1901. intense, though interrupted, spell in July, which may have increased the powers of resistance and enabled the residents of Baltimore to with- stand the debilitating effects of an additional period after the interval of two weeks of moderate summer weather between the 2 2d of Jidy MARYLAND WEATHER SERVICE 465 and 6th of August. In the case of the hot period of June 2 6- July 7, 1901, there were practically no excessively hot days in the months of May and the first three weeks of June; the city was overwhelmed with- out previous preparation, by one of the most intense heated terms exper- ienced in Baltimore. A comparison of the chief climatic features of the two periods by means of statistics and diagrams will enable the reader to understand more fully the points of difference and similarity. THE HOT PERIOD OP AUGUST, 1900. Date Aug. Max. Temp. Min. Temp. Relative Humid- ity. Wind Direc- tion. Hourly Wind V^elocity. Cloud- iness. Rain- fall. Thun- der Storms (Degrees Fahr.) (Per cent.) (Miles.) * 6 97 67 64 SW 2.9 Clear 7 100 76 58 W 4.9 Clear 8 99 . 80 52 NW 5.5 Pt. cldy 9 100 81 54 W 4.0 Pt. cldy 10 100 80 50 Var. 5.1 Clear 11 100 82 44 W 4.7 Clear 12 99 73 70 'w 6.1 Clear 0.14 * 13 92 72 63 SW 3.7 Pt. cldy 0.03 14 94 76 60 SE 4.0 Pt. cldy 15 91 76 72 S 4.3 Pt. cldy T * 16 92 71 74 w 4.0 Pt. cldy 0.36 * 17 91 75 73 N 4.5 Pt. cldy 18 92 72 75 S 3.2 Pt. cldy Total ige 95.0 75.5 62 w 4.4 Pt. cldy 0.53 THE HOT PERIOD OF JUNE-JULY, 1001. June 26 92 70 58 SW 3.7 Pt. cldy " 27 92 73 60 SW 5.5 Pt. cldy " 28 93 73 70 SE 4.0 Pt. cldy " 29 96 74 68 SW 5.2 Clear " 30 99 77 58 W 3.1 Clear July 1 103 80 60 SW 3.6 Pt. cldy 2 103 80 65 Var. 4.5 Pt. cldy " 3 97 74 60 SW 4.9 Pt. cldy T 4 96 77 66 w 4.3 Pt. cldy T 5 94 76 61 SW 5.9 Pt. cldy 6 96 69 76 W 6.0 Pt. cldy 0.65 7 90 66 83 SW 5.8 Pt. cldy 0.50 Total Average 95.9 74.1 65 SW 4.7 Pt. cldy 1.15 466 THE CLIMATE OF BALTIMORE THE ANNUAL DISTRIBUTION OF DAYS WITH A MAXIMUM TEMPERATURE Tear. Apr. May. OF 90° OR ABOVE. (Baltimore, Md., 1871-1907.) June. July. Aug. Sept. Oct. Annual. Absolute Maximum. 1871 2 5 2 9 92, July 16. 1872 5 10 12 2 .. 29 97, July 2. 1873 2 15 4 2 .. 23 96, July 3. 1874 , . 9 7 3 1 20 98, June 9. 1875 7 7 2 16 97, June 27. 1876 5 18 2 .. 25 99, July 9. 1877 2 4 8 3 17 95, June 26. 1878 1 16 4 21 98, July 18. 1879 1 4 10 5 .. 20 99, July 16. 1880 7 9 10 2 3 .. 31 99, July 13. 1881 3 3 11 8 6 31 101, Sept. 7. 1882 6 8 1 15 97, June 25. 1883 1 7 2 10 96, July 22. 1884 4 3 3 3 .. 13 95, July 24. 1885 4 15 3 22 99, July 21. 1886 1887 2 4 10 4 3 2 10 15 92, 102, rJuly 7, jAug. 27 July 18. 1888 1 8 5 10 24 96, Aug. 16. 1889 2 5 1 10 93, July 9. 1890 4 8 2 14 98, July 8, 1891 5 5 1 11 94, ^ June 16, ) Aug. 10, 11 1892 6 10 o 19 99, July 26. 1893 5 9 2 16 98, June 20. 1894 8 11 2 2 23 98, June 24. 1895 2 5 5 10 7 29 97, June 1, 3. 189G 2 5 3 10 10 3 33 98, Aug. 7. 1897 2 4 1 4 1 12 97, Sept. 11. 1898 1 7 10 9 8 35 104, July 3. 1899 1 8 8 8 2 27 98, June 6. 1900 o 3 15 17 4 . . 42 100, (July 16, 17 / Aug. 7, 9, 1 1901 6 18 1 1 26 103 July 1, 2. 1902 1 4 10 1 1 17 99 July 18. 1903 1 1 10 2 14 94 Aug. 25. 1904 4 6 10 97 July 19. 1905 4 8 1 13 98 July 18. 1906 2 5 3 4 2 16 96 Aug. 6. 1907 1 6 3 1 11 93 July 8, 11. Totals 4 31 158 325 153 57 1 729 Average . . 1 4 9 4 2 20 makyland weather service 467 The Cold Summer of 1816. There are numerous records in local annals showing that the summer of 1816 was phenomenally cold — in fact the coldest of which we have any authentic records. Systematic instrumental observations did not begin in Baltimore until the year 1817 (see pages 91-95) ; however, it is not a difficult matter to reduce reliable records of a neighboring station to contemporary conditions in Baltimore. We fortunately have a very complete and trustworthy series of daily records for Philadelphia, which go back to the year 1790. In the main, weather changes in Philadelphia and Baltimore are synchronous, and similar in kind; there is, however, a uniform difference of 1° to 2° in the average monthly temperatures of the two stations due to difference in latitude. By adding this difference to the average monthly Philadelphia temperatures we obtain a reliable value for con- temporary Baltimore temperatures. An interesting little book published in Philadelphia in 1847 by Mr. Charles Peirce and entitled " A meteorological account of the weather in Philadelphia from 1790 to 1847,^' contains a valuable record of. the general weather conditions for each month during this period of 57 years. The following extracts are made concerning the character of the weather conditions in 1816, with special reference to the three summer months of June, July, and August : The Year. The temperature of the whole year was only 49°; it being the coldest year we have on record. Although there was no uncommonly cold weather during the three winter months, yet there was ice during every month in the year, not excepting June, July, and August. There was scarcely a vegetable came to perfection north and east of the Potomac. The cold weather during summer, not only extended through America, but throughout Europe. It was also the coldest summer ever known in the West Indies and in Africa. June, 181G. The medium temperature of the month was only 64°, and it was the coldest month of June we ever remember; there were not only severe frosts on several mornings, but on one morning there was said to be ice. Every green herb was killed, and vegetables of every description very much injured. All kinds of fruit had been previously destroyed, as not a month had passed without producing ice. From 6 to 10 inches of snow fell in 468 THE CLIMATE OF BALTIMORE various parts of Vermont; 3 inches in the interior of New York; and several inches in the interior of New Hampshire and Maine. July, 1816. The medium or average temperature of this month was only 68°, and it was a month of melancholy forebodings, as during every previous month since the year commenced, there were not only heavy frosts, but ice, so that very few vegetables came to perfection. It seemed as if the sun had lost his warm and cheering influences. One frosty night was succeeded by another, and thin ice formed in many exposed situations in the country. On the morning of the 5th there was ice as thick as window glass in Pennsyl- vania, New York, and through New England. Indian corn was chilled and withered, and the grass was so much killed by repeated frosts, that grazing cattle would scarcely eat it. Northerly winds prevailed a great part of the month; and when the wind changed to the west, and produced a pleasant day, it was a subject of congratulation by all. Very little rain fell during the month. August, 1816. The medium temperature of this month was only 66°, and such a cheerless, desponding, melancholy summer month, the oldest inhabi- tant never, perhaps, experienced. This poor month entered upon its duties so perfectly chilled, as to be unable to raise a warm, foggy morning, or cheerful sunny day. It commenced with a cold northeast rain storm, and when it cleared the atmosphere was so chilled as to produce ice in many places half an inch thick. It froze the Indian corn, which was in the milk, so hard, that it rotted up on the stock, and farmers mowed it down and dried it for cattle fodder. Every green thing was destroyed, not only in this country, but in Europe. Newspapers received from England said: "It will be remembered by the present generation, that the year 1816 was a year in which there was no summer." Indian corn, raised in Pennsylvania in 1815, sold (for seed to plant in the spring of 1817) for four dollars per bushel in many places. The departures of the year 1816 from the normal summer temperature are compared with the departures for some of the coolest summers on record in Baltimore in the following table : DEPARTURES FROM THE NORMAL TEMPERATURE. June. July. August. Season 1816 —8° —8° —8° —8.0° 1836 —5° —3° —5° —4.3° 1846 —4° —3° —1° —2.7° 1886 —3° —3° —2° —2.7° 1891 —1° —6° —2° —3.0° 1903 —6° —1° —3° —3.3° The mean daily temperature for each month of the summer of 1816 fell decidedly below that of any summer month during the period of observations — from 1790 to 1906. MARYLAND WEATHER SERVICE 469 DlSTRIBUTIOX OF PRESSURE DURIXG THE CoOL JUXE OF 1903. In the normal distribution of pressure during the summer months the western edge of the Atlantic high area extends to the South Atlantic states while the barometer over the Central states is comparativeh' low. Hence the atmosphere which flows over the ^liddle Atlantic states comes from the warm southeast. In June of, 1903 the barometer was com- paratively high over the Xorth-Central states, with a maximum in the Upper Mississippi and in the Missouri valleys, and relatively low in the Lower Lake region and the Atlantic coast states. During the same period the western portion of the Atlantic high area was found far north- ward toward the Gulf of St. Lawrence, causing cool easterly in place of the usual warm southerly winds to blow over the Middle Atlantic states ; in addition an unusual flow of cool air was derived from the high area to the northwest over the central portion of the North American continent. The effect of this abnormal distribution of pressure is reflected in the temperature departures recorded in the following table: COOL JUNE OF 1903. Districts. Normal Temp. Departure. New England States 57.8° —5.4° Middle Atlantic States 65.7° —5.2° South Atlantic States 73.4° —3.2° Gulf States 74.5° —4.5° Ohio Valley and Tennessee 68.4° — 5.6° Lake Region 60.8° —3.8° At Baltimore the month was cool, wet, and cloudy. The afternoon temperatures were high on a number of days, but the warm periods were of brief duration. Light frosts occurred in the mountains at the beginning of the month. The rainfall was considerably in excess of the normal seasonal amounts. The mean temperature of the month was 67.0°, 5.8° below the average value for a period of 86 years. This large departure from the average June temperature in Baltimore marks the month of June, 1903, as the coldest since the beginnipg of systematic instriitnciital observations in 1817. (See PL XX1\.) 470 the climate of baltimore Distribution of Pressure During the Normal June of 1902. The temperatures during the month of June, 1903, were very near the normal for a long series of years; they were remarkably uniform, the month being without marked departures from the seasonal average either above or below the normal. The distribution of pressure was in marked contrast with that of the cool June of 1903, described in the preceding paragraphs. The pressure was highest over the South Atlantic states and low over the St. Lawrence Valley and the extreme Southwest. The pressure distribution was such as to give prevailing southwest winds over the Middle Atlantic states. (See PI. XXIV.) While the immediate cause of departures from the normal seasonal temperatures over wide areas may be traced back to abnormal distribu- tion of pressure, it is not so easy to find a cause for these abnormal movements in the positions of the so-called permanent areas of high and low pressure. This slow shifting about of the large areas of high and low barometric pressure is sufficient cause for the greatest observed depar- tures from the seasonal temperature of a given locality, without calling * in the airl of extra-terrestrial influences, such as the moon, the planets, or sun-spots. The Variability of Summer Temperatures. While marked temperature changes from day to day during the summer months are not as frequent or as large as they are during the winter and spring seasons, there is still considerable variability due to the different types of pressure distribution. With a pronounced area of high pres- sure over the Upper Mississippi Valley and the Lake region the tem- peratures in the Middle Atlantic states fall below the seasonal average; with a well developed high area over the South Atlantic states, or just off the coast to the southeast, the temperatures over the Middle Atlantic and the Central states rise above the normal. These two types of pres- sure distribution, with resulting departures from the normal seasonal temperatures, are well defined in the weather maps of July 1, 1885, and July 1, 1901. On the 1st of July, 1885, an area of high pressure cov- VOLUME 2, PLATE XXIV. ^■ V. / Fig. 10.— Cold October, 1905 (—4°). Fig. 11.— Normal October, 1894 (— 0°.l). Fig. 12— Warm October. 1900 (-f 4°.5). D States, BObars, or lines of equal pressure / MARVLAND WEATHER SERVICE E 2, PLATE XXIV. Fio. 3.— Warm December of 1889 (-f 8°.0). Pio. 6.— Warm March of 1898 (+G°.5). Fia. 9.— Warm June of 1899 (+2°) Pio. 12— Warm October. 1900 (+4''.6). DiSTRIBDTION OF PrKSSURE, WiNDS AND TEMPERATURE DURING NORMAL, (^OLD AND WaRM SEASONS IN THE UNITED STATES. Black lines are isotherms, or lines of equal temperature. Red lines are isobars, or lines of equal pressur Arrows fly with the wind. MARYLAND WEATHER SERVICE 471 FiG. 1G3.— The Cold July 1, 18S5. 31 Fic. 1G4.— The Wiirni July ], 1901. 472 THE CLIMATE OF BALTIMORE ered most of the country east of the Kocky Mountains, excepting the New England states, the barometer being highest over the Middle Mis- sissippi Yalley. This distribution of pressure caused a steady flow of cool northwest winds over the Middle Atlantic states. The tempera- tures for the day were abnormally low. The early morning minimum at Baltimore was 56°, the lowest minimum recorded on the first day of July in a period of 36 years. The distribution of pressure noted above is typical for periods of cool weather in all seasons of the year. TEMPERATURES ON JULY 1, 1885. (A cool day with high barometer in Northwest.) 7 a.m. o p. m. 11p.m. Max. Min. 62 76 68 79 56 Mean 67.5° On the other hand the warm weather type is represented by the dis- tribution of pressure seen in the weath?r map showin.u- conditions on the morning of July 1, 1901. In this type the' high area covers the South Atlantic states. With such a distribution of pressure the winds in the Middle Atlantic sta'tes are light and prevailingly from the south or southwest and abnormally warm. The minimum temperature on this day at Baltimore was 80°, while the afternoon maximum reached 103°, the liighest recorded in Baltimore upon the first day of July. TEMPERATURES ON JULY 1, 1901. (A warm day with high barometer in the Southeast.) Hours. 2 4 6 8 10 13 A. M. 82 81 82 88 96 103 Noon P.M. 103 99 96 91 Mean 90.9° 88 87 Midnight THE WEATHER OF JULY 4. The variability of weather conditions may be illustrated in another way, by charting the various climatic factors for a given typical summer day during a long series of years. Thus in the accompanying diagram we have noted the maximum, the mean, and the minimum temperatures, the barometric pressure, the amount of cloudiness, the prevailing wind direction and the amount of rainfall recorded upon each fourth of July MARYLAND WEATHER SERVICE 473 TOE WEATHER OF JULY 4. Year. 1871 Max. Temp. 82 Miu. Temp. (Degrees 71 Mean Temp. Fahr.) 76 Character of Day. Pt. cldy Wind Direction. E&SW Daily Wind Movement. (Miles) 120 Precip- itation. (Inches) 0.03 1872 93 78 86 Pt. cldy SW 117 0.05 1873 92 76 84 Cloudy W 168 1874 92 67 80 Cloudy S&W 100 1.14 1875 83 69 76 Cloudy SE 137 1876 95 73 84 Pt. cldy SW 103 0.22 1877 86 70 78 Pt. cldy N 185 1878 92 73 82 Clear SE 127 1879 98 74 86 Pt. cldy SW&W 170 1880 87 66 76 Clear N 145 1881 93 74 84 Cloudy NW 137 1882 71 61 66 Cloudy NE 178 1.09 1883 94 74 84 Clear SW 207 1884 85 73 79 Cloudy E-SE-S 115 0.04 1885 88 65 77 Clear N&NW 85 1886 86 66 76 Pt. cldy N-E-S 70 1887 86 72 79 Cloudy SE 212 1888 85 64 74 Pt. cldy S 143 . 1889 84 74 79 Cloudy SW 143 0.36 1890 91 71 81 Pt. cldy SW&NW 77 1891 79 59 69 Pt. cldy W&NW 324 0.08 1892 75 61 68 Pt. cldy W&NW 191 1893 84 64 74 Clear NW 155 0.01 1894 86 72 79 Pt. cldy W 243 1895 76 64 70 Cloudy NW 169 0.11 1896 88 70 79 Cloudy SW 177 1897 86 72 79 Pt. cldy E 138 1898 100 74 87 Pt. cldy SW 102 0.07 1889 90 68 79 Pt. cldy S 115 1900 97 77 87 Pt. cldy SW&W 114 1901 96 77 86 Pt. cldy w 103 1902 86 70 78 Cloudy s 117 0.17 1903 87 73 80 Cloudy SE 123 0.08 1904 88 61 74 Clear SW 156 1905 83 69 76 Pt. cldy SE 204 1906 85 73 79 Pt. cldy SW&W 217 0.02 1907 78 59 68 Clear N&E 121 Average 87 70 78 150 0.25 O o O 05 O 00 o o «5 b O o O o - i 1 j 1 ^<^^^ i r- !?^ i ^ ly" y\ : \i ;v7:^ r Hi J '■ ■^ /: < T ; o u \ K K V : i 3; ^Y ■ ^S i i / ; ^ : i ! cc JT : 5 ! —> -1 : V :\ i y *-^ VV : ^ V 1 : ii i 3 1 ■\ 1 A CD cr: ^-» V^/ ' j "^ 2i >^ H 2 S'' y R A \ ' ■ 1 '■ ■ ! ^ < \ : : ■(_)■ Y' ^ i SI \ \i '■ ^ .\ u ;\ ;9i . ; o /^ x' /^ \ 'n ' * ■ £ V :/ i f i^:^ 1 1 \ \i ^rS^ /< A 1 iV' ' i I < \i 'n ij ^ |: ] \\ O '^ 1 I i / i 4' / \ :[ \i ; iAi, f —^ ;;^ $^ ^> \\ J rih ^ \ -< ^ ^ A ^rn' \ ■ o \\ \' u ; rS'^ IB / >-|- i/ /; :\ 1 iVi . \ \^ 1 u ^XiJ /' • 3 ; v;'' >^\ '^ 1/ ''■l g -CH- OI Ja;» i^ lO -> s: A k i : (XL *ci — i — 22 •^^ ^ m i QC: i — r? rr- |r ■q= rly ; L ^ 1 i : y /i i 1 J-H 00 :u — i — i i ll iS- ^ i MARYLAND WEATHEE SERVICE 475 i'rom 1S71 to 1903. We see that the lowest temperature recorded on any -ith of July was 58° and that this occurred in 1891; the highest temperature, namely 100°, is credited to 1898. Between these extremes we have had in the past 36 years all degrees of temperature conditions. There appears to be no regularity in the fluctuations in temperature from year to year, although there are indications of irregular periods of steadily decreasing or increasing temperature. The barometric pressure has varied but little above or below the normal seasonal values, the entire range being less than half an inch. The days with a clear sky have numbered but six in 36 years; with partly cloudy sky, 18 ; and with an overcast sky, 12. The prevailing wind direction has been from the southwest. Kain fell in amounts vary- ing from a light sprinkle to heavy showers on somewhat less than half the total number of days, making the rainfall probability for the 4th of July less than 50 per cent. Thunderstorms were recorded but five times during the period of 36 years on this day. West Indian Hurricanes. Hurricanes do not differ, in essential features, from the temperate region cyclones described in preceding pages. They are more restricted in area, but relatively more intense in energy and destructive power. These storms have their origin in the vicinity of the Windward Islands; they move toward the west or northwest at the rate of 10 to 12 miles per hour — less than half the average rate of temperate region cyclones — and curve northward and then northeastward approximately in the neighborhood of Florida, as a rule, following the Atlantic coast, enlarg- ing in area after recurving until they resemble in every detail the storms common to the higher latitudes. While these, the most disastrous of all storms, have occurred in all seasons of the year, they are confined almost entirely to the months of August, September, and October. The abrupt increase in their frequency in August is phenomenal as shown in the following extract from one of the publications of the United States Weather Bureau.' ^ E. B. Garriott. West Indian Hurricanes. Bull. H., U. S. Weather Bureau. 4to. Washington, D. C, 1902. 476 THE CLIMATE OF BALTIMORE FREQUENCY OF HURIilCANES (1878-1900). Jan. Feb. Mar. Apr. May. June. July. Aug. Sept. Oct. Nov. Dec. Vear. 3 1 3 3 25 25 32 3 3 98 Fortunately the path of the hurricane rarely falls within the limits of the Middle Atlantic states until it has lost some of its violence. By the time it has reached the latitude of Baltimore the center is generally well off the coast, and we experience only the ordinary storm winds of the western quadrant of the liurricane. When there happens to be a well developed area of high barometric pressure over the eastern half of the country on the approach of a hurri- cane the storm is prevented from recurving near the Florida Peninsula and moves slowly westward into the Gulf of Mexico, or even entirely across the Gulf, before recurving northward. Under such circumstances the hurricane is apt to gather in force in its journey across the Gulf. The storm which destroyed Galveston in September, 1900, was of this type. A typical storin of this class passed over Maryland on the 13th of October, 1893 ; it is described in the following paragraphs. THE HURRICANE OF OCTOBER 13, 1893. The first indication of the approach of this storm from the West Indies was contained in a report from Saint Thomas on the 5th of October; on the following day additional information was received from Antigua. The storm advanced slowly westward.' On the Tth it was southeast of Port au Prince, and on the 8th southeast of Santiago de Cuba. On the 9th it had reached the Bahama Islands and Southern Florida, and storm signals were ordered up along the Florida and east Gulf coasts by the Chief of the Weather Bureau. By the evening of the 10th the wind had freshened to a gale along the Florida coast. On the morning of the 11th the storm center was east of the Bahama Islands and the barometer was falling rapidly along the Atlantic coast as far north as Xew Jersey. During the 12th severe nortlieast gales and heavy '.See.- Lake Storm Bulletin. No. 2, 1893, U. S. Weather Bureau. MAKYLAND WKITHER SERVICE 477 rains prevailed along the coast of the South Atlantic states in connec- tion with a rapidly falling barometer. On the morning of the 13th the storm center reached the South Caro- lina coast, the barometer at Charleston indicating 28.88 inches. From this point the storm took an unusual course, moving northward into the interior, the center passing over the Carolinas and the Middle Atlantic states. Xortheast storm signals were ordered for all stations on the Middle Atlantic and Kew England coasts. Special w^arnings were sent to all Weather Bureau observers in the Middle Atlantic states and Kew England, and observers from Southern New England to Maryland were authorized to use tlie telegraph at their discretion in distributing these warnings in the most effectual manner possible. During the evening of the 13th the center of the storm passed over Western Maryland, the barometer falling to 28.88 inches at Baltimore. Moving due north it crossed Pennsylvania and Western N'ew York to the north of Lake Ontario on the 14th. On the 15th the st>orm disappeared in the direction of Labrador. The storm was attended by high winds and heavy rains all along its path across the United States. Some of the records are quoted below from the official report of the United States Weather Bureau. niGII WINDS AND HEAVY RAINS DURING THE STORM OF OCTOBER 13 AND 14, 1893. (S a. m. to 8 p. m. October 13.) „. .. Velocity. Direc- > 5 C > 6 o O irt O iTi «^j - — d o ■^. ■■ 1 ■ , 1 '' C\'' ' ' ; %y^ ; > F-^ / \ \ iVT": i^ i i o o f ( ( 1 \ \^ \ i ^ ! ~az- CD T. UJ IZ) UJ o. Z L.J V ^ X / ■<: y'^ ( ,^ ■f^ i. ^ CO CD K : ; « s ^-s ^s y * UJ :y: T ; \ , ? \ :Y'' * • ■rn' i ■ ^ // \ a: :%\i\ z 1 ;2:J: o ^ i f /{/ / :1 'Vij'; s^ \^ '^' "^ i V wP \ : : ;^: .H : ! \ S^ ) ; M •^' y : / ^/ \ V :T' >: ^ N N,\ I ; / V ■ ^ i)^ 2 /?/ / X ; 1 Hv ?^V ) * : ; ^ ^ Q- n . ^<^'/ / ^'■ o i / : ; £ ; ■ ^"^' y ^ ^i s o / / f' J^ > ;C)":*1 : -i- UJ O- "uJ :/; ■i' 1 i V \^: r : 1 \ \ 2Y' ' - 2 X^ i: \ \ "A: <^ ^^^ /'; : : p" ^N : / 1 : 1 ; ■ "^ '• 00 ■ - ; ■ -'—J Ti'i ■ : ^ ■■ CI — ' -