key: cord-0474279-5v3f3cw5
authors: Liu, Zhu; Ciais, Philippe; Deng, Zhu; Davis, Steven J.; Zheng, Bo; Wang, Yilong; Cui, Duo; Zhu, Biqing; Dou, Xinyu; Ke, Piyu; Sun, Taochun; Guo, Rui; Boucher, Olivier; Breon, Francois-Marie; Lu, Chenxi; Guo, Runtao; Boucher, Eulalie; Chevallier, Frederic
title: Carbon Monitor: a near-real-time daily dataset of global CO2 emission from fossil fuel and cement production
date: 2020-06-13
journal: nan
DOI: nan
sha: fb856d3f94611a3bc16f5623e106121f33f1f80a
doc_id: 474279
cord_uid: 5v3f3cw5

We constructed a near-real-time daily CO2 emission dataset, namely the Carbon Monitor, to monitor the variations of CO2 emissions from fossil fuel combustion and cement production since January 1st 2019 at national level with near-global coverage on a daily basis, with the potential to be frequently updated. Daily CO2 emissions are estimated from a diverse range of activity data, including: hourly to daily electrical power generation data of 29 countries, monthly production data and production indices of industry processes of 62 countries/regions, daily mobility data and mobility indices of road transportation of 416 cities worldwide. Individual flight location data and monthly data were utilised for aviation and maritime transportation sectors estimates. In addition, monthly fuel consumption data that corrected for daily air temperature of 206 countries were used for estimating the emissions from commercial and residential buildings. This Carbon Monitor dataset manifests the dynamic nature of CO2 emissions through daily, weekly and seasonal variations as influenced by workdays and holidays, as well as the unfolding impacts of the COVID-19 pandemic. The Carbon Monitor near-real-time CO2 emission dataset shows a 7.8% decline of CO2 emission globally from Jan 1st to Apr 30th in 2020 when compared with the same period in 2019, and detects a re-growth of CO2 emissions by late April which are mainly attributed to the recovery of economy activities in China and partial easing of lockdowns in other countries. Further, this daily updated CO2 emission dataset could offer a range of opportunities for related scientific research and policy making.

The main cause of global climate change is the excessive anthropogenic emission of CO2 to the atmosphere from geological carbon reservoirs, from the combustion of fossil fuel and cement production. Dynamic information on fossil fuel-related CO2 emissions is critical for understanding the impacts on climate due to different human activities, and their variability, on the forcing of climate change. Further, the combustion processes of fossil fuel also emit short-lived pollutants such as SO2, NO2 and CO. Therefore, such information would also allow a more accurate quantification and better understanding of air quality changes 1,2 .

Estimates of CO2 emissions from fossil fuel combustion and cement production 2-8 are based on both activity data (e.g., the amount of fuel burnt or energy produced) and emission factors (See Methods) 9 . The sources of these data are mainly national energy statistics, although a number of organizations such as CDIAC, BP, EDGAR, IEA and GCP also produce and compile estimates for different groups of countries or for all countries 1, [10] [11] [12] . The reported fossil fuel-related CO2 emissions are usually on an annual basis while lagging the very year's emissions by at least one year.

The uncertainty associated with CO2 emissions from burning fossil fuel and producing cement is small when considering large emitters or the global totals, smaller than that of co-emitted combustion-related pollutants for which uncertain technological factors influence the ratio of emitted amounts to fossil fuel burnt [13] [14] [15] . The uncertainty of global carbon emissions from fossil fuel burning and cement production varies between ±6% and ±10% 5,7,16,17 (±2σ) . The uncertainty is attributed to both the activity data and the emission factors. For the activity data, the amount of fuel burnt is recorded by energy production and consumption statistics, hence the uncertainties are introduced by errors and inconsistencies in the reported figures from different sources. For the emission factors, the different fuel types, quality and combustion efficiency together contribute to the overall uncertainty. For example, coal used in China is of variable quality and so is its emission factors, both before (raw coal) and after cleaning (cleaned coal) varies significantly, which was found to cause a 15% uncertainty range for CO2 emissions. On the other hand, there is very limited temporal change of emission factors. For example, the annual difference of emission factors for coal consumption was within 2% globally 18 while the variation of emission factors for oil and gas was found to be much smaller.

Given the fact that the uncertainty of CO2 emissions from fossil fuel burning and cement production is in general under ±10% 10, 19, 20 , and the annual difference of emission factors is less than 2% 18 , the CO2 emission thus can be estimated directly by estimating the absolute amount and the relative change of activity through time. This method has been widely used for scientific products that update recent changes of CO2 emissions estimates 1,21, 22 23 , understanding that official and comprehensive CO2 national inventories reported by countries to the UNFCCC become available with a lag of two years for Annex-I countries and several years for non-Annex-I ones 24 . As such, a higher spatial, temporal and sectoral resolution of CO2 emission inventories beyond annual and national level can be obtained by spatial, temporal and sectoral data to disaggregate the annual national emissions 9, 14, 23, 25 . The level of granularity of spatially explicit dynamic emission inventories depends on available data, such as location and operations of point sources 23 (i.e. power generation for a certain plant), regional statistics of energy use (i.e. monthly fuel consumption) 9, 25 , and knowledge of proxies for the distribution of emissions such as gridded population density, night lights, urban forms and GDP data etc. 9, 14, 23, 25 .

Gaining from past experiences of constructing annual inventories and newly compiled activity data, we present in this study a novel daily dataset of CO2 emissions from fossil fuel burning and cement production at national level. The countries/regions include China, India, U.S., Europe (EU27 & UK), Russia, Japan, Brazil, and rest of world (ROW), as well as the emissions from international bunkers. This dataset, known as Carbon Monitor, is separated into several key emission sectors: power sector (39% of total emissions), industrial production (28%), ground transport (18%), air transport (3%), ship transport (2%), and residential consumption (10 %). For the first time, daily emissions estimates are produced for these six sectors, based on dynamically and regularly updated activity data. This is made possible by the availability of recent activity data such as hourly electrical power generation, traffic indices, airplane locations and natural gas distribution, with the assumption that the daily variation of emissions is driven by the activity data and that the contribution from emission factors is negligible, as they evolve at longer time scales, e.g. from policy implementation and technology shifts.

The framework of this study is illustrated in Fig 1. We calculated national CO2 emissions and international aviation and shipping emissions since the Jan 1 st 2019, drawing on hourly datasets of electricity power production and their CO2 emissions in 29 countries (thus including the substantial variations in carbon intensity associated with the variable mix of electricity production), daily vehicle traffic indices in 416 cities worldwide, monthly production data for cement, steel and other energy intensive industrial products in 62 countries/regions, daily maritime and aircraft transportation activity data, and either previousyear fuel use data corrected for air temperature to residential and the commercial buildings. Together, these data cover almost all fossil fuels and industry sources of global CO 2 emissions, except for the emission from land use change (up to 10% of global CO 2 emissions) and non-fossil fuel CO 2 emissions of industrial products (up to 2% of global CO 2 emissions) 26 in addition to cement and clinker (i.e. plate glass, ammonia, calcium carbide, soda ash, ethylene, ferroalloys, alumina, lead and zinc etc.).

While daily emission can be directly calculated using near-real-time activity data and emission factors for the electricity power sector, such an approach is difficult to apply to all sectors. For the industry sector, emissions can be estimated monthly in some countries. For the other sectors, we used proxy data instead of daily real activity data, to dynamically downscale the annual or monthly CO2 emissions totals on a daily basis. For instance, traffic indic es in cities representative of each country were used instead of actual vehicle counts and categories, combined with annual national total sectoral emissions, to produce daily road transportation emissions. As such, for the road transportation, air transportation and residential use of fuels sectors in most countries, we downscaled monthly or annual total emission data in 2019 to calculate the daily CO2 emission in the very year. Subsequently, we scaled monthly totals of 2019 by daily proxies of activities to obtain daily CO2 emissions data in the first four months of 2020, during the unprecedented disturbance of the COVID-19 pandemic. The Carbon Monitor near-real-time CO2 emission dataset shows a 7.8% decline of CO2 emission globally from January 1 st to April 30 th in 2020 when compared with the same period in 2019, and detects a re-growth of CO2 emissions by late April which are mainly attributed to the recovery of economy activities in China and partial easing of lockdowns in other countries. 

Where , , are indices for regions, sectors and fuel types respectively. can be further separated into the net heating values for each fuel type (the energy obtained per unit of fuel), the carbon content per energy output (t C/TJ) and the oxidization rate (the fraction (in %) of fuel oxidized during combustion):

Due to the lag of more than two years in publishing governmental energy statistics, we started from the latest CO2 emissions estimates up to 2018 from current CO2 databases 1,10-12 . For 2019, we completed this information to obtain annual total emissions based on literature data and disaggregated the annual total into daily emissions (see below). For 2020, we estimated daily CO2 emissions by using daily changes of activity data in 2020 compared to 2019. The In this study, the EDGAR sectors were aggregated into four sectors ( ): power sector, industry sector, transport sector (ground transport, aviation and shipping), and residential sector. This is consistent with the new activity data we used below to compute daily variations. We used the sectoral distribution in 2018 from EDGAR to infer the sectoral emissions in 2019 for each country/region (Equation 4 ), assuming that the sectoral distribution remained unchanged in these two years. 

According to IPCC Guidelines 4 , the CO2 emissions for sector could be calculated by multiplying sectoral activity data by their corresponding emission factors following Equation 5:

The emissions were here calculated following this equation, separately for the power sector, the industry sector, the transport sector, and the residential sector.

The CO2 emissions from power sector can be calculated by adapting Equation 5 with sector specific activity data (i.e. electricity production/thermal electricity production) and corresponding emission factors (Equation 6):

Normally the emission factors change slightly over time but can be assumed to remain constant over the two years period considered in this study, compared to the huge changes in activity data. Thus, we assumed that emission factors remained unchanged in 2019 and 2020, and calculated the daily emissions as follows:

The data sources of daily activity data in power sector are described as Table 2 . The countries/regions listed in Table 2 account for more than 70% of the total CO2 emissions in the power sector. For emissions from other countries (ROW), which are not listed in Table 2, we estimated the power sector emission changes in 2020 based on the period of the national lock-down. For daily emission changes of ROW in 2019, we firstly assumed a linear relationship between daily global emission and daily total emissions of the ROW countries listed in Table 2 . Then we classified each country according to whether they adopted lockdown measures, based on official reports. Based on daily emission data of the power sector of the countries listed in Table 2 , we calculated the respective average change rates of power sectors in ROW countries between January and April, assuming changes started since the date of lock-down in each country. Emissions from countries with no lock-down were left unchanged. We then applied these country-specific January to April emissions growth rates to estimate daily changes for each ROW country in 2020, based on their lock-down measures, and aggregated them into daily emission for ROW. 

While daily production data is not directly available for industrial and cement production, the monthly CO2 emissions from industry and cement production sector could be calculated by using monthly statistics of industrial production, and daily data of electricity generation to disaggregate the monthly CO2 emissions into daily values. This calculation assumes a linear relationship between daily electricity generation for industry and daily industry production data to compute daily industry production.

The emissions from industrial production during the fossil fuel combustion were calculated 

The emissions from cement production during the chemical process of calcination of calcite were calculated with the same Eq. 

For the monthly emissions in 2020 in country/region , we used the following equation:

where is the industrial production in different industrial sectors (in China) or a total Industrial Production Index (in other countries) as listed in Table 3 . In China's case, the January and February estimates were combined as no individual monthly data was reported by sources listed in Table 3 for these two months. The monthly industrial emissions were disaggregated to daily emissions using daily electricity data, as explained above.

Lacking in the latest Industrial Production Index in April 2020 for Europe, India, Japan,

Russia and Brazil, we adopted monthly growth rates of industrial output from Trading Economics (https://tradingeconomics.com) based on preliminary survey data. For other countries not listed in Table 3 , we used the same method as described for the power sector to calculate the daily industry emissions from ROW.

To allocate monthly emissions into daily emissions, we use the weight of daily electricity production to monthly electricity production: represents the extra time spent on a trip, in percentage, compared to uncongested condition.

TomTom congestion level data were obtained for 416 cities across 57 countries at a temporal resolution of one hour. Of note that a zero-congestion level means that the traffic is fluid or 'normal', but does not mean there was no vehicle and zero emissions. It is thus important to identify the lower threshold of emissions when the congestion level is zero. To do so, we compared the time series of daily mean TomTom congestion level , with the daily mean car flux (called hereafter in vehicle per day) from publicly available real-time data from an average of 60 roads in the Paris megacity. Those daily mean car counts were reported by the City's service (https://opendata.paris.fr/pages/home/). We used a sigmoid function to fit the relationship between and (Fig 2) :

where a, b, c and d are the regression parameters (Table 4 ). We verified that the empirical fit from Eq. 

For countries not included in the TomTom dataset, we assumed that the emission changes follow the mean changes of other countries. For example, Cyprus, as an EU member country, had no city reported in TomTom dataset, so its relative emission change was assumed to follow the same pattern of the total emissions from other EU countries included in TomTom dataset (which covers 98% of EU total emissions). Similarly, the relative emission changes of countries in ROW but not reported by TomTom were assumed to follow the same pattern of the total emissions from all TomTom reported countries (which cover 85% of global total emissions). 

We calculated CO2 emissions from commercial aviation following a commonly used approach: reconstructing the emission inventories from bottom up based on the knowledge of the parameters of individual flights. We collected the FlightRadar24 data (https://www.flightradar24.com/) for the departure and landing airports for each flight, the calculate the distance flown assuming the shortest distance for each flight, and then CO2 emissions per flight 31 . Flights were grouped per country, and for each country between domestic or international traffic. The daily CO2 emission was computed as the product of distance flown, by a CO2 emission factor per km flown, according to:

We acquired monthly individual commercial flight information from FlightRadar24. The FlightRadar24 database has incomplete data for some flights and may miss altogether a small fraction of actual flights 31 , so we scaled the EDGAR estimate of CO2 emissions (inflated by 5.7% for the year 2019) with the total estimated number of kilometers flown in 2019 (67.91 million km) and apply this scaling factor to 2020 data. We assumed that the fraction of missed flights was the same in 2019 and 2020, which is reasonable.

We collected international CO2 shipping emissions from 2016-2018 based on the EDGAR's international emissions. We also. collected global shipping emissions during the period of 2007-2015 from IMO 33 and ICCT (https://theicct.org/sites/default/files/publications/Global-shipping-GHG-emissions-2013-2015_ICCT-Report_17102017_vF.pdf). According to the Third IMO GHG Study 33 , CO2 emissions from international shipping accounted for 88% of global shipping emissions, domestic and fishing accounts for 8% and 4%, respectively. We calculated international CO2 shipping emissions from 2007-2015 from global shipping emissions and the ratio of international shipping and global shipping emissions. We extrapolated emissions from linear fits 2007-2018 to estimate the emissions in 2019. The data sources of shipping emissions are in Table 6 . We obtained emissions for the first quarter of 2019 based on the assumption the equal distribution of monthly shipping CO 2 emissions. The equations are as follows:

is the increasing rate of international shipping emissions in 2019 based on the linear extrapolation of data from the period 2007-2018, estimated to be of 3.01%.

ℎ represents the ratio of the months to be calculated in the whole year. Given this, we estimated the shipping emissions for the first quarter of 2019, ℎ equals 121/365. We assumed that the change in shipping emissions was linearly related to the change in ships. Traffic volume. The change of international shipping emissions for the first four months of 2020 was calculated according to the following equation:

,2020 = ,2019 × (17) Where represents the ratio of the change in shipping emissions, estimated to the end of Apr by -15% compared to the same period of last year according to https://www.theedgemarkets.com/article/global-container-shipments-set-fall-30-next-fewmonths. The main assumption is this approach is that residential emissions did not change from other factors than heating degree days variations in 2020, when people time in houses dramatically increased during the lock-down period. In order to test the validity of this assumption, we compiled natural gas daily consumption data by residential and commercial buildings for France (https://www.smart.grtgaz.com/fr/consommation) (unfortunately such data could not be collected in many countries) during 2019 and 2020. Natural gas consumption in kWh per day was transformed to CO 2 emissions using an emission factor of 10.55 kWh per m 3 and a molar volume of 22.4 10 -3 m 3 per mole.

Firstly, we verified that the temporal variation of those 'true' residential CO 2 emissions was similar to that given by equations (18) to (20) . Secondly, after fitting a piecewise model to those natural gas residential emission data using ERA5 air temperature data, we removed the effect of temperature to obtain an emission corrected for temperature effects. Even if the lock down was very strict in France, we found no significant emission anomaly, meaning that the fact that nearly the entire population was confined at home did not increase or decrease emissions. This complementary analysis tentatively suggests that residential emissions can be well approximated in other countries by equations (18) to (20) based only on temperature during the lock down period. 

Currently there are 27484 data records provided in this dataset:

-268 records are daily mean CO2 emissions (from fossil fuel combustion and cement production process) 1751-2020.

-4374 records are the daily emissions for 9 countries/regions (China, India, US, EU27&UK, Russia, Japan, Brazil, ROW and Globe) and 486 days (from January 1 st 2019

to April 30 th 2020).

-22842 records are daily emissions in power sector, ground transport sector, industry sector, residential sector, aviation sector and international shipping sector respectively, for 9 countries/regions (China, India, US, EU27&UK, Russia, Japan, Brazil, ROW and Globe) and 486 days (from January 1 st 2019 to April 30 th 2020).

We followed the 2006 IPCC Guidelines for National Greenhouse Gas Inventories to conduct the uncertainty analysis of the data. Firstly, the uncertainties were calculated for each sector:

-Power sector: the uncertainty is mainly from inter-annual variability of coal emission factors. Based the UN statistics the inter-annual variability of fossil fuel is within (±1.5%), which been used as uncertainty of the CO2 from power sectors.

-Industrial sector: Uncertainty of CO2 from Industry and cement production comes from the monthly production data. Given CO2 from Industry and cement production in China accounts for more than 60% of world total industrial CO2, and the fact that uncertainty of emission in China is t Uncertainty from monthly statistics was derived from 10000 Monte Carlo simulations to estimate a 68% confidence interval (1-sigma) for China. from monthly statistics was derived from 10000 Monte Carlo simulations to estimate a 68% confidence interval (1-sigma) for China. We calculated the 68% prediction interval of linear regression models between emissions estimated from monthly statistics and official emissions obtained from annual statistics at the end of each year, to deduce the one-sigma uncertainty involved when using monthly data to represent the whole year's change. The squared correlation coefficients are within the range of 0.88 (e.g., coal production) and 0.98 (e.g., energy import and export data), which represent that only using the monthly data can explain 88% to 98% of the whole year's variation 37 

Where and are the percentage uncertainties and the uncertain quantities (daily mean emissions) of sector respectively.

Eq. (22) is used to derive for the uncertainty of the multiplication, which is used to combine the uncertainties of all sectors and of the projected emissions in 2019: 

All data generated or analyzed during this study are included in this article.

The code generated during and/or analyzed during the current study are available from the corresponding author. After peer-reviewed the code will be open accessible on the Carbon Monitor website (www.carbonmonitor.org or www.carbonmonitor.org.cn).

EDGAR v4. 3.2 Global Atlas of the three major Greenhouse Gas Emissions for the period 1970-2012

Carbon dioxide emissions from fossil fuels: a procedure for estimation and results for 1950-1982

Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories

Intergovernmental Panel on Climate Change (IPCC). 2006 IPCC Guidelines for National Greenhouse Gas Inventories. (Intergovernmental Panel on Climate Change

Carbon dioxide emissions from fossil fuels: A procedure for estimation and results for 1950-1982

China: Emissions pattern of the world leader in CO2 emissions from fossil fuel consumption and cement production

A new evaluation of the uncertainty associated with CDIAC estimates of fossil fuel carbon dioxide emission

Inventory of China's Energy-Related CO2 Emissions in 2008

Monthly, global emissions of carbon dioxide from fossil fuel consumption

CO2 Emissions from Fuel Combustion

Variations of China's emission estimates: response to uncertainties in energy statistics

High resolution temporal profiles in the Emissions Database for Global Atmospheric Research

Quantifying the uncertainties of China's emission inventory for industrial sources: From national to provincial and city scales

A synthesis of carbon dioxide emissions from fossil-fuel combustion

Trends in global CO2 emissions: 2013 report

Reduced carbon emission estimates from fossil fuel combustion and cement production in China

Trends in global CO2 and total greenhouse gas emissions

Gridded uncertainty in fossil fuel carbon dioxide emission maps, a CDIAC example

A multiyear, global gridded fossil fuel CO2 emission data product: Evaluation and analysis of results

Global carbon budget 2019

The Open-source Data Inventory for Anthropogenic CO2, version 2016 (ODIAC2016): a global monthly fossil fuel CO2 gridded emissions data product for tracer transport simulations and surface flux inversions

IPCC guidelines for national greenhouse gas inventories

A method for estimating the temporal and spatial patterns of carbon dioxide emissions from national fossil-fuel consumption

China's non-fossil fuel CO2 emissions from industrial processes

Fossil CO2 and GHG emissions of all world countries

Statistical Communiqué of the People's Republic of China on the National Economic and Social Development

Reduced carbon emission estimates from fossil fuel combustion and cement production in China

ICCT CO2 commercl aviation

Third IMO Greenhouse Gas Study

ERA5: Fifth generation of ECMWF atmospheric reanalyses of the global climate

Fossil CO2 and GHG emissions of all world countries

Taking advantage of the improved availability of census data: a first look at the gridded population of the world, version 4

New dynamics of energy use and CO2 emissions in China

Los-angeles, New-york, San-francisco, San-jose, Seattle, Miami, Chicago, Washington, Honolulu, Atlanta, Baton-rouge, San-diego, Boston, Austin, Portland, Philadelphia, Sacramento, Houston, Riverside, Tampa, Nashville, Orlando, Charleston, Denver, Cape-coral-fort-myers, Pittsburgh, New-orleans, Las-vegas, Boise, Fresno, Baltimore, Tucson, Providence, Charlotte, Dallas-fortworth, Oxnard-thousand-oaks-ventura, Bakersfield, Greenville, Jacksonville, Detroit, Albuquerque, Columbus, San-antonio, Salt-lake-city, Phoenix, Mcallen, Raleigh, Virginia-beach, Hartford, Colorado-springs, Birmingham, New-haven, Louisville, Minneapolis, Cincinnati, El-paso, Allentown, Buffalo, Memphis, Worcester, Grand-rapids, Albany, St-louis, Milwaukee, Omaha-council-bluffs, Indianapolis, Rochester, Columbia, Oklahoma-city, Cleveland, Tulsa, Kansascity, Knoxville, Richmond, Winston-salem, Dayton, Little-rock, Syracuse, Akron, Greensboro-high-point

Authors declare no conflict of interests.

Zhu Liu and Philippe Ciais designed the research, Zhu Deng coordinated the data processing. Zhu Liu, Philippe Ciais and Zhu Deng contributed equally in this research, all authors contributed to data collection, analysis and paper writing.