key: cord-0746450-fcth35am
authors: Zangari, Shelby; Hill, Dustin T.; Charette, Amanda T.; Mirowsky, Jaime E.
title: Air quality changes in New York City during the COVID-19 pandemic
date: 2020-06-25
journal: Sci Total Environ
DOI: 10.1016/j.scitotenv.2020.140496
sha: 4b07b099b197b624fb5ef4b3397a136fae04b5c9
doc_id: 746450
cord_uid: fcth35am

Abstract In December 2019, a new, severe coronavirus (COVID-19) appeared in Wuhan, China. Shortly after, the first COVID-19 case was confirmed in the United States. The emergence of this virus led many United States governors to enact executive orders in an effort to limit the person-to-person spread of the virus. One state that utilized such measures was New York, which contains New York City (NYC), the most populous city in the United States. Many reports have shown that due to the government-backed shutdowns, the air quality in major global cities improved. However, there has been only limited work on whether this same trend is seen throughout the United States, specifically within the densely populated NYC area. Thus, the focus of this study was to examine whether changes in air quality were observed in NYC resulting from New York State's COVID-19-associated shutdown measures. To do this, daily concentrations of fine particulate matter (PM2.5) and nitrogen dioxide (NO2) were obtained from 15 central monitoring stations throughout the five NYC boroughs for the first 17 weeks (January through May) of 2015–2020. Decreases in PM2.5 (36%) and NO2 (51%) concentrations were observed shortly after the shutdown took place; however, using a linear time lag model, when changes in these pollutant concentrations were compared to those measured during the same span of time in 2015–2019, no significant difference between the years were found. Therefore, we highlight the importance of considering temporal variability and long-term trends of pollutant concentrations when analyzing for short-term differences in air pollutant concentrations related to the COVID-19 shutdowns.

J o u r n a l P r e -p r o o f

In December 2019, a new severe coronavirus appeared in Wuhan, China (Martelletti and Martelletti, 2020) . Shortly after, on January 21, 2020, the first COVID-19 case was confirmed in the United States (US) (Schuchat, 2020) . From January 21 st to February 23 rd , a total of 14 cases of COVID-19 emerged in six US states, and many of these cases were from travelers who were either arriving from abroad or had contact with people having confirmed infections (Schuchat, 2020) . However, in late February, COVID-19 cases were being reported from people who had no recent travel or links to those with known confirmed cases, signaling the start of a community-based spread (Schuchat, 2020) . The World Health Organization (WHO) declared the disease a pandemic on March 11, 2020 (Cucinotta and Vanelli, 2020) .

The first case of COVID-19 confirmed in New York (NY) State was on March 1, 2020, and by March 20, the NY governor officially put "NY on PAUSE"; this executive order included directives such as stay-at-home orders and lockdown measures, social distancing measures, and a 100% reduction of in-person work forces within non-essential businesses. At the peak of the outbreak in NY, which occurred on April 4 th , there were as many as 12,000 new COVID-19 cases per day (The New York Times, 2020). As of May 12 th , 2020, there had been a total of 191,320 confirmed cases of COVID-19 and 19,736 deaths due to the virus in NYC; these numbers constituted 55.7% of the total cases and 72.3% of the total deaths in NY state (The New J o u r n a l P r e -p r o o f 4 NY was not the only region with limitations restricting people's movements, as these or similar policies were enacted throughout many other US states as well as globally (International Organization for Migration, 2020). As a result, researchers began investigating the impact of these measures on air quality under the assumption that fewer people would be driving and thus there would be lower levels of traffic-related air pollutants (Abdullah et al., 2020; Dantas et al., 2020; Zambrano-Monserrate et al., 2020) . Research groups in major cities such as Sao Paulo, Brazil; Wuhan, China; and Barcelona, Spain reported decreases in traffic-related air pollutant concentrations from January to May 2020 (Abdullah et al., 2020; Cadotte, 2020; Kambalagere, 2020; Li et al., 2020; Tobías et al., 2020; Zambrano-Monserrate et al., 2020; Zheng et al., 2020) , which they suggest is a result of reduced vehicle emissions. Several studies have also shown similar reductions in pollutant concentrations by comparing 2020 concentrations to those measured in previous years Dantas et al., 2020; Dutheil et al., 2020; Freitas et al., 2020; Isaifan, 2020; Muhammad et al., 2020; Nakada and Urban, 2020; Sharma et al., 2020; Xu et al., 2020) . Similar work has been completed in the US, but to a much lesser extent (Bechle et al., 2013; Berman and Ebisu, 2020; EarthSky Team, 2020; McLinden et al., 2014; Muhammad et al., 2020; Porterfield, 2020; Stein, 2020; Volcovici, 2020) . Only one study conducted in the North China Plain showed no significant change in air quality before and after shutdowns took place (P. .

Although general COVID-related decreases in the concentrations of air pollutants have J o u r n a l P r e -p r o o f 6 there has been a significant short-term decline in pollution levels in 2020 because of the shutdown due to COVID-19. This model included a dummy covariate for year. The model statement is:

Where 0 is the intercept for 2020, 1 is the coefficient for time, , (in our case is our covariate, days from January to May), 1 is the slope for 2020. We compared all years to 2020; therefore, , is the intercept for each year (2015-2019), is the slope term for each

year (2015-2019), is the time lag for each year (2015-2020), and is the error term.

Thus, using this model, varying the intercept by year allows for testing for differences in the level of pollution for each year compared to 2020, and varying the slope tests for differences in the rate of change of pollution for each year compared to 2020. We conducted an ANCOVA using an F-test for Type III sums of squares on the regression model testing for homogeneity of intercepts and homogeneity of slopes for each year and the results for the varying intercept. For a full model statement and output, refer to Appendices 1 and 2. All analyses were completed in RStudio (Version 4.0.0) (R Core Team, 2020).

The average daily PM 2.5 concentrations during the first full 17 weeks of 2015-2020 in the NYC area can be seen in Figure 3 . First, we wanted to assess how the average PM 2 concentrations between the years could have been a continuation of the improvement in air quality rather than as a result of the COVID-19 shutdowns. In this current study, we found that compared to 2020, the years 2015, 2016, 2017, and 2018 had significantly higher intercepts than that observed in 2020 (Table 1) Our study is not the first to observe temporal variability in PM 2.5 mass concentrations in and around NYC. In Mirowsky et al. (2013) , different size fractions of PM were collected during a winter and summer season at five locations in the NYC metropolitan area; these locations included both rural and urban sampling sites. This group found that PM 2.5 mass concentrations J o u r n a l P r e -p r o o f 8 were greater at all five locations in the winter compared to the summer months. In another study, using several of the same DEC monitoring sites as was used in this current work, researchers assessing the long-term variability of air pollutant concentrations in NY found that colder months had higher PM 2.5 concentrations than the warmer months (Schwab et al., 2012) .

Every year between January and May, NY State transitions between the winter and spring seasons. Due to decreased levels of solar flux and photochemical activity and less instances of long-range pollutant transport, winter months typically have higher levels of air pollution than those measured in the summer; this is especially evident in the northeast region of the country where NY is located (Rattigan et al., 2010; Schwab et al., 2012; Squizzato et al., 2018) . Thus, although social distancing and lockdown measures might have resulted in a decrease in NYC traffic, these measures may not necessarily have had a large enough or direct impact on the air quality of the NYC metropolitan during this time.

The average daily NO 2 concentration measured in Region 2 during the first full 17 weeks of 2015-2020 can be seen in Figure 4 . We found that compared to 2020, NO 2 concentrations were higher in all the subsequent years going back to 2015 (Table 1) Previous studies of nitrogen oxides (NO x ) in NY have found that the concentrations of this pollutantsimilar to PM 2.5vary temporally, with NO 2 levels being higher in the winter and lower in the summer (Masiol et al., 2017; Squizzato et al., 2018) . Our results corroborate those prior studies and support the conclusion that social distancing and shutdown measures implemented as a result of the COVID-19 pandemic may not have had a direct impact on commonly measured traffic-related air pollutants in the NYC metropolitan area.

Overall, we found that concentrations of PM 2.5 and NO 2 decreased from January 2020 to May 2020; however, this decrease is similar in magnitude to that observed during the same time in the previous five years at this location. This suggests that no or minimal improvements in NYC's air quality were observed as a result of the COVID-19 government-backed shutdowns. nearly two times lower than Brazil (IQAir, 2019). Thus, we can speculate that major reductions in air quality were only found in places that had higher levels of air pollutants before COVID-19

hit, compared to locations with relatively clean air to begin with; this observation has also been suggested in other work (IQAir, 2020) .

This study has several limitations. We focused on two main pollutants commonly associated with traffic in a major metropolitan area in the US. Other studies have also included analyses for ozone (O 3 ), which is a secondary pollutant and highly variable based on season and temperature.

Other pollutants commonly associated with changes in air quality, including NO, NO x , CO, and SO 2 were not examined as part of this work because of the lack of air quality monitoring stations for these pollutants in NYC. Additionally, NO 2 was only monitored at three of the 15 locations in our study area. We did not report changes in traffic levels or social distancing metrics in and around NYC to support whether changes in traffic or people's movements changed during our study time.

This study also has several strengths. In comparison to similar studies, this work is the first detailed investigation of the impacts of COVID-19 shutdowns on the ground-level air quality in a major US city. One recently published manuscript did assess changes in PM 2.5 and NO 2 concentrations across the continental US (Berman and Ebisu, 2020) ; however, that work utilized J o u r n a l P r e -p r o o f 11 EPA AirNow data that has not been evaluated for quality assurance. In addition, the researchers of that study averaged air quality data across the country to evaluate changes in pollutant concentrations over time rather than focus exclusively on one geographic area. Additionally, several other studies outside the US have utilized satellite data to estimate pollutant concentrations at ground level. By using monitoring station data, we are able to measure direct effects at the ground level, thus allowing us to measure time-resolved changes in air quality with better precision than satellite estimates (Bechle et al., 2013; McLinden et al., 2014) .

Despite studies throughout the world demonstrating that air quality is improving because of the social distancing and shutdown measures put in place due to COVID-19, we found no significant changes in NYC. Although the air quality may have improved in other regions, our lack of findings may be because our analysis accounted for both short-and longer-term changes in air quality in our models and NYC has lower baseline concentrations of air pollutants compared to the other locations being studied.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

J o u r n a l P r e -p r o o f 12 analysis review and Terry Gordon and Mary Collins for reading through this manuscript prior to publication. 

Air quality status during 2020 Malaysia Movement Control Order (MCO) due to 2019 novel coronavirus (2019-nCoV) pandemic

Changes in U.S. air pollution during the COVID-19 pandemic

Early evidence that COVID-19 government policies reduce urban air pollution

Air Pollution Reduction and Mortality Benefit during the COVID-19 Outbreak in China 1

WHO Declares COVID-19 a Pandemic

The impact of COVID-19 partial lockdown on the air quality of the city of Rio de Janeiro

COVID-19 as a factor influencing air pollution?

Satellite data show 30% drop in air pollution over northeast U.S. | Earth | EarthSky [WWW Document

Mobility restrictions and air quality under COVID-19 pandemic in São Paulo

International Organization for Migration

REPORT: COVID-19 Impact On Air Quality In 10 Major Cities | IQAir

WORLD AIR QUALITY REPORT: Region & City PM2.5 Ranking

The dramatic impact of Coronavirus outbreak on air quality: Has it saved as much as it has killed so far? Glob

A Study on Air Quality Index (AQI) of Bengaluru, Karnataka during Lockdown Period to Combat Coronavirus Disease (Covid-19): Air Quality Turns 'Better' from 'Hazardous

Analysis of major air pollutants and submicron particles in New York City and Long Island

Improved satellite retrievals of NO2 and SO2 over the Canadian oil sands and comparisons with surface measurements

The effect of particle size, location and season on the toxicity of urban and rural particulate matter

COVID-19 pandemic and environmental pollution: A blessing in disguise?

COVID-19 pandemic: Impacts on the air quality during the partial lockdown in São Paulo state

Department of Environmental Conservation • Air Monitoring Website

Table 2: Population, Land Area, and Population Density by County

NYCdata: Metropolitan Statistical Area (MSA) Population

Coronavirus Shutdown Means Cleaner Air For Some U.S. Cities

R: A language and environment for statistical computing

Multi-year hourly PM 2.5 carbon measurements in New York: Diurnal, day of week and seasonal patterns

Coronavirus (COVID-19) Deaths -Statistics and Research -Our World in Data

Public Health Response to the Initiation and Spread of Pandemic COVID-19 in the United States

New York State Urban and Rural Measurements of Continuous PM 2.5 Mass by FDMS, TEOM, and BAM

Effect of restricted emissions during COVID-19 on air quality in India

PM2.5 and gaseous pollutants in New York State during 2005-2016: Spatial variability, temporal trends, and economic influences

NOAA exploring impact of COVID-19 response on the environment

The New York Times, 2020. New York Coronavirus Map and Case Count -The New York Times

Changes in air quality during the lockdown in Barcelona (Spain) one month into the SARS-CoV-2 epidemic

CDC COVID Data Tracker [WWW Document

Urban air quality improves as coronavirus empties U.S. highways: NOAA

Severe air pollution events not avoided by reduced anthropogenic activities during COVID-19 outbreak

Changes in air quality related to the control of coronavirus in China: Implications for traffic and industrial emissions

Impact of the COVID-19 Event on Air Quality in Central China

Indirect effects of COVID-19 on the environment

Significant changes in the chemical compositions and sources of PM2

The authors would like to acknowledge the Center for Environmental Medicine and Informatics at SUNY ESF for financial support. We would also like to thank Michael Petroni for statistical