High-Spatial-Resolution Methane Emissions Calculation Using TROPOMI Data by a Divergence Method

Abstract

Methane (CH₄) is the second-largest greenhouse gas emitted by human activity and natural sources after carbon dioxide (CO₂). Its relatively short lifetime in the atmosphere (about 12 years) means that we can mitigate the human impacts of climate change in a relatively short period of time by reducing CH₄ emissions. The creation of CH₄ emissions management policies can be based on the distribution maps of surface CH₄ concentration that are in large-scale and at high-resolution. The estimate of CH₄ emissions with broad coverage are provided by currently extensively used satellite data supplemented with data from model simulations. However, it is at low spatial resolution. In this paper, through the combination of atmospheric CH₄ observations from the TROPOMI sensor and wind data from the ECMWF global reanalysis, a straightforward divergence method is proposed to estimate the surface CH₄ emissions in China from March 2019 to September 2022 at a resolution of 7 km × 7 km. This method was compared with the average annual CH₄ emissions of Emissions Database for Global Atmospheric Research (EDGARv7.0), and the Root Mean Square Error (RMSE) is 2.53 kg/km²/h and within error envelop (EE) is 72.93%, which represents the proportion of reliable values under certain uncertain conditions. We estimated that the average annual CH₄ emissions in China from 2019 to 2022 is 81 Tg, with the lowest emissions in 2021 (75 Tg) due to the impact of COVID-19. In 2021, the largest anthropogenic emissions in China are from agriculture, energy activities and livestock, accounting for 28% (20.8 Tg), 25% (18.9 Tg) and 19% (13.9 Tg) of total emissions, respectively, while wetlands, as the largest natural source, produce 14% (10.5 Tg) of CH₄ emissions.

1. Introduction

Methane (CH₄), which makes up more than 20% of all greenhouse gases (GHGs) from both natural and anthropogenic sources, is the second most prevalent GHGs after carbon dioxide (CO₂) [1,2]. The World Meteorological Organization’s most recent study of observation data from the Global Atmospheric Watch Program reveals that atmospheric CH₄ concentration has climbed from 722 ppb (part per billion) in 1750 to 1908 ± 2 ppb in 2021, an increase of 18 ppb with respect to the previous year and the record is constantly increasing [3]. In addition, the radiative forcing from atmospheric CH₄ is 0.48 ± 0.05 W m⁻² and makes up 17% of the global radiative forcing of GHGs [4]. Due to significant global warming potential and brief atmospheric lifetime (approximately 12 years) of CH₄, reducing its emissions can partially alleviate the effects of climate change on humans in a short amount of time. China is one of the largest CH₄ emissions in the world, and will actively reduce CH₄ emissions in the next decade [5]. Plans for reducing CH₄ emissions will be more effective if they take into account the precise temporal and spatial variations of surface emissions.

Major sources of CH₄ emissions in China include agriculture, energy activities, livestock, waste and wetlands. To effectively and accurately monitor CH₄ emissions, we need detailed analysis and monitoring of major sources. Agriculture includes agricultural waste burning and agricultural soils. Agricultural waste burning is mainly the burning of straw of rice, corn and other crops, while the water content of agricultural waste is generally high, burning agricultural waste will also produce a large amount of CH₄. CH₄ emissions from agriculture are mainly caused by rice paddy soil because of warm, moist paddy soils providing ideal conditions for CH₄ production [6]. Energy activities include industry, transport, energy for buildings, fuel exploitation [7], iron and steel production, etc. CH₄ is a large part of the world’s energy, used to heat homes, cook food, heat water, generate electricity and even power transportation. Industries that contribute to anthropogenic CH₄ emissions are mainly the Oil refineries and transformation industry, power plants, manufacturing plants and steel mills, whose combustion converts CH₄ energy into carbon dioxide for release into the atmosphere. Cattle contribute the majority of greenhouse gas emissions from Enteric fermentation, followed by other ruminants such as buffalo, sheep and goats. Enteric fermentation occurs when anaerobic microorganisms, called methanogens, break down and ferment food present in the animal’s digestive tract to produce compounds that are then absorbed by the animal host. This digestive process allows ruminants to eat more plant material that would otherwise be indigestible. About 7–10% of ruminants’ energy is lost in the form of CH₄ and excreted by enteric fermentation through burping [8]. In 2020, the waste sector, municipal solid waste and wastewater, accounted for nearly 20 percent of all human-related CH₄ emissions, the third largest global source of CH₄ emissions after enteric fermentation and energy activities [9]. Wetlands are one of the largest natural sources of CH₄-producing emissions globally, which are estimated to make up 20% to 30% of the global CH₄ emissions budget [10]. In wetlands, CH₄ is produced by microbial activity and can be transported from the soil to the atmosphere through a variety of methods. According to the government data of the National Forestry and Grassland Administration, China’s wetland area is about 56.35 million hectares, accounting for 4% of the world’s wetland area. Thirteen cities in China have been elected to the Convention on Wetlands, making it the country with the largest number of international wetland cities in the world. The NASA Carbon Monitoring System (CMS) program describes, quantifies, understands and predicts the evolution of global carbon sources and sinks through satellite observations and modeling analysis capabilities, providing estimates of monthly CH₄ emissions from global wetlands.

The two primary techniques used to measure atmospheric CH₄ concentration currently are bottom-up ground-based methods [11,12] and top-down satellite platforms [13,14,15,16,17]. The advantage of satellite-based monitoring is that it can identify the characteristics of temporal and spatial CH₄ concentration variations on a local, regional, and global scale. The Atmospheric Infrared Sounder (AIRS) on EOS/Aqua satellite has been used to analyze the distribution of CH₄ in the middle and upper troposphere of China (MUT-CH4) from 2003 to 2008 [14]. The results show that the growth in China is generally more significant than in other parts of the world. Based on CH₄ vertical columns measured by the Scanning Imaging Absorption spectroMeter for Atmospheric CHartographY (SCIAMACHY) onboard Envisat, Zhao et al. presented temporal and spatial variations of the three-order trapezoid spatial gradient of CH₄ in China from 2003 to 2005 (0.5° latitude × 0.5° longitude) [18]. Miller et al. used CH₄ observations from the Greenhouse gases Observing SATellite (GOSAT) for aggregate 2.0° × 2.5° latitude-longitude grid boxes between September 2009 and September 2015 to show the latest trends in Asia’s total anthropogenic and natural emissions, with a particular focus on China, largely attributed to coal mining [19]. Qin et al. obtained the spatial-temporal characteristics of seasonal variation of atmospheric CH₄ concentration over China from 2003 to 2021 using the CH₄ column Homogeneous Dry Air Mole Fraction (XCH₄) product of SCIAMACHY, GOSAT and the Sentinel 5 Precursor (S5P) [20]. Since its October 2017 launch, the TROPOspheric Monitoring Instrument (TROPOMI) on board the S5P satellite has measured CH₄ concentrations at an unprecedented 7 km × 7 km resolution (upgraded to 5.5 km × 7 km in August 2019), outperforming previous widely used instruments such as GOSAT and SCIAMACHY. Chen et al. quantify CH₄ emissions in China and the contributions from different sectors by inverse analysis of 2019 TROPOMI satellite observations of atmospheric CH₄ at 0.25° × 0.3125° resolution [21].

However, previous studies have shown that (1) the spatial resolution of previous satellite data is too low, resulting in low coverage of CH₄ emissions data; (2) The model’s simulation data and satellite data are combined to cover the entire study area, but the spatial resolution of the results is lower than that of the satellite itself. In this study, a straightforward divergence method is used successfully to estimate surface CH₄ emissions from TROPOMI satellite data, and the spatial resolution is consistent with the satellite data.

2. Data and Methodology

2.1. Data

2.1.1. Surface Atmospheric XCH₄ from TROPOMI

The S5P is the first of the atmospheric composition Sentinels, launched on 13 October 2017. For this study, we used S5P OFFO(L2) CH₄ products from March 2019 to September 2022 with a spatial resolution of 7 km × 7 km, as shown in

Table 1. Description of data used in the paper.

Dataset	Parameter(s)	Resolution	Time Range	Download Link
TROPOMI CH4	Priori CH4 profile	7 km × 7 km	2019.03–2022.09	https://s5phub.copernicus.eu/dhus/#/home accessed on 8 October 2022.
	Dry air columns
ECMWF Wind	10 m u-component of wind	0.25° × 0.25°	2019.03–2022.09	https://cds.climate.copernicus.eu/ accessed on 1 October 2022.
	10 m v-component of wind
EDGAR CH4	CH4 emissions	0.1° × 0.1°	2019–2021	https://edgar.jrc.ec.europa.eu/ accessed on 1 August 2022.
WetCHARTs v1.3.1	Wetlands CH4 emissions	0.5° × 0.5°	2019	https://doi.org/10.3334/ORNLDAAC/1915 accessed on 1 October 2022.

. Focusing on the XCH₄ bias corrected product, the accuracy offered in the data product (defined as the standard deviation of the retrieval noise, which describes the impact of the measurement noise on the retrieval) and the quality descriptor (“qa_value”) is suitable for data use. The “qa_value” indicates the status and quality of the retrieval output. To assure that the highest quality data is used, pixels that are classified with qa_value < 0.5 were filtered to ensure data quality. The entire atmospheric column was divided into 12 layers in the TROPOMI XCH₄ retrieval, and the lowest layer is used to represent the surface atmospheric column concentration. We used the mole content of CH₄ in atmosphere bottom layer in INPUT_DATA divided by the mole content of dry air in atmosphere bottom layer to obtain the surface column averaged dry air mixing ratio of CH₄ (surface atmospheric XCH₄).

2.1.2. Wind Data

ERA5 is the fifth generation ECMWF reanalysis for global climate and weather, which uses the laws of physics to combine model data with observations from across the world to form a globally complete and consistent dataset [22]. ERA5 has been re-gridded to a regular latitude-longitude grid of 0.25 degrees for the reanalysis. Eastward wind, 10 U parameter from ECMWF (grib variable 165), is the horizontal component of the wind at 10-meter height in the eastward direction. Northward wind, 10 V parameter from ECMWF (grib variable 166), is the horizontal component of the wind at 10-meter height in the northward direction, whose data format is NC FLOAT. The TROPOMI orbit timestamp, which is generally one hour before the Riyadh overpass, was interpolated to using a linear interpolation from the six-hourly output.

2.1.3. CH₄ Emissions from EDGAR

The gridded CH₄ emissions database in China was from the Emissions Database for Global Atmospheric Research (EDGAR), which is a widely used inventory for anthropogenic emissions of greenhouse gases on Earth. EDGAR provides CH₄ emission estimation from sector-specific grid maps at 0.1° × 0.1° resolution with yearly, monthly, and hourly data using the IPCC methodology [23,24]. This study collected data on yearly anthropogenic CH₄ emissions from all sectors for 2019–2021, and we divided them into Agricultural, energy activities, livestock and Waste.

2.1.4. Wetland CH₄ Emissions

The CMS program aims to make a significant contribution to characterizing, quantifying, understanding, and predicting the evolution of global carbon sources and sinks by improving the monitoring of carbon storage and fluxes. WetCHARTs v1.3.1 provides global monthly wetland CH₄ emissions estimates at 0.5 by 0.5-degree resolution for the period 2001–2019, whose model drivers are replaced from using ERA-interim to ERA5 reanalysis data [25]. This product is the intended use is as a process of information wetlands CH₄ emissions data set, used for atmospheric chemistry and transport modeling.

2.2. Methodology

2.2.1. Surface Atmospheric XCH₄ Model

The TROPOMI CH₄ data product is given in the form of total column-averaged dry-air mole fraction, XCH₄. It is calculated from the CH₄ vertical sub-column elements $x_i$ and the dry-air column $V_{air,dry}$ calculated with meteorology input from ECMWF:

$${\mathrm{XCH}}_4={\displaystyle \sum }_{i=0}^n\frac{x_i}{V_{air,dry}}$$

(1)

The entire atmospheric column was divided into 12 layers in the TROPOMI CH₄ profile data and the bottom layer represents the concentration of the surface atmospheric column, whose average height is about 700 m. In order to obtain surface XCH₄ quickly, we obtain atmospheric CH₄ a priori vertical profile data and dry air density column from TROPOMI official supporting data. To assure that the highest quality data is used, pixels that are classified with qa_value < 0.5 were filtered to ensure data quality. XCH₄ in surface is derived as follow:

$${\mathrm{XCH}}_4=\frac{V_{C{H}_4}}{V_{air}}$$

(2)

$V_{C{H}_4}$ and $V_{air}$ stand for the bottom layer of the TROPOMI CH₄ profile data and dry air density used by the retrieval of TROPOMI XCH₄, respectively.

2.2.2. A Divergence Method to Quantify CH₄ Emissions

The divergence can be used to describe the degree of divergence of vector field at different points in space. According to the continuity equation for steady state, the divergence (D), emission (E) and sink (S) can be expressed as: D = E + S [26]. Because CH₄ is a long-lived gas that can exist in the atmosphere for 12.5 years, there is atmospheric CH₄ emission and background at the surface. The sink term can be ignored and the background (B) concentration is assumed to be completely homogeneous [27]. As D is a linear operator, the relationship between D, E and B can be expressed as: E = D − B and the average emissions over time can be calculated.

Here, we extract top-down CH₄ emission maps on high spatial resolution based on TROPOMI CH₄ columns V combined with wind fields $\omega$ from the European Center for Medium Range Weather Forecasts (ECMWF). The CH₄ fluxes is given as $\mathrm{F}=\mathrm{V}\omega$ and the divergence D works on horizontal fluxes (F): $\mathrm{D}=\nabla F$, thus $\mathrm{E}=\nabla F=\nabla \left(\mathrm{V}\overrightarrow{\omega }\right)$, where $\overrightarrow{\omega }$ is divided into vector eastward wind ($\overrightarrow{u}$) and northward wind ($\overrightarrow{v}$) at 10 m height.

In this study, numerical derivatives for D were calculated as the second-order central difference. For each grid cell, its neighboring grid cells along south-north ($\overrightarrow{v}$ component) and east-west ($\overrightarrow{u}$ component) directions are firstly used to obtain the divergence. However, it is hard to know the exact background, so we use the regional background to approximate the background of surface XCH₄, that is defined as the average of the lower 10 percentile of its surrounding ±5 grid cells (11 × 11 = 121 grid cells in total by taking the current grid cell as the center). So, the daily variations of CH₄ emission can be written as:

$$\mathrm{E}=\nabla \left(\left(\mathrm{V} - \mathrm{B}\right)\overrightarrow{\omega }\right)$$

(3)

where V stands for column of CH₄ in the surface and B stands for divergence of the regional background.

2.2.3. Validation by EDGAR Anthropogenic CH₄ Emissions

To evaluate the feasibility of our approach, anthropogenic CH₄ emission data from EDGARv7.0 were used to validate estimated CH₄ emission data we calculated through the divergence method. EDGAR reports on the evolution of national CH₄ emission inventories around the world over time and provides a 0.1° × 0.1° grid representing CH₄ emission sources. EDGARv7.0 provides CH₄ emissions per sector and country from 1970 to 2021 and grid maps are expressed in kg substance /km²/h for the network Common Data Form (NetCDF) files.

Using the modified divergence method, daily CH₄ emissions in 2021 were estimated, and their annual mean values were compared with EDGAR’s 2021 annual emissions, totaling 10,709 measurements. Due to the complex terrain, we choose the eastern plain as the verification target. Figure 1 shows the scatter plot between our estimated CH₄ emissions and EDGAR’s anthropogenic CH₄ emissions, with the root mean square error (RMSE) of 2.53 kg/km²/h and within error envelop (EE) of 72.93%, indicating that our annual estimation is slightly higher than EDGAR’s due to uncounted anthropogenic and natural CH₄ emissions in EDGAR’s emissions and the error of satellite-based background concentration.

3. Results

3.1. Spatial Distribution Patterns of Surface Atmospheric XCH₄ across China

In order to analyze the surface spatial distribution patterns of XCH₄ in China, the surface spatial distribution of multi-year-averaged XCH₄ concentrations from S5p from January 2019 to September 2022 is shown in Figure 2. The calculation details are provided in Section 2.2.1. The range and average of XCH₄ concentrations from S5P were 1900 to 2150, and 1998 ppb. The XCH₄ concentrations showed greatly spatial variation in different regions of China and presented spatial cluster in different regions. Overall, there were low XCH₄ concentrations in southwest China, high concentrations in east China and moderate concentrations in the north. Because CH₄ is a long-lived gas that can exist in the atmosphere for 12.5 years, the spatial variations of atmospheric CH₄ concentrations might reflect the accumulation of CH₄ emissions and transportations.

According to the available data (from Jan. 2019 to Sept. 2022, qa_value > 0.5), the high concentration areas in China mainly include central China, East China, the Beijing-Tianjin-Hebei region, the Sichuan Basin and Part of the northern part of Xinjiang Uygur Autonomous Region, with an average concentration of 2064 ppb, which is 66 ppb higher than the China’s average; The areas with consistently low concentrations are Tibet and Qinghai Province, with an average concentration of 1934 ppb, which are 64 ppb lower than the China’s average; The rest of China was in the middle, with an average of 1975 ppb, 23 ppb lower than the China’s average.

3.2. Analysis of Divergence Method of Calculating CH₄ Emissions in China

In order to analyze the spatial and temporal distribution of CH₄ emissions in China, the divergence method is used to calculate China’s CH₄ emissions through TROPOMI surface XCH₄ concentrations from March 2019 to September 2022, as shown in Figure 3. Excluding the missing data, the regions with high emission rates are the Tibetan Plateau, parts of Yunnan, parts of northern Xinjiang, the Sichuan Basin, central China, eastern China, and the three provinces in northeast China. China’s northwest and west region has virtually no CH₄ emissions, according to EDGAR CH₄ emissions data. Although these areas do not have significant anthropogenic CH₄ emissions, they have a relatively high surface CH₄ concentration levels and relatively high wind speeds and change of wind direction throughout the years (Figure 3). In the Tibetan Plateau, the daily amount of TROPOMI secondary data is sufficient, but it is very sparse with a large number of vacancy values. The divergence method is used to estimate CH₄ emissions on a daily basis, which leads to the abnormal divergence (vacancy value is set to zero, resulting in the value points identified as high divergence points). Similarly, in Yunnan Province, not only is the daily data very sparse, but also the amount of data is very small, resulting in high divergence. In the hilly areas of southern China, there are also large data missing, resulting in the loss of computing continuity. In addition, in areas with large altitude changes, such as Kunlun Mountains, Qilian mountains, Hengduan Mountains, Greater Khingan Mountains and hilly areas in southern China, the data quality is poor, resulting in abnormal divergence calculation (positive or negative values are both too high). Wind speed and wind direction are also the most important factors affecting divergence. Large divergence errors will be caused at the interfaces of different terranes (where there are high wind speeds in opposite directions), such as China’s First Ladder and the second ladder, as well as China’s second ladder and the third ladder (Kunlun Mountains and Greater Khingan Mountains, which typically connect the Qinghai-Tibet Plateau and the Tarim Basin).

3.3. CH₄ Emissions in China

Due to missing data, for each grid cell, the divergence is calculated only when there are valid values along the north-south and east-west directions. Excessive wind speed will affect the conservation of mass in the region, so the wind speed is set to be less than 10 m/s. A negative value represents a reduction in CH₄ concentration in the area, indicating that more CH₄ was transported than emitted over a period of years.

Based on daily TROPOMI observation from March 2019 to September 2022, we estimated multi-year averages CH₄ emissions in the Chinese region with a resolution of 7 km × 7 km. The calculation details are provided in Section 2.2.2. When the divergence of CH₄ is obtained, CH₄ emissions can be obtained by averaging them over time. As shown in Figure 4a,b, TROPOMI observations are mainly distributed in the northern, central and eastern regions of China, where CH₄ emissions can be better estimated. In addition, in central, eastern and northern China, where a small amount of complex terrain exists, daily data volumes are plentiful and of higher quality than in other regions, resulting in smoother and more robust performance on averages. Since the daily data of the Qinghai-Tibet Plateau is very sparse and there is almost no anthropogenic CH₄ emissions, we choose to mask this part. The accumulation of divergence over time manifests as local CH₄ emissions. We estimated that the average annual CH₄ emissions in China is 81 Tg y⁻¹, which is higher to EDGARv7.0 average annual CH₄ anthropogenic emissions from 2019 to 2021 (66 Tg y⁻¹), which can be due to the uncertainty of satellite observation and the existence of natural sources. The average annual CH₄ emissions in China from 2019 to 2022 are 78 Tg, 87 Tg, 75 Tg, 85 Tg respectively, and the emissions in 2021 due to the epidemic were less than the previous year.

To analyze the differences in CH₄ emissions between northern and southern China, we compared the CH₄ emissions characteristics of the Beijing-Tianjin-Hebei region and Jiangsu Province in 2021. The terrain in the Beijing-Tianjin-Hebei region is very different in height, being high in the northwest and low in the southeast. Figure 4c shows that the distribution characteristics of CH₄ emissions in the Beijing-Tianjin-Hebei region are completely different from those in Jiangsu Province. The ratio of plateau, plain, and mountainous land in Hebei Province is 1:4:5, which blocks the air flow in the west and north and makes it difficult for haze pollution to spread. Due to the influence of mountainous terrain, a large number of data gaps exist in the northern part of Hebei Province. Hebei province is rich in mineral resources, with large numbers of coal mines and coal-fired power plants in the cities of Shijiazhuang, Xingtai, and Handan, causing a lot of pollution. Jiangsu Province is abundant in water resources and has the highest percentage of plain land in China, up to 86.9%. Since paddy fields are the primary crop, agriculture accounts for the majority of Jiangsu Province’s CH₄ emissions. Two significant freshwater lakes in Jiangsu Province, Taihu Lake and Hongze Lake, have substantial CH₄ emissions levels close by, which may be brought on by the interaction of rice fields and natural wetlands. In 2021, We estimated that Hebei and Jiangsu Province emitted 6.2 Tg and 3.8 Tg of CH₄ annually, respectively. CH₄ emissions from energy activities in Hebei Province are higher than those from paddy fields and natural wetlands in Jiangsu Province.

EDGARv7.0 average annual CH₄ anthropogenic emission was divided into 22 sectors in 2021, which we summarize into four sectors, including agriculture, energy activities, livestock andwaste, as shown in Figure 5a–d.

The Sichuan Basin, southern and northeastern China are large areas where rice is grown. Soils in the south generally contain more moisture than those in the north, and the entire Yangtze River basin, in particular, is the worst emitter of CH₄. The most serious areas of agriculture waste burning are southern Guangdong Province, Guangxi Zhuang Autonomous Region and Taiwan Province. Our estimate for agriculture in 2021 is about 20.8 Tg, the highest anthropogenic CH₄ emissions of any industry. Anthropogenic CH₄ emissions from energy for buildings, like those from most factories, are distributed in and around provincial capitals, largely because of increasing urbanization. CH₄ emissions from road transport such as rail and road transport occur on major national and provincial highways across the country. Important fuel exploitation activities caused more than 80 percent of CH₄ emissions in Shanxi Province, with a smaller amount in Henan and Shandong provinces. Anthropogenic CH₄ emissions from energy activities have a characteristic national distribution, totaling about 18.9 Tg in 2021, but are mainly concentrated in central and eastern China. Nationwide, CH₄ emissions from enteric fermentation are mainly distributed in Jiangsu, Anhui, Shandong and Henan regions, followed by southwest China, and a small amount of CH₄ emissions exists in northeast China, with a total estimated CH₄ emission of 13.9 Tg in 2021. In China, CH₄ emissions generated by waste in all cities are estimated to be 11.7 Tg y⁻¹. Due to China’s large population and increasing purchasing power, a large amount of consumption-related waste is accumulated in major cities across the country, especially in super-large cities such as Beijing, Shanghai, Guangzhou and Chengdu. According to the data, the estimated amount of China’s wetland CH₄ emissions is 10.5 Tg y⁻¹, mainly distributed in central China, including Hubei, Anhui and Jiangxi provinces, as shown in Figure 5e.

4. Conclusions

CH₄ is the second most important greenhouse gas causing greenhouse effect, with a short lifespan of about 12 years. Therefore, effective monitoring of CH₄ emissions can provide a basis for the government’s carbon emission reduction policies. In order to obtain high resolution surface CH₄ emission data in China, a simple divergence method is proposed in this paper. Firstly, the surface CH₄ column concentration can be obtained by combining the CH₄ profile data of TROPOMI with the dry air density column. The precise background concentration was then replaced with a regional CH₄ concentration of less than 10 percent of XCH₄ in a regional grid; Finally, using the surface CH₄ column concentration of TROPOMI and wind data of ECMWF, the divergence method is used to estimate surface CH₄ emissions.

CH₄ emissions calculated by divergence method were evaluated against CH₄ emissions data from EDGAR, showing a good result in the overall estimation. However, there is a phenomenon of high CH₄ concentration but few emissions in complex topographic areas, so the estimation of this method is a certain overestimated.

The surface CH₄ column concentrations distribution shows that the high concentration of CH₄ is in eastern China and the low concentration is in southwest China. Based on the divergence method, the average annual surface estimated CH₄ emissions in China from 2019 to 2022 were 78 Tg, 87 Tg, 75 Tg and 85 Tg, respectively. Affected by the pandemic, the average annual estimated CH₄ emissions in 2021 were significantly reduced. In 2021, agriculture, energy activities and livestock are the three sectors with the largest anthropogenic estimated CH₄ emissions in China, which emit 20.8 Tg, 18.9 Tg and 13.9 Tg respectively, while wetland, the largest natural source of CH₄ emissions, releases 10.5 Tg, accounting for 14% of the total estimated emissions.

The advantage of the method proposed in this paper is that the high resolution and wide coverage data of TROPOMI can be used to estimate surface CH₄ emissions in a large range. However, CH₄ emissions are overestimated in areas with sparse TROPOMI data and areas with complex terrain, so we will improve this method in the future.