An Analysis of the Spatial-Temporal Characteristics and Regulatory Strategies Pertaining to CH₄ Emissions in China from 2000 to 2023

Abstract

Methane (CH₄), the second-largest global greenhouse gas and a key driver of tropospheric ozone formation, critically influences climate change and air quality. As the world’s largest CH₄ emitter, China must develop targeted mitigation strategies to support its carbon peak and neutrality goals while reducing ozone pollution. Here, we analyzed the spatiotemporal evolution of provincial CH₄ emissions in China from 2000 to 2023 using spatial autocorrelation, hotspot detection, trend analysis, and K-means clustering. Our results revealed a triphasic emission trajectory—rapid growth followed by stabilization and a recent resurgence—with all provinces except Tibet showing increasing trends. The energy sector emerged as the primary contributor, particularly in Inner Mongolia, Shanxi, and Shaanxi, whereas agricultural emissions dominated in pastoral regions, such as Inner Mongolia and Sichuan, and rice-growing areas, such as Hunan and Hubei. Coastal provinces, including Shandong, Jiangsu, and Guangdong, exhibited waste disposal as their predominant CH₄ source. Based on these patterns, we classified the emission zones into four distinct typologies: coal-dominant, waste-dominant, oil-agriculture composite, and multifactorial systems, proposing tailored mitigation frameworks that integrate CH₄ and ozone co-reduction. This study provides a spatially resolved foundation for synergistic climate and air quality governance in China.

1. Introduction

Methane (CH₄) represents the second most significant anthropogenic greenhouse gas, contributing 16% of total radiative forcing. While its atmospheric lifetime is shorter than carbon dioxide (CO₂), CH₄ possesses a 20-year global warming potential 82 times greater—making it a critical driver of near-term climate change [1]. Beyond its direct radiative effects, CH₄ oxidation fuels tropospheric ozone production, creating cascading impacts on air quality, human health, and agricultural productivity [2,3]. Approximately 60% of global CH₄ emissions originate from human activities, with dominant sources spanning enteric fermentation, fossil fuel systems, and organic waste decomposition [4]. The IPCC AR6 quantifies CH₄’s substantial climate influence, attributing 0.5 °C of the observed 1.07 °C anthropogenic warming to this potent gas from 2010 to 2019 [1]. Without immediate mitigation, projected CH₄ growth threatens to compromise the 1.5 °C Paris Agreement threshold, positioning CH₄ reduction as both a climate imperative and public health opportunity [5].

As a global leader in CH₄ emissions, China has progressively strengthened its climate governance framework since establishing its first CH₄ mitigation targets in 2007. Through successive Five-Year Plans spanning 2011–2025, the nation has systematically implemented emission control measures, culminating in the 2023 CH₄ Emission Control Action Plan that established comprehensive mitigation pathways. This policy evolution has been accompanied by continuous refinement of regulatory standards and active engagement in international climate cooperation. Official reports reveal that China’s 2020 CH₄ emissions reached 60.4 million tons CO₂-eq, with the energy sector contributing nearly half of total emissions at 46.2%, followed by agriculture at 39.3% and waste management at 12.2% [6]. Despite these advancements, achieving carbon neutrality requires overcoming persistent challenges including technological limitations, policy gaps, and financing constraints. Addressing these barriers requires a precise understanding of emission spatial-temporal patterns and the development of sector-specific mitigation strategies—critical steps for simultaneously advancing climate goals, supporting the carbon peak and neutrality goals, and reducing ozone-related health impacts.

Accurate quantification of CH₄ emissions and their sectoral contributions is critical for tracking temporal trends and evaluating mitigation efficacy. Leveraging multi-source datasets-including remote sensing (e.g., GOSAT satellite observations), ground-based monitoring (e.g., EDGAR), and emission inventories—researchers have systematically analyzed the spatiotemporal patterns and driving factors of China’s CH₄ emissions. For instance, Peng et al. [7]. and Guo et al. [8]. utilized C3S CH₄ concentration products and GOSAT near-surface data, respectively, to identify distinct spatial (higher in eastern regions, lower in western regions) and seasonal (peaks in summer/autumn, troughs in winter/spring) variations in CH₄ emissions. Similarly, Han et al. [9], using EDGAR data (2011–2018), demonstrated that fossil fuel extraction, agriculture, and waste management dominate China’s anthropogenic CH₄ sources. Complementary studies employing national greenhouse gas inventories, such as those by Li et al. [10] (following IPCC Inventory Guidelines), revealed persistent increases in oil/gas-related CH₄ emissions (2003–2021) in western provinces (e.g., Shaanxi, Xinjiang, Sichuan), contrasting with declining trends in northeastern and northern China. Recent research has shifted from foundational data analysis to applied domains, including emission projection, abatement technologies, and policy design [6]. Sector-specific studies have adopted diverse methodologies to dissect emission drivers [11] and mitigation pathways [12]. In the Energy Sector, Fu et al. [13] highlighted coal production as a key emission driver, while energy mix diversification (renewable adoption) and stringent industrial policies effectively curbed emissions. They advocated for integrated strategies, including enhanced coalbed CH₄ (CBM) tiered utilization technologies. In Agriculture and Waste Sectors, Research has predominantly focused on livestock management [14], reflecting its outsized role in CH₄ generation [15]. Growing evidence underscores CH₄ mitigation as a crucial strategy for reducing ground-level ozone pollution [16]. As a potent ozone precursor, CH₄ emission controls offer dual benefits—simultaneously decreasing radiative forcing and improving air quality [17]. Recent studies by Wang et al. [3] reveal striking spatial-temporal correlations: CH₄ and ozone hotspots consistently co-occur in urban clusters, with peak concentrations observed during summer and autumn months, contrasting with lower winter–spring levels.

Although existing studies have analyzed the spatiotemporal characteristics, influencing factors, and mitigation potential of CH₄ emissions in China, several limitations remain. First, in terms of temporal scale, most existing research focuses on data from 2018 or earlier, with only a few studies updated to 2021. There is a lack of dynamic analysis of CH₄ emissions since the proposal of China’s carbon peaking and carbon neutrality goals. Second, regarding spatial scale, existing research predominantly concentrates on specific regions (such as the Sichuan Basin [18] and Dongbei region [19]) or individual sectors like energy and agriculture. Systematic research on the spatiotemporal heterogeneity of CH₄ emissions across the country and differences in regional control strategies is still lacking. Third, from a research perspective, existing analyses mainly focus on the impact of natural factors on CH₄ emissions. In contrast, the driving mechanisms of social factors such as economic development levels remain largely unexplored. Based on the EDGAR emission inventory, this study employs a suite of methods including spatial autocorrelation analysis (Moran’s I), hotspot detection (Getis-Ord G_i*), Sen trend analysis, and K-Means clustering. It systematically investigates the spatiotemporal evolution patterns of anthropogenic CH₄ emissions in China from 2000 to 2023. The study also deeply dissects the emission characteristics of different sectors and identifies CH₄ emission partition types and their main features by integrating socio-economic driving factors. Finally, it proposes targeted CH₄ control pathways and prioritization strategies to provide decision-making support for improving the CH₄ control system and enhancing control efficiency.

2. Materials and Methods

2.1. Data Collection

2.1.1. CH₄ Emission Data

EDGAR https://edgar.jrc.ec.europa.eu/ (accessed on 26 May 2025) is a database developed by the European Commission to estimate anthropogenic greenhouse gas emissions. It employs bottom-up emission estimation models and adheres to the methodologies established by the Intergovernmental Panel on Climate Change (IPCC) to aggregate global anthropogenic emissions [20]. These emissions are calculated based on statistical data, emission factors, and are spatially distributed worldwide on a 0.1 × 0.1 grid [21]. EDGAR comprehensively addresses all reporting categories outlined in the IPCC (2006) guidelines [22], categorized by emission sector. The time series data is updated annually to incorporate the most recent data sources [23].

This research examined CH₄ emission data spanning from 2000 to 2023 across multiple sectors worldwide, in accordance with the guidelines established by the IPCC (2006) [22]. The data was sourced from the European Commission’s Joint Research Centre (JRC) EDGAR database (

Table A1. Sector names of CH₄ emission sources and their IPCC codes.

Source		Sector	IPCC Sector Code	Sector ID
Energy Activities	Fossil Fuel Combustion	Petroleum refining, solid fuel manufacturing, mobile sources (others), uncontrolled combustion and coal pile burning, others, processing, solid fuel transportation	1A1b + 1A1ci + 1A1cii + 1A5biii + 1B1b + 1B2aiii6 + 1B2biii3 + 1B1c	a11
	Transportation	Rail, other transport	1A3c + 1A3e	a21
		Aviation	1A3a_LTO + 1A3a_CRS + 1A3a_CDS + 1A3a_SPS	a22
		Water transport	1A3d	a23
		Road transport	1A3b_noRES	a24
Industrial Production Processes	Coal Mining Fugitive Emissions	Coal fuel extraction	1B1a	a31
	Oil and Gas System Fugitive Emissions	Oil & natural gas (venting, flaring)	1B2aiii2 + 1B2aiii3 + 1B2bi + 1B2bii	a41
	Other Production Processes	Manufacturing industries and construction	1A2	b11
		Iron & steel production, ferroalloy production	2C1 + 2C2	b12
		Chemical industry	2B	b13
		Main activity electricity and heat production	1A1a	b14
Agriculture	Rice Cultivation	Lime application, urea application, direct emissions from managed soils, rice cultivation	3C2 + 3C3 + 3C4 + 3C7	c11
	Enteric Fermentation	Enteric fermentation	3A1	c21
	Manure Management	Manure management	3A2	c31
LULUCF	Forest Conversion Carbon Emissions (Combustion Emissions)	Biomass burning in croplands	3C1b	d11
Waste Management	Solid Waste	Waste incineration and open burning	4C	e11
		Solid waste disposal, biological treatment of solid waste	4A + 4B	e12

).

2.1.2. Influencing Factors Data

Based on the current landscape of CH₄ emission sources categorized by sector and data availability, this study identifies eight factors influencing CH₄ emissions from various sectors and one factor indicative of social and economic development that have been selected for quantitative analysis.

Specifically, fossil energy consumption and freight turnover indicate energy activity levels and transportation intensity, respectively. Additionally, raw coal, crude oil, and natural gas outputs represent industrial production activities, reflecting the extraction of these resources. In the agricultural sector, the area dedicated to rice cultivation and the number of large livestock at the end of the year are utilized as indicators of agricultural production levels. Furthermore, domestic waste management practices include removing household garbage, which signifies urban solid waste disposal efforts. Given the correlation between regional economic development and human activities, the secondary industry’s added value has been chosen to represent the degree of regional development. The data utilized for this analysis primarily derives from the China Statistical Yearbook, China Energy Statistical Yearbook, and Provincial Energy Statistical Yearbooks (

Table 1. Potential influencing factors of CH₄ emissions.

Type	Name	Unit
Energy	Fossil energy consumption	104 tons of standard coal
	Freight turnover serves	104 persons
Industrial	raw coal production	kt
	Crude oil production	kt
	Natural gas production	kt
Agricultural	Rice Sown Area	103 hectares
	Large Livestock Year-end Inventory	104 heads
Waste	Domestic Waste Collection Volume	kt
Economic Development	Added value of the secondary industry	Yuan/person

).

2.2. Research Methods

2.2.1. Trend Analysis

This study utilized the Theil–Sen slope method to examine the spatial trends of CH₄ emissions in China from 2000 to 2023 [24]. The calculation formula for the Sen slope is as follows:

$$\mathsf{\beta }=\mathrm{M}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{n}\left({\displaystyle \frac{{\mathrm{x}}_{\mathrm{j}}-{\mathrm{x}}_{\mathrm{i}}}{\mathrm{j}-\mathrm{i}}}\right)$$

(1)

β represents the slope, where i and j denote the years. If β > 0, it indicates an increasing trend in CH₄ emissions within the time series data; if β < 0, it signifies a decreasing trend. The larger the absolute value of β, the more pronounced the trend.

The Mann–Kendall test is a non-parametric statistical method widely used to assess the significance of trends in long-term time series data. Its major advantages include the fact that it does not require the data to follow a normal distribution, does not assume linearity in the trend, and is robust against missing values and outliers. The test procedure is described as follows [25]:

For a time series X with data points indexed by i = 1, 2, …, n, the standardized test statistic Z is defined as:

$$Z=\left\{\begin{array}{ll}{\displaystyle \frac{\mathrm{S}}{\sqrt{\mathrm{v}\mathrm{a}\mathrm{r}\left(\mathrm{S}\right)}}}& (\mathrm{S}>0)\\ 0& (\mathrm{S}=0)\\ {\displaystyle \frac{\mathrm{S}+1}{\sqrt{\mathrm{v}\mathrm{a}\mathrm{r}\left(\mathrm{S}\right)}}}& (\mathrm{S}<0)\end{array}\right.$$

(2)

$$\mathrm{S}=\sum _{\mathrm{i}=1}^{\mathrm{n}-1}\sum _{\mathrm{j}=\mathrm{i}+1}^{\mathrm{n}}\mathrm{s}\mathrm{i}\mathrm{g}\mathrm{n}({\mathrm{x}}_{\mathrm{j}}-{\mathrm{x}}_{\mathrm{i}})$$

(3)

$$\mathrm{E}\left(\mathrm{S}\right)=0$$

(4)

$$\mathrm{v}\mathrm{a}\mathrm{r}(\mathrm{S})={\displaystyle \frac{\mathrm{n}(\mathrm{n}-1)(2\mathrm{n}+5)}{18}}$$

(5)

x_i and x_j represent the time series data, and n denotes the number of data points. When n > 8, the test statistic S approximately follows a normal distribution.

2.2.2. Spatial Autocorrelation

The Global Moran’s I statistic was used to evaluate spatial autocorrelation across the study region, measuring the degree of similarity between each geographic unit and its neighboring units. This method is widely applied to characterize the spatial structure of georeferenced data. In this study, it was employed to assess the spatial dependence and heterogeneity of CH₄ emissions among internal spatial units in China. The significance of spatial autocorrelation was determined based on the Moran’s I index, along with the corresponding z-score and p-value [25,26]. The formula is as follows:

$$\mathrm{I}={\displaystyle \frac{\left[\begin{array}{c}\mathrm{n}\sum _{\mathrm{i}=1}^{\mathrm{n}}\left({\mathrm{x}}_{\mathrm{i}}-\stackrel{\mathrm{‾}}{\mathrm{x}}\right)\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{W}}_{\mathrm{i}\mathrm{j}}\left({\mathrm{x}}_{\mathrm{j}}-\stackrel{\mathrm{‾}}{\mathrm{x}}\right)\end{array}\right]}{\left[\begin{array}{c}\sum _{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathrm{x}}_{\mathrm{i}}-\stackrel{\mathrm{‾}}{\mathrm{x}}\right)}^2\sum _{\mathrm{i}=1}^{\mathrm{n}}\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{W}}_{\mathrm{i}\mathrm{j}}\end{array}\right]}}$$

(6)

n represents the number of raster pixels of CH₄ emissions; i and j denote the CH₄ emissions of spatial units i and j, respectively; x is the mean value of CH₄ emissions for each raster pixel; and W_ij is the spatial weight matrix.

The Local Moran’ s I statistic, an extension of the Global Moran’ s I, was employed to identify local clustering patterns and spatial heterogeneity in China’s CH₄ emissions. This approach allowed detection of both high-value and low-value spatial clusters, highlighting regions with significant spatial dependence and variation in emission levels [27]. Statistical significance of local clustering was assessed based on the computed index, z-scores, and p-values. The formula for the local Moran’s index is computed as follows:

$${\mathrm{I}}_{\mathrm{i}}=\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{W}}_{\mathrm{i}\mathrm{j}}{\mathrm{x}}_{\mathrm{j}}/\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{x}}_{\mathrm{i}},$$

(7)

When I_i is greater than 0, it indicates a positive spatial correlation; when I_i is less than 0, it indicates a negative spatial correlation.

2.2.3. Spatial Hotspot Analysis

The Getis-Ord ${\mathrm{G}}_{\mathrm{i}}^{\ast }$ statistic, a spatial analysis method based on spatial statistics, was employed to identify clustering patterns and spatial distribution characteristics within the dataset. This approach was used to assess methane emission hotspots across China, visually highlighting regions where values were significantly higher or lower than expected on a large-scale map, thereby revealing spatial patterns of methane emissions [28,29]. The formula utilized for calculating high-value clusters (High) and low-value clusters (Low) of CH₄ in China is computed as follows:

$$\begin{array}{rrr}& {\mathrm{G}}_{\mathrm{i}}^{\ast }& ={\displaystyle \frac{\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{w}}_{\mathrm{i},\mathrm{j}}{\mathrm{x}}_{\mathrm{j}}-\overline{\mathrm{X}}\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{w}}_{\mathrm{i},\mathrm{j}}}{\mathrm{S}\sqrt{\frac{\left[\begin{array}{c}\mathrm{n}\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{w}}_{\mathrm{i},\mathrm{j}}^2-{\left(\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{w}}_{\mathrm{i},\mathrm{j}}\right)}^2\end{array}\right]}{\mathrm{n}-1}}}},\\ & & \overline{\mathrm{X}}={\displaystyle \frac{\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{x}}_{\mathrm{j}}}{\mathrm{n}}},\\ & & \mathrm{S}=\sqrt{{\displaystyle \frac{\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{x}}_{\mathrm{j}}^2}{\mathrm{n}}}-{\left(\overline{\mathrm{X}}\right)}^2},\end{array}$$

(8)

W_i,j represents the spatial relationship between i and j, corresponding to the elements of the spatial adjacency weight matrix; x_j is the distance expression of the spatial unit. When the calculated result ${\mathrm{G}}_{\mathrm{i}}^{\ast }$ > 0, the study area exhibits a high-value cluster, referred to as a “hot spot”; when ${\mathrm{G}}_{\mathrm{i}}^{\ast }$ < 0, the study area exhibits a low-value cluster, referred to as a “cold spot.” A Z-test is applied to perform statistical inference on the ${\mathrm{G}}_{\mathrm{i}}^{\ast }$ results. If the difference is not statistically significant, the study area is considered to exhibit a random distribution.

2.2.4. K-Means Analysis

K-means is an unsupervised machine learning algorithm based on iterative optimization, which employs Euclidean distance to measure the similarity between data samples for the purpose of clustering [30]. To further explore and identify the similar characteristics of CH₄ emissions and influencing factors across various provinces, K-means clustering algorithm has been employed to conduct a cluster analysis of CH₄ emissions alongside nine selected influencing factors across 30 provinces and municipalities in 2022. K-means algorithm categorizes data points into k distinct clusters, aiming to enhance the similarity among samples within each cluster while concurrently reducing the similarity between clusters [31].

The concrete process of K-means algorithm is as follows:

1.

Select the objects, which number is k, as the initial clustering centers;

2.

According to the value of clustering center, each object is assigned to the bunch which is the most similar with;

3.

The average value of each bunch is re-calculated, which is used a new clustering center;

4.

Repeat the steps (2), (3), until the clustering centers no longer vary and the average error function E has already restrained [32].

This study presents a maximum iteration coefficient designed to mitigate the issue of excessively slow convergence. The primary aim of the K-means algorithm is to minimize the Within-Cluster Sum of Squares (SWCS), which represents the sum of squared errors within the clusters, and is mathematically articulated as follows:

$${\mathrm{S}}_{\hspace{0.25em}\mathrm{W}\mathrm{C}\mathrm{S}}=\sum _{\mathrm{i}=1}^{\mathrm{k}}\sum _{{\mathrm{x}}_{\mathrm{j}}\in {\mathrm{C}}_{\mathrm{i}}}{∥{\mathrm{x}}_{\mathrm{j}}-{\mathsf{\mu }}_{\mathrm{i}}∥}^2$$

(9)

C_i denotes the ith cluster; μ_i represents the centroid of cluster C_i; xi indicates the data point within cluster C_i. The optimal assignment is determined through iterative optimization.

Selecting an appropriate k value is crucial. In this study, the Elbow Method was utilized to determine the optimal k by identifying the inflection point at which the Sum of Squares due to Error (SSE) demonstrates a notable decrease, commonly referred to as the “elbow” point [33,34].

3. Results

3.1. Spatiotemporal Distribution Characteristics of CH₄ Emissions

China’s CH₄ emissions exhibited an overall fluctuating upward trend from 2000 to 2023, with distinct phase-specific variations. Rapid Growth Phase (2000–2010): CH₄ emissions increased at an average annual rate of 2.1%, exceeding 50 Mt for the first time in 2008. This surge was primarily driven by rapid growth in emissions from the energy sector and waste management sector. Slowdown Phase (2011–2016): The growth rate declined to 1.4% per year, with emissions stabilizing around 55.12 Mt. During this period, emissions from the energy and agricultural sectors plateaued, while growth in the waste sector decelerated. Resurgence Phase (2017–2023): The annual growth rate rebounded to 1.9%, reaching a historic peak of 62.15 Mt in 2023. This resurgence was mainly attributed to a sharp rebound in energy sector emissions, continued growth in waste-related emissions, and an incremental rise in agricultural sector emissions. Notably, emissions from the industrial processes sector remained relatively stable throughout the study period (Figure 1). The observed phase-specific variations in CH₄ emissions were closely linked to China’s economic development cycles, energy structure optimization policies, and improvements in waste management systems.

The spatial analysis of China’s provincial CH₄ emissions revealed distinct geographical patterns across 2000 (Figure 2a), 2010 (Figure 2b), 2020 (Figure 2c), and 2023 (Figure 2d). The 31 provinces exhibited pronounced regional disparities, with emissions generally following an east-high, west-low gradient that gradually expanded northwestward over time. Energy-intensive provinces, including Shanxi, Inner Mongolia, and Shaanxi, consistently maintained the highest emission levels, reaching 8.07 Mt, 7.07 Mt, and 4.71 Mt, respectively, by 2023. Secondary emission hotspots emerged in Anhui, Shandong, and Hebei provinces, where emissions ranged between 3 and 3.5 Mt in 2023, whereas the lowest emission values (<0.3 Mt) persisted in southern and western regions, such as Hainan, Qinghai, and Tibet. This spatial distribution reflects the complex interplay between regional energy production activities, varying economic development patterns, and differential implementation of emission control policies across provinces. The observed northwestward shift in emission hotspots corresponds with the relocation of energy extraction activities, establishment of new industrial bases, and evolving regional environmental regulations, collectively shaping China’s dynamic CH₄ emission landscape over the study period.

Analysis of CH₄ emission trends using the Theil–Sen estimator (Figure 3) revealed significant spatiotemporal heterogeneity in emission patterns across China’s 31 provinces during 2000–2023 (p < 0.05). The Tibet Autonomous Region was the only province to demonstrate a declining trend in CH₄ emissions (Sen Slope < 0), whereas the remaining 30 provinces exhibited increasing trends (Sen Slope > 0). The most pronounced growth occurred in the northwestern provinces (Shanxi, Shaanxi, Ningxia, and Gansu), eastern coastal regions (Shandong, Jiangsu, Shanghai, and Anhui), and northern municipalities (Beijing and Tianjin) of China (Z > 2.58, p < 0.001). Zhejiang Province showed moderately strong growth (Z > 1.96, p < 0.05), while southern China displayed more modest increases, with Guangdong being the only province in this region that exhibited slow but consistent growth (Z > 1.65, p < 0.01).

Further analysis of the 2023 provincial CH₄ emissions revealed statistically significant clustering patterns across China’s 31 provinces, as evidenced by a Moran’s I index of 0.614 (Z-score = 7.89, p < 0.01) and a General G statistic of 0.01 (Z-score = 5.32, p < 0.01), confirming both global clustering tendencies and local aggregation of high emission values. The analysis identified distinct spatial patterns: Shaanxi, Shanxi, and Henan formed a contiguous high-high cluster, where both these provinces and their neighboring areas exhibited elevated emissions, whereas Guangdong emerged as a spatial outlier with high-low clustering characteristics, where this high-emission province was surrounded by lower-emitting regions (Figure 4a). Hotspot analysis further identified Shaanxi, Shandong, Inner Mongolia, Shanxi, and Hebei as the primary emission hotspots, with Shaanxi displaying the most pronounced hotspot intensity, demonstrating the marked spatial heterogeneity in China’s provincial CH₄ emission distribution (Figure 4b).

3.2. Distribution Characteristics of CH₄ Emissions by Sectors

Based on the IPCC 2006 sectoral classification [22], an analysis of the spatial distribution characteristics of CH₄ emissions across China’s energy (Figure 5a), agriculture (Figure 5b), industry (Figure 5c), and waste (Figure 5d) sectors in 2023 revealed distinct patterns. In the energy sector, emissions are predominantly concentrated in coal-rich regions, such as Inner Mongolia, Shanxi, and Shaanxi, each with emissions exceeding 0.8 Mt, followed by secondary hotspots in Xinjiang (a major natural gas producer) and Anhui (a significant coal-producing region), both emitting approximately 0.2 Mt. Moderate emission levels (0.15–0.2 Mt) are observed in Shandong, Henan, and Ningxia. Agricultural CH₄ emissions, influenced by farming systems and livestock distribution, exhibit a generally high in the south and low in the north pattern, with elevated emissions in northern pastoral regions such as Inner Mongolia (0.3 Mt) and Hebei (0.17 Mt). Simultaneously, southern rice-producing provinces, including Hunan, Hubei, Sichuan, Anhui, Jiangsu, Guangdong, and Guangxi, each emit 0.1–0.2 Mt. Industrial emissions remain relatively low nationwide (<0.03 Mt) and are spatially homogeneous, with slightly higher levels in traditional industrial bases, such as Liaoning and Hubei. Waste-related emissions are driven by economic development, with the highest contributions from Shandong, Jiangsu, and Guangdong (0.09–0.18 Mt each). This sectoral and spatial heterogeneity underscores the need for region-specific mitigation strategies tailored to the dominant emission sources.

3.3. Driving Factors and Cluster Analysis of CH₄ Emissions

Given the declining trend in total CH₄ emissions in Tibet from 2000 to 2023, this study conducted a K-means cluster analysis of CH₄ emissions and nine influencing factors for 30 Chinese provinces (excluding Tibet) in 2022, based on the latest available provincial energy balance data. The two principal components (PC1 and PC2), derived from methane emissions and influencing factors, collectively explain 57.4% of the total variance attributable to the nine factors including fossil energy consumption, freight turnover serves, raw coal production, crude oil production, natural gas production, rice sown area, large livestock year-end Inventory, domestic waste collection volume and added value of the secondary industry (Figure 6).

The elbow method identified four optimal clusters, with descriptive statistics for each cluster’s CH₄ emissions and influencing factors, as shown in

Table 2. Descriptive statistics of CH₄ emissions and influencing factors by cluster.

CH4 Emission and Its Influencing Factors	Type	Resource-Dependent, Coal-Driven Regions	Economically Active, Waste-Driven Regions	Balanced-Development, Multi-Factor Regions	Ecology-Rich, Agro-Petro Complex Regions
Emissions/Mt	Min	4,602,180	1,605,693	555,974	222,045
	max	7,935,822	3,012,190	3,449,280	1,904,494
	mean	6,800,180	2,464,772	1,926,427	932,727
Fossil energy consumption/104 tonsof standard coal	Min	15,186.29	31,635.38	9877.30	2091.27
	max	37,668.03	49,040.49	33,400.14	25,196.59
	mean	25,248.51	39,198.00	19,501.70	10,053.97
freight turnover serve/104 persons	Min	4368.94	11,829.27	2931.57	702.64
	max	6473.00	28,078.11	32,369.69	9963.70
	mean	5354.22	16,931.45	11,242.64	2914.58
raw coal production/kt	Min	74,876.00	0.00	0.00	0.00
	max	132,009.00	8753.00	11,177.00	41,305.00
	mean	109,413.00	2429.25	3375.44	6251.43
crude oil production/kt	Min	0.00	0.00	0.00	0.00
	max	2537.00	2200.00	984.00	3575.00
	mean	861.00	1059.50	208.33	841.21
Natural gas production/kt	Min	132.00	0.00	0.00	0.00
	max	307.00	124.00	19.00	554.00
	mean	248.67	33.25	4.56	91.43
Rice Sown Area/103 hectares	Min	2.16	106.43	76.59	0.00
	max	117.23	2221.42	3967.67	3601.37
	mean	75.16	1198.23	1558.77	743.00
Large Livestock Year-end Inventory/104 heads.	Min	154.13	15.54	5.80	8.44
	max	956.88	287.04	443.60	943.91
	mean	421.85	109.90	249.35	444.19
Domestic Waste Collection Volume/kt	Min	348.56	1553.53	527.70	117.25
	max	654.88	3280.64	1087.87	1259.20
	mean	490.04	2129.23	862.49	473.79
added value of the secondary industry/Yuan/person	Min	11,241.80	33,205.20	11,458.40	1310.90
	max	15,933.10	55,888.70	25,465.00	21,157.10
	mean	13,671.90	44,237.90	18,242.03	6989.77

. The key characteristics of each cluster are as follows.

Resource-dependent, coal-driven regions (Inner Mongolia, Shaanxi, and Shanxi): This cluster exhibited the highest mean CH₄ emissions (0.68 Mt), reflecting its heavy reliance on coal and fossil energy. These regions led in raw coal and natural gas production, with high energy consumption and crude oil outputs. Simultaneously, non-energy factors, such as rice planting area and municipal waste disposal, remained low.

Economically active, waste-driven regions (Shandong, Guangdong, Jiangsu, and Zhejiang): These provinces had advanced socioeconomic development, featuring the highest secondary industry GDP and waste disposal volumes, and the second-highest mean emissions (0.25 Mt). Freight turnover and fossil energy consumption were also elevated, whereas coal/gas production and livestock numbers were relatively low. Shandong’s crude oil output was higher owing to its petroleum resources.

Balanced-development, multi-factor regions (Hebei, Henan, Liaoning, Anhui, etc.; 9 provinces): emissions averaged 0.19 Mt, with moderate levels in freight turnover, rice cultivation, and waste disposal. Energy production, fossil fuel consumption, and secondary industry GDP were generally lower than those in other clusters.

Ecology-rich, agro-petro complex regions (Xinjiang, Sichuan, Heilongjiang, etc.; 14 provinces): This cluster had the lowest emissions (mean: 0.09 Mt), with strong agricultural and pastoral economies, including the highest livestock inventories. While coal, crude oil, and natural gas production remained substantial, economic development indicators—freight turnover, waste disposal, and secondary industry GDP—were comparatively low.

4. Discussion

4.1. Characteristics of CH₄ Emissions

China’s CH₄ (CH₄) emissions exhibit substantial magnitude and structural complexity, with marked sectoral heterogeneity in emission drivers. Temporally, the evolution of CH₄ emissions can be divided into three distinct phases. During Phase I (2000–2010), China’s accession to the WTO accelerated industrialization and urbanization, driving rapid economic growth. This period witnessed sustained increases in CH₄ emissions due to expanded energy combustion and agricultural activities, with socioeconomic development prioritized over environmental governance. The absence of targeted mitigation policies resulted in a notably high annual growth rate of CH₄ emissions. Phase II (2011–2016) saw structural shifts in emission drivers as economic growth moderated and environmental policies strengthened. Key interventions included the 12th Five-Year Plan, which for the first time incorporated non-CO₂ greenhouse gases into regulatory frameworks; the 2013 Air Pollution Prevention and Control Action Plan, which reduced coal consumption through coal-to-gas conversion; and the 2016 coal overcapacity reduction policy, which closed 5000 outdated coal mines, cutting coal mine CH₄ emissions by 9.6%. Concurrently, enhanced resource utilization policies—such as increased coalbed CH₄ extraction and expanded rural biogas coverage—collectively slowed emission growth, leading to a stabilization period [12,35]. Phase III (2017–2023) was characterized by fluctuating emissions. Despite institutional advancements under the “carbon peaking and carbon neutrality goals”, inertia in energy transition sustained fossil fuel dependence, with coal-fired power capacity continuing to grow. The COVID-19 pandemic in late 2019 triggered a surge in medical waste landfilling, while rising meat consumption intensified CH₄ emissions from concentrated animal feeding operations. Inadequate monitoring coverage and delayed deployment of mitigation technologies further contributed to emission rebound [36].

From a spatial perspective, China’s CH₄ emissions exhibit marked geographical imbalances, with regional emission profiles showing distinct heterogeneity driven by factors such as resource endowments, industrial distribution, and climatic conditions. Coal mining-related CH₄ fugitive emissions, China’s largest CH₄ source, are predominantly concentrated in major coal-producing regions, such as Inner Mongolia, Shanxi, Shaanxi, and Xinjiang [9], reflecting the nation’s geospatial dependence on energy production. Agricultural emissions align with national food security zones and regional livestock policies: enteric fermentation emissions cluster in the pastoral regions of Inner Mongolia, Hebei (North China Plain), and Sichuan (Southwest China), which are the core areas of intensive animal husbandry, consistent with the findings of Zhang et al. [14]. Rice-derived CH₄ emissions demonstrate a southeast-high/northwest-low pattern, concentrated in the Yangtze River Delta (Hunan and Hubei), Pearl River Valley, and delta regions, where continuous flooding of southern paddies enhances anaerobic CH₄ production. Waste sector emissions predominantly occur in coastal eastern economic hubs, such as Guangdong Province, as noted by Li et al. [37]. Despite lower total emissions, the rising population density from high-tech industry agglomeration and lagging waste treatment technology adoption have gradually increased landfill CH₄ emissions. Simultaneously, the aforementioned areas are regions with a high incidence of ozone pollution. Conversely, the western regions (Tibet and Qinghai) show minimal emissions owing to sparse populations, underdeveloped economies, and the absence of major emission sources such as large-scale coal mining or intensive agriculture [38]. This pronounced regional disparity challenges nationwide uniform mitigation policies, necessitating place-specific strategies that consider local industrial structures, resource availability, and technological capacities. Simultaneously, advanced CH₄ reduction and ozone pollution control should be implemented in a coordinated manner.

4.2. Regional CH₄ Control Strategies

Given the distinct CH₄ emission characteristics and driving factors across the four regional types, targeted regional CH₄ control measures can be implemented.

For resource-dependent coal-dominant regions influenced by China’s coal distribution pattern of “rich in north and west, poor in south and east,” the large coal bases in Shanxi, Shaanxi, and Inner Mongolia exhibit significant CH₄ emissions mainly from two key processes [39]: CH₄ release from coal fractures due to mining disturbances and CH₄ leakage through overlying rock fractures and abandoned shafts in closed mines [12]. Additionally, the direct venting of ultra-low concentration coalbed CH₄ (below 9% volume fraction) owing to technical and economic constraints further exacerbates emissions [40]. Based on these resource-dependent emission characteristics, these regions should focus on coal mine CH₄ reduction by incorporating carbon emission reductions from coal mine gas utilization projects into the national carbon market, developing key technologies for low-concentration gas purification and abandoned mine resource utilization, and a comprehensive methane monitoring, reporting, and verification (MRV) framework in high-emission areas such as Shanxi-Shaanxi-Inner Mongolia. Integration of unmanned aerial vehicles (UAVs) is recommended to enhance both the spatial resolution and accuracy of coal methane emission monitoring [41]. This multi-platform approach will support precise and verifiable methane management.

For economically active waste-dominant regions, primarily coastal developed cities with high population density and substantial municipal waste generation, CH₄ emissions mainly originate from urban solid waste treatment processes [42]. The low waste incineration rate in these areas leads to the accumulation of organic matter (such as food waste and paper) in landfills, which continuously produces CH₄ under anaerobic conditions. Furthermore, Guangdong’s beverage processing and textile industries generate large amounts of high-COD organic wastewater and sludge [37], and their open storage or simple landfilling further increases CH₄ generation. These regions should prioritize waste sector CH₄ control by transforming landfills and wastewater treatment plants from simple pollutant reduction to resource recovery, enhancing CH₄ capture, and improving treatment efficiency. New landfills should be completely banned, and waste-to-energy incineration should be promoted as the primary treatment method, with strengthened CH₄ recovery and emission control from existing landfill facilities in China.

For balanced-development multi-factor dominant regions where various indicators reach medium-high levels, CH₄ emissions come from multiple sources, including industry, agriculture, and waste. These regions rely heavily on external fossil fuel supplies, resulting in significant CH₄ emissions from transportation. Combined with the delayed energy transition in old industrial bases such as Liaoning, inefficient industrial boilers cause incomplete coal combustion, while industrial processes such as steel production in Hebei lead to industrial CH₄ leaks [43]. As major rice production areas, Hubei and Hunan contribute to agricultural CH₄ emissions through continuous flooding rice cultivation [44]. Emerging megacities, such as Zhengzhou, with high population densities but low waste incineration rates, also generate landfill CH₄ emissions. Given the complexity of multiple emission sources, these regions should implement precise “one-region-one-policy” mitigation strategies: accelerating energy transition and phasing out inefficient industrial boilers in industrial emission hotspots like Hebei and Liaoning; piloting “intermittent irrigation + rice-fish co-culture” and “biogas + organic fertilizer” models in agricultural regions such as Hubei and Hunan [45]; and promoting food waste anaerobic digestion for bio-natural gas (BNG) in megacities such as Zhengzhou with supporting financial incentives.

For ecologically rich oil-agriculture composite regions with substantial oil/gas production and large livestock inventories, CH₄ emissions mainly originate from two sectors: oil/gas operations and agriculture. In Xinjiang and Heilongjiang, where oil and gas production is concentrated, aging equipment and low associated gas recovery rates lead to significant CH₄ leakage [46]. Moreover, natural gas pipeline networks in remote areas, such as Xinjiang and Sichuan, have low monitoring coverage and high leakage risks owing to frequent geological activity [47]. Deployment of unmanned aerial vehicles (UAVs) enhances methane monitoring from oil and gas leaks and landfills with higher spatial resolution, providing critical complementary data to support MRV systems [48,49]. In the agricultural sector, regions such as Sichuan, Qinghai, and Tibet, with large ruminant populations, face significant CH₄ emissions from enteric fermentation, whereas the low coverage of manure treatment facilities and biogas projects results in substantial CH₄ potential from untreated manure. Additionally, traditional continuous-flooding rice cultivation in Sichuan’s rice-growing areas prolongs CH₄ emission periods under anaerobic soil conditions [50]. CH₄ control in these regions should balance ecological protection with oil/gas and agricultural development through an “oil/gas reduction priority with agricultural synergy” approach: implementing rigorous leak detection and repair (LDAR) programs in oil/gas sectors [51]; promoting yak feed additives and herder subsidies to reduce enteric fermentation; adopting psychrophilic microbial degradation to lower landfill CH₄ generation rates; increasing biogas project coverage; and optimizing rice cultivation techniques to reduce flooding duration and associated CH₄ emissions.

Approximately 60% of global anthropogenic methane emissions can be abated through low-cost measures, with nearly half of these interventions yielding net-negative costs [5]. Future efforts should not only advance deep emission reductions under China’s Methane Emission Control Action Plan, but also enhance co-governance of methane and ground-level ozone pollution. It is recommended to prioritize cost-effective and negative-cost mitigation strategies based on a systematic analysis of the spatiotemporal relationships between methane and ozone, thereby maximizing synergies among environmental improvement, climate mitigation, and economic development.

This study has several limitations. Although the EDGAR database offers relatively high spatial resolution, it may still fail to capture fine-grained regional emission patterns. Future work could integrate multi-source data, such as MEIC, for cross-validation and improved accuracy. Moreover, the use of provincial-level analysis may introduce uncertainties through the modifiable areal unit problem (MAUP), particularly in hotspot identification. Subsequent studies should conduct multi-scale sensitivity analyses at the prefectural level to enhance the robustness of the findings.

5. Conclusions

This study analyzed the spatiotemporal characteristics of China’s CH₄ emissions from 2000 to 2023 using the EDGAR and driver datasets, and proposed region-specific control policies. Key findings reveal:

1.

Temporally, total CH₄ emissions followed a fluctuating upward trajectory with three distinct phases—rapid growth (2000–2010), stabilization (2011–2016), and resurgence (2017–2023)—closely aligned with economic cycles, energy restructuring policies, and waste management advancements.

2.

Spatially, emissions exhibited an east-high-west-low gradient with northwestward expansion, where hotspots were concentrated in coal-mining regions (Inner Mongolia, Shanxi, Shaanxi), accounting for the largest share, followed by agricultural emissions from livestock (Inner Mongolia, Sichuan) and rice paddies (Hunan, Hubei), while waste sector emissions dominated coastal provinces (Shandong, Jiangsu, Guangdong), and industrial contributions remained minimal and spatially uniform. Low-emission zones (Tibet and Qinghai) were correlated with sparse populations and underdevelopment, with Tibet demonstrating a declining trend.

3.

Cluster analysis identified four spatially heterogeneous emission typologies: (i) resource-dependent, coal-driven regions, (ii) economically active, waste-driven regions, (iii) balanced-development, multi-factor regions, and (iv) ecology-rich, agro-petro complex regions. To achieve precise mitigation, we recommend (a) scientifically delineating priority control zones, (b) focusing on key emission sectors for prioritized regulation, and (c) deploying tailored emission-reduction technologies. In addition, in key areas affected by ozone pollution, coordinated control of CH₄ reduction and ozone pollution should be strengthened.