Spatiotemporal Patterns and Inﬂuencing Factors of Agriculture Methane Emissions in China

: Explaining the methane emission pattern of Chinese agriculture and the inﬂuencing factors of its spatiotemporal differentiation is of great theoretical and practical signiﬁcance for carbon neutrality. This paper uses the IPCC coefﬁcient method to measure and analyze the spatial and temporal differentiation characteristics of agricultural methane emission, clarify the dynamic evolution trend of the kernel density function, and reveal the key inﬂuencing factors of agricultural methane emission with geographical detectors. The results show that China’s agricultural methane emissions showed a ﬁrst increasing and then declining trend. Agricultural methane emissions decreased from 21.4587 million tons to 17.6864 million tons, with an upward trend from 2000 to 2005, a signiﬁcant decline in 2006, a slow change from 2007 to 2015, and a signiﬁcant decline from 2015 to 2019. In addition, the emissions pattern of the three major grain functional areas is characteristic; in 2019, agricultural methane emissions from main producing area, main sales area, and balance area were 10.8406 million tons, 1.2471 million tons, and 5.599 million tons, respectively. The main grain producing area is the main area of methane emissions, and the emission pattern will not change in the short term. The variability of grain functional areas is the decisive factor for the difference in agricultural methane emissions. The state of industrial structure is the key inﬂuencing factor for adjusting the spatial distribution—the explanatory power of the industrial structure to the main producing areas reached 0.549; the level of agricultural development is the most core inﬂuencing factor of the spatial pattern of the main grain sales area—the explanatory power reached 0.292; and the level of industrialization and the industrial structure are the core inﬂuencing factors of the spatial pattern of the balance area—the explanatory power reached 0.545 and 0.479, respectively.


Introduction
Climate change has become the biggest threat to global sustainable development [1][2][3][4].The Paris Agreement stipulates that the state parties should keep the global average warming to 2 degrees Celsius higher than the pre-industrial revolution level and strive to limit it to 1.5 degrees Celsius [5].The "climate critical points" of 1.5 and 2 degrees Celsius are the key time points for catastrophic climate events but are also related to the threat of human survival, which may produce a chain reaction once exceeded [6].In addition, frequent occurrence of global extreme climate caused by climate change has brought significant economic and social losses to human beings [7].How to mitigate and adapt to climate change has become a major issue for countries all over the world.Carbon peaking and carbon neutrality is a solemn commitment of China to address global climate change, and a major declaration of the transformation of social and economic development mode [8].To achieve carbon peaking by 2030 and strive to carbon neutrality by 2060 is not only an essential part of China's high-quality growth but also a reform requirement for major adjustments in different industries [9].
Methane is the second largest greenhouse gas responsible for climate change after carbon dioxide [10].Strengthening methane emissions reduction has become a necessary item in the 21st century [11].In 2020, global methane emissions were 570 million tons, of which human activities produced 340 million tons.IPCC pointed out in Working Group I of the Sixth Assessment Report that within 20 years after emission, the greenhouse effect of 1 ton of methane is comparable to 84 tons of carbon dioxide and its warming effect is still 28 times that of carbon dioxide even after 100 years [12].In addition, since it is easier to reduce methane than carbon dioxide, the International Energy Agency (IEA) notes that 75% of methane leaks in the global oil and gas industry can be controlled with existing technologies, and 50% of methane emissions are reduced at net zero cost.[13].Methane has a shorter lifetime in the atmosphere than carbon dioxide and can be a priority for emissions reduction.Emission reductions of non-carbon dioxide greenhouse gases such as methane are essential for the goal of controlling global warming to 1.5 • C by the end of this century [14][15][16].
As agriculture is the main source of methane emissions, it is of great significance to clarify its spatial distribution and influencing factors [17].Methane emissions from agriculture account for about 1/5 of the total global emissions, and the main sources of agricultural methane emissions are farming and animal husbandry.The carbon emissions produced by the planting industry are mainly generated by paddy planting, while the carbon emissions produced by the breeding industry are mainly generated by the intestinal fermentation and manure management of livestock and poultry [18].
Scholars from all over the world have studied agricultural methane emissions from different perspectives.Some scholars have estimated the intestinal methane emissions of livestock and poultry in the Southern African Development Community (SADC) [19].Some scholars have studied methane emissions from wetland rice fields [20].In addition, some researchers have compared the intestinal methane emissions of livestock and poultry in Germany in the late 19th century with the current situation [21].As a major agricultural country, China's methane emissions from agricultural activities account for 50.15% of the country's total emissions.Agricultural methane emission patterns and influencing factors have become a hot topic in academic research [22][23][24][25][26][27].After estimating the methane emissions of various regions in China, it is indicated that the total methane emissions from agricultural interaction in 2018 were 18.22 million tons, among which the intestinal fermentation emissions from livestock and poultry were the largest, accounting for 50.69%, the emissions from paddy were 35.17%, and the emissions from livestock and poultry management were 14.14% [28].By measuring methane and nitrous oxide emissions from 1980 to 2018, Li Yang pointed out that the methane missions generated from agriculture increased from 0.56 × 10 9 CO 2 -eq to 0.73 × 10 9 CO 2 -eq, which were mainly affected by efficiency factor, structure factor, and population size factor [29].In addition, scholars' research focused on the provincial level, and calculated the changes in Jiangxi, Guangdong, Heilongjiang, and other provinces.Some scholars have also measured methane emissions from lake farms and their response to ecological restoration [30], which provided a solid foundation to study agricultural methane emissions.However, at present, there are relatively few estimates based on the national level, and there is a lack of comparative studies on the differences among main grain producing area, main sales area, and balance area; thus, relevant influencing factors need to be further clarified.In this paper, the spatial and temporal patterns of methane emission in China were analyzed by using the IPCC method and the calculation method of the Provincial Greenhouse Gas Inventory Guide.The influencing factors were calculated by using the analysis method of the geographic detector, and the emission reduction countermeasures and suggestions were put forth to provide some references for methane emission reduction in China.

Measurement Method of Agricultural Methane Emission
The CH 4 emissions of the national agricultural system mainly come from paddy fields and livestock and poultry manure management and intestinal fermentation [28].Among them, the CH 4 emissions from paddy fields generally follow the basic methods and requirements determined by the IPCC guidelines, and the formula is calculated as: where E CH 4 rice represents CH 4 emissions from paddy fields (×10 4 t), EF i represents methane emission factors from paddy fields (kg/hm 2 ), and AD i represents the sown area of this type of methane emission factor (×10 3 hm 2 ),The methane emission factors of paddy fields are listed in Table 1.The calculation formula of CH 4 generated by livestock and poultry manure management is calculated as follows: where E CH 4 ,manure,i refers to the amount of methane produced by manure management of species i (×10 4 t), EF CH 4 ,manure,i refers to manure management methane emission factor of species i, and AP i refers to the number of species i, the methane emission factors for manure management are listed in Table 2.The calculation formula of CH 4 produced by intestinal fermentation of livestock and poultry is calculated as follows: where E CH 4 ,enteric,i is the amount of methane produced by intestinal fermentation of species i (×10 4 t), EF CH 4 ,enteric,i,j is methane emission factor from intestinal fermentation of livestock and poultry of species I, AP i is the number of species I, and R j is the breeding proportion of this livestock and poultry, The methane emission factor from enteric fermentation of livestock and poultry are listed in Table 3.

Kernel Density Function
The kernel density function, due to its non-parametric estimation power of the probability density, can be used to measure the distribution form of random variables, with the expression of: where n represents the number of the sample, h represents bandwidth, h = 0.9SN 4/5  (N represents the number of the sample and S represents sample standard deviation).K( x−x i h ) represents kernel density function, and the Epanechnikov kernel density form is used.The dynamic evolution of methane emission zones can be reflected by the distribution interval, morphology, and kurtosis extension of kernel density function.If the function presents a "single peak" as a whole, there are no multiple equilibrium states; if the function appears a "double peak" or "multi-peak" state, there are two or more equilibrium points.

Geographic Detector
To deeply explore the causes of regional differences in agricultural methane emissions, geographic detector analysis is used in this study.Geographic detectors can not only test the spatial differentiation of the delay variables but also excavate the influencing factors of spatial differentiation, explain the decisive role of influencing factors, and provide panoramic display for the exploration of the spatial differences.In this study, the factor detection method is adopted to determine the fundamental factor of spatial differentiation of agricultural methane emissions through the determinant force index q, in which Y represents the agricultural methane emission, X = {X m } represents influencing factors,(m = 1, 2, . . .L;), L refers to the number of partitions of factor X, and X m refers to the different partitions of factor X. The determinant force index q of factor X on Y is as follows: where N is the number of provinces in the study, N m is the number of provinces contained in the m-th partition of factor X, σ 2 is the variance of region Y, σ 2 m is the variance of the driver X in the m subdomain, and SSW and SST represent the sum of variance in each grain function and the total variance in the whole region, respectively.In general, the larger the q-value of factor X, the stronger the factor of driving force for the spatial analysis.When the driving factors X and Y have driving effects on each other, the sum of the variances within regions are usually less than the sum of the variances between regions.In order to compare whether the cumulative variance of different grain production regions is significantly different from the overall variance of the whole region, the F-statistic test is introduced in this paper.

Data
In this paper, the paddy fields area, livestock and poultry quantity, and breeding proportion are from the China Statistical Yearbook (2000-2019), State Administration of Grain, China Agriculture Yearbook, and China Animal Husbandry and Veterinary Yearbook (2000-2019).Methane emission factors in different paddy fields, from livestock and poultry manure management in different regions, and from livestock and poultry intestinal methane emission factors under different feeding methods are from the Provincial Guidelines for Greenhouse Gas (National Development and Reform Commission, 2011), and some missing data are supplemented by a linear interpolation method.gions is significantly different from the overall variance of the whole region, the F-statis test is introduced in this paper.

Data
In this paper, the paddy fields area, livestock and poultry quantity, and breedi proportion are from the China Statistical Yearbook (2000-2019), State Administration Grain, China Agriculture Yearbook, and China Animal Husbandry and Veterinary Yea book (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019).Methane emission factors in different paddy fields, from livestock a poultry manure management in different regions, and from livestock and poultry intes nal methane emission factors under different feeding methods are from the Provinc Guidelines for Greenhouse Gas (National Development and Reform Commission, 201 and some missing data are supplemented by a linear interpolation method.From 2006 to 2015, there was a slow change period, with a slight increase in metha emissions, and since 2015, methane emissions have fallen significantly.From the perspe tive of evolution, the No.1 central document in 2004 "opinions on several policies of pr moting farmers' income " was put forward after 18 years, and its main body returned "the field of agriculture, rural areas and farmers".The document put forward the "thr From 2006 to 2015, there was a slow change period, with a slight increase in methane emissions, and since 2015, methane emissions have fallen significantly.From the perspective of evolution, the No.1 central document in 2004 "opinions on several policies of promoting farmers' income" was put forward after 18 years, and its main body returned to "the field of agriculture, rural areas and farmers".The document put forward the "three subsidies" of direct subsidies for farmers, subsidies for improved varieties, and subsidies for agricultural machinery.It also proposed some measures to reduce the agricultural tax burden, which effectively enhanced the enthusiasm of farmers to grow grain and prompted significant improvement in the level of green development in 2006.In 2015, in order to cope with the problems such as the "ceiling" of prices, the rise of the "floor" of production costs, and the intensified challenges of the "hard constraints" of resources and the environment, the General Office of the State Council of the People's Republic of China issued "the Opinions on Accelerating the Transformation of Agricultural Development Mode", which provided an important starting point for promoting agricultural development on quantity, quality, and efficiency.Under the guidance of the document, the structural adjustment to the agriculture led to a further decline in methane emissions.In terms of the sources of methane emissions, from 2000 to 2019, emissions from paddy fields fell from 6.51 million tons to 6.29 million tons, emissions from livestock and poultry manure decreased from 2.72 million tons to 1.88 million tons, and emissions from livestock and poultry intestinal fermentation dropped from 12.23 million tons to 9.53 million tons (Figure 2).The decline of livestock and poultry intestinal emissions played a crucial role in reducing agricultural methane emissions.

Measurement of Agricultural Methane Emission in China
issued "the Opinions on Accelerating the Transformation of Agricultural Developme Mode", which provided an important starting point for promoting agricultural develo ment on quantity, quality, and efficiency.Under the guidance of the document, the stru tural adjustment to the agriculture led to a further decline in methane emissions.In term of the sources of methane emissions, from 2000 to 2019, emissions from paddy fields f from 6.51 million tons to 6.29 million tons, emissions from livestock and poultry manu decreased from 2.72 million tons to 1.88 million tons, and emissions from livestock an poultry intestinal fermentation dropped from 12.23 million tons to 9.53 million tons (Fi ure 2).The decline of livestock and poultry intestinal emissions played a crucial role reducing agricultural methane emissions.

Spatial Analysis of Different Production Areas
Considering the differences between agricultural development among different r gions in China, the analysis between spatial regions is more targeted.Based on the clas fication of agricultural grain functional areas, the whole country is divided into main pr ducing area, main sales area, and balance area.
Agricultural methane emissions in the main producing area, main sales area, an balance area showed a downward trend, but the changes had their own characteristi (Figure 3).The main producing area was the region with the highest agricultural metha emissions.From 2000 to 2019, the main producing area decreased the most, from 12. million tons to 10.84 million tons.From the perspective of the planting industry, the sow area of double-cropping early rice and double-cropping late rice in the main producin areas decreased by 27.61% and 31.83%,respectively, but the sown area of single-crop ri increased by 48.38%, which led to an increase in methane emissions from paddy fields the main producing areas.From the perspective of the breeding industry, although t methane emissions from paddy fields in the main producing areas increased to a certa extent, an increase of 14.36%, the methane emissions from livestock and poultry manu

Spatial Analysis of Different Production Areas
Considering the differences between agricultural development among different regions in China, the analysis between spatial regions is more targeted.Based on the classification of agricultural grain functional areas, the whole country is divided into main producing area, main sales area, and balance area.
Agricultural methane emissions in the main producing area, main sales area, and balance area showed a downward trend, but the changes had their own characteristics (Figure 3).The main producing area was the region with the highest agricultural methane emissions.From 2000 to 2019, the main producing area decreased the most, from 12.79 million tons to 10.84 million tons.From the perspective of the planting industry, the sown area of double-cropping early rice and double-cropping late rice in the main producing areas decreased by 27.61% and 31.83%,respectively, but the sown area of single-crop rice increased by 48.38%, which led to an increase in methane emissions from paddy fields in the main producing areas.From the perspective of the breeding industry, although the methane emissions from paddy fields in the main producing areas increased to a certain extent, an increase of 14.36%, the methane emissions from livestock and poultry manure management and enteric fermentation decreased by 507,500 tons and 2,028,000 tons, respectively.From 2000 to 2019, the number of livestock and poultry breeding dropped significantly, and the number of non-dairy cows, goats, pigs, poultry, horses, camels, and donkeys decreased by 43.83%, 20.43%, 29.04%, 16.78%, 62.97%, 88.65%, and 70.96%, respectively.The number of poultry breeding fell the most.The main production areas bear the heavy responsibility of China's food security, and the decline in emissions shows that the industrial structure has been improved, to a certain extent, and agricultural green production has gradually improved.From the perspective of planting industry, the sown area of double-cropping early rice and double-cropping late rice in the main producing areas decreased by 27.61% and 31.83%,respectively, but the sown area of single-crop rice increased by 48.38%, which led to an increase in methane emissions from paddy fields in the main producing areas.From the perspective of the breeding industry, although the methane emissions from paddy fields in the main producing areas increased to a certain extent, with an increase of 14.36%, the methane emissions from livestock and poultry manure management and enteric fermentation decreased by 507,500 tons and 2,028,000 tons, respectively.From 2000 to 2019, the number of livestock and poultry breeding dropped significantly, and the number of non-dairy cows, goats, pigs, poultry, horses, camels, and donkeys decreased by 43.83%, 20.43%, 29.04%, 16.78%, 62.97%, 88.65%, and 70.96%, respectively, and the number of poultry breeding fell the most.The main production areas bear the heavy responsibility of China's food security, and the decline in emissions has shown the improved industrial structure and increased agricultural green production.Methane emissions in the balance area fell from 6.29 million tons to 5.60 million tons.In the balance area, methane emissions from rice fields, livestock manure management, and livestock intestinal fermentation all decreased by 23,600 tons, 169,800 tons and 285,400 tons, respectively.The balance area is located mainly in Central and Western China, and the change in emissions was partly due to the implementation of China's western development strategy.In addition, methane emissions in the main sales area dropped from 2.37 million tons to 1.25 million tons, decreasing by 47.48%.The methane emissions from paddy fields, livestock manure management, and livestock enteric fermentation in the main sales area decreased by 41%, 48.14%, and 61.25%, respectively.Except for dairy cows, the number of livestock and poultry has decreased significantly.This was mainly due to the fact that the main sales areas, such as Beijing, Tianjin, and Shanghai, are relatively developed areas in China.With limited land and other planting resources and a large food shortage, it is necessary for these developed areas to rely on other provinces to transport food, forcing the region to adjust the industrial structure and objectively promoting the reduction of agricultural methane emissions.The decline in methane emissions in main producing and sales regions was the main reason for the overall decline.Therefore, the main producing areas have high methane emissions and are key areas for the green transformation and development of agriculture.
From 2000 to 2019, the emissions of the main producing areas, main sales areas, and balanced areas were all trending downward, of which the largest decline rate was the main sales area (47.48%), and the largest decline amount was the main producing area (1.9539 million tons).Although they all showed a downward trend, the changes had their own characteristics.For example, the main grain producing areas were the areas with the highest agricultural methane emissions which have dropped by approximately 15.27% in the past 20 years, with an average annual decline of 0.83%, of which 2000-2005 and 2010-2015 were the rising periods, with increases of 10.78% and 2.34%, respectively; 2005-2010 and 2015-2019 were the decreasing periods, and the decline was relatively large, with 13.14% and 13.96%, respectively.Although the total amount of emissions had a downward trend, the discharge of paddy fields had an upward trend, and the proportion of emissions was also different from the total amount of emissions.2000-2005 and 2015-2019 were decreasing periods, 2005-2015 was the rising period, where the total amount increased from 4.0507 million tons in 2000 to 4.6324 million tons in 2019, and the proportion of emissions increased from 31.66% to 42.73%.Manure management emissions and intestinal fermentation emissions were on a downward trend, decreasing by approximately 30.02% and 28.75%, respectively.In short, the emission structure of the main producing areas was optimized, where livestock and poultry intestinal emissions were the main source of emissions; the proportion of paddy field emissions and intestinal emissions was close, with 42.73% and 46.36%, respectively in 2019.Paddy field management will continue to be an important measure to reduce emissions in major producing areas.Agricultural methane emissions from the main sales areas were the lowest, and although the agricultural methane emissions from paddy fields were not the highest among the three grain divisions, the ratio of emission sources in the main sales areas was the highest, from 58.67% in 2000 to 65.91% in 2019.In addition, 2000-2005 was a decline period, reducing from 58.67% to 52.99%, followed by an increase period in 2005-2019, with a growth of 8.91%.The trend of changes in the proportion of manure management emissions and intestinal emissions was similar, showing the trend of first rising and then declining, of which the total amount of manure management emissions did not change much in 2000-2010, but the proportion of emissions increased from 14.42% to 19.32%.Intestinal fermentation emissions fell by 23,500 tons between 2000 and 2005, but the proportion of emissions increased from 26.91% to 29.99%, and then in 2019, the proportion of emissions fell to 19.85%.Overall, intestinal fermentation emissions fell by 391,300 tons, with the proportion decreasing by 7.05%.The total decline (691,200 tons) and the decline rate (10.99%) in the balance area in 20 years were the lowest, similar to the change trajectory of the main producing areas where the increase periods were 2000-2005 and 2010-2015, and the decline periods were 2005-2010 and 2015-2019.The emission sources were similar to those in the main producing areas, with intestinal fermentation emissions as the main emission source, but the change amplitude of the three emission sources was not high.In addition, it should be noted that the proportions of intestinal emissions and paddy field emissions were different from the change trend of the main producing areas and the main sales areas; the change range from 2000 to 2019 was between 3.81% and −2.12%, and the three emission sources were reduced by 236,000 tons, 169,800 tons, and 285,400 tons, respectively, from which can be seen that the emission reduction effect of the balance area was not good.In addition, in 2000, the main producing area and the main sales area were 5.39 times and 2.65 times, respectively, the total emissions of the balanced area and by 2019, they had developed into 8.69 times and 4.49 times, respectively; this, to a certain extent, also reflects the widening of spatial differences.

Spatial Analysis of Different Provinces
In view of the different emphases of industrial structure and economic development in each province, the changes in agricultural methane emissions in each province are analyzed at the provincial level (Figure 4).In 2000, the top provinces in agricultural methane emissions were Hunan, Henan, Sichuan, Guangxi, and Shandong, and by 2010, the top provinces were Hunan, Sichuan, Henan, Inner Mongolia, and Guangxi.The emissions of Shandong decreased from 5.70% to 3.90%, and the emissions of Inner Mongolia increased from 3.27% to 5.68%.In addition, Heilongjiang rose significantly from 3.19% to 5.01%.In 2019, the top emission provinces were Hunan, Sichuan, Inner Mongolia, Heilongjiang, and Jiangxi.The reason for the highest agricultural methane emissions in Hunan was that it is a large paddy production and marketing province in China, with the paddy planting area of 3.8556 million hm2 in 2019.The large-scale breeding rate of cattle and sheep in Sichuan Province was low, and the proportion of large-scale pig breeding in the first half of 2020 (53.8%) exceeded the free-range ratio (46.2%) for the first time, resulting in high intestinal methane emissions from livestock and poultry; however, intensive production is gradually becoming the mainstream of production, and greening has been further improved.In 2018, Henan guided farmers to implement the plan of advocating intensive agricultural production, and at the end of 2018, 980,000 livestock and poultry free-range households had been "withdrawn from the village", which caused agricultural methane emissions in Henan to decrease by 2.31% from 2010 to 2019.The number of sheep raised in Inner Mongolia was much higher than that of other provinces; the large-scale breeding rate needs to be improved, which was the main reason for the high emission of agricultural methane.Therefore, scale is the only way for the transformation and development of animal husbandry.It is impossible for retail farming to completely withdraw, and financial means should be strengthened to support and promote new industrial models.For example, building a comprehensive platform for investment funds and technologies by industry experts and attracting retail investors to join the company can not only reduce the risk and cost of retail investors but also increase the rate of large-scale breeding and effectively reduce methane emissions.The agricultural methane emissions of Tianjin, Beijing, Shanghai, and other Eastern developed provinces were relatively small and remained at the bottom of the emission proportion in 2000, 2010, and 2019.The reason lay in the composition of industrial structure.For example, in 2019, the proportion of the first industry in the GDP of the three provinces was 1.3%, 0.3%, and 0.3%, respectively.The number of livestock and poultry in Ningxia and Hainan was low; in 2019, the number of livestock and poultry was 20.707 million and 63.913 million, respectively, far lower than the national average of 233.2885 million, so the methane emission was low.Liaoning and Jilin, also major grain-producing regions in Northeast China, had relatively low emissions, ranking lower (20th and 21st, respectively, in 2019).Compared with Heilongjiang, the two provinces had a lower paddy sown area, with only 22% and 13%, respectively, than Heilongjiang in 2019.In addition, among the main producing areas, Hebei accounted for 4.05% of emissions in 2000, and the number of livestock and poultry decreased by 37.26% in 2010 compared with 2000, which was the main reason for its low emissions, ranking 18th in 2019.
million tons to 1.25 million tons, decreasing by 47.48%.The methane emissions from paddy fields, livestock manure management, and livestock enteric fermentation in the main sales area decreased by 41%, 48.14%, and 61.25%, respectively.Except for dairy cows, the number of livestock and poultry has decreased significantly.This was mainly due to the fact that the main sales areas, such as Beijing, Tianjin, and Shanghai, are relatively developed areas in China.With limited land and other planting resources and a large food shortage, it is necessary for these developed areas to rely on other provinces to transport food, forcing the region to adjust the industrial structure and objectively promoting the reduction of agricultural methane emissions.The decline in methane emissions in main producing and sales regions was the main reason for the overall decline.Therefore, the main producing areas have high methane emissions and are key areas for the green transformation and development of agriculture.From 2000 to 2019, the emissions of the main producing areas, main sales areas, and balanced areas were all trending downward, of which the largest decline rate was the main sales area (47.48%), and the largest decline amount was the main producing area (1.9539 million tons).Although they all showed a downward trend, the changes had their own characteristics.For example, the main grain producing areas were the areas with the highest agricultural methane emissions which have dropped by approximately 15.27% in the past 20 years, with an average annual decline of 0.83%, of which 2000-2005 and 2010- From the perspective of emission sources, in 2019, the agricultural methane emission in 12 of the 31 provinces was dominated by paddy field emissions, represented by Jiangsu (82.73%), and the remaining 19 provinces were dominated by intestinal fermentation emissions, represented by Qinghai, Gansu, Tibet, Xinjiang, and Inner Mongolia, which accounted for more than 90%, and the intestinal fermentation emissions caused by them reached 18.57% of the national total emissions.

Dynamic Evolution of Spatial Differentiation
To effectively reflect the spatial characteristics of agricultural methane emissions, the dynamic distribution of 2000-2019 and 2010-2019 was analyzed by non-parametric kernel density.As can be seen from Figure 5, the kernel density showed two peaks in the two time periods, and then decreased continuously.The distribution curve showed the right tail, the peak moved significantly to the right, and the gap of provincial agricultural methane emissions showed a widening trend.

Dynamic Evolution of Spatial Differentiation
To effectively reflect the spatial characteristics of agricultural methane emissions, the dynamic distribution of 2000-2019 and 2010-2019 was analyzed by non-parametric kernel density.As can be seen from Figure 5, the kernel density showed two peaks in the two time periods, and then decreased continuously.The distribution curve showed the right tail, the peak moved significantly to the right, and the gap of provincial agricultural methane emissions showed a widening trend.

Model Building of Influencing Factors
Referring to relevant references, the factors affecting agricultural methane emissions include agricultural development level, residents' living conditions, industrial structure, urbanization level, industrialization level, environmental regulation, agricultural machinery input, and agricultural science and technology R&D investment, etc.
Agricultural development level.There is a double-sided effect between agricultural development level and agricultural methane emissions.The improvement of agricultural development level is often accompanied by the expansion of production scale or structural adjustment, which may produce more methane emissions in this process.On the contrary, the improvement of agricultural development level drives the improvement of infrastructure construction and agricultural production capacity and promotes the decline of methane emissions.In this paper, (agr) is used to represent the per capita gross output value of agriculture, forestry, animal husbandry and fishery.Residents' living conditions.There is an "inverted U-shaped" relationship between residents' living conditions and agricultural methane emissions.With the improvement of residents' living conditions, a greener and better ecological environment is needed.Therefore, there is a negative relationship between them, which is represented by (inc), the per capita income of rural residents.Industrial structure.Methane emissions come from planting and breeding.Therefore, the internal structure of the agricultural industry will affect methane emissions, which is represented by (ind), the proportion of planting industry in the total output value of agriculture, forestry, animal husbandry, and fishery.Urbanization level.Urbanization will occupy agricultural land, thus reducing agricultural methane emissions, which is measured by the proportion of urban population in total population and represented by (ubr).Industrialization level.With the development of industrialization, the emergence of green technology will reduce methane emissions in the process of agricultural production, which is represented by (iind), the proportion of non-agricultural total output value.Environmental regulations.Environmental regulations will have an impact on agricultural Agricultural development level.There is a double-sided effect between agricultural development level and agricultural methane emissions.The improvement of agricultural development level is often accompanied by the expansion of production scale or structural adjustment, which may produce more methane emissions in this process.On the contrary, the improvement of agricultural development level drives the improvement of infrastructure construction and agricultural production capacity and promotes the decline of methane emissions.In this paper, (agr) is used to represent the per capita gross output value of agriculture, forestry, animal husbandry and fishery.Residents' living conditions.There is an "inverted U-shaped" relationship between residents' living conditions and agricultural methane emissions.With the improvement of residents' living conditions, a greener and better ecological environment is needed.Therefore, there is a negative relationship between them, which is represented by (inc), the per capita income of rural residents.Industrial structure.Methane emissions come from planting and breeding.Therefore, the internal structure of the agricultural industry will affect methane emissions, which is represented by (ind), the proportion of planting industry in the total output value of agriculture, forestry, animal husbandry, and fishery.Urbanization level.Urbanization will occupy agricultural land, thus reducing agricultural methane emissions, which is measured by the proportion of urban population in total population and represented by (ubr).Industrialization level.With the development of industrialization, the emergence of green technology will reduce methane emissions in the process of agricultural production, which is represented by (iind), the proportion of non-agricultural total output value.Environmental regulations.Environmental regulations will have an impact on agricultural production and will encourage farmers engaged in agricultural production, especially aquaculture production, to adopt green technology and improve new technologies in the process of agricultural production.In this paper, (enr) is used to characterize the proportion of pollution control projects completed in each region in the year in the regional GDP.Agricultural machinery input.The input of agricultural machinery is mainly used to reduce methane emissions in agricultural planting.The input of great horsepower machinery can effectively improve planting efficiency, which is represented in this paper by (am), the total power of agricultural machinery.Agricultural science and technology R&D investment.Investment in agricultural science and technology R&D is the key to promoting agricultural technology progress.In this paper, (ati) is used to represent the proportion of government investment and financial support to agriculture in the total expenditure of provincial researchers.

Analysis of Influencing Factors of Spatial Differences of Agricultural Methane
On the basis of the geographical detector, this paper takes the three grain production areas as the main hierarchical method, uses the natural breakpoint method to treat variables by stratification, and divides the impact factors into four categories(Table 4).It can be seen that 9 factors play a driving role in the spatial differentiation of methane emissions, and that the influencing factors are divided into key and secondary influence factors according to the size of the driving factor explanatory power (q-statistical value).Among them, spatial factors (spa), industrial structure status (ind), industrialization level (iind), and agricultural science and technology R&D investment (ati) are the key influencing factors; agricultural development level (agr), residents' living conditions (inc), urbanization level (ubr), environmental regulation (enr), and agricultural machinery input (am) are the secondary influencing factors.The explanatory power of spatial factors reaches 0.438, which is the most key influencing factor, indicating that the production differences in different grain production zones played a decisive role in agricultural methane emissions.The explanatory power of industrial structure, industrialization level, and agricultural R&D investment are 0.122, 0.241, and 0.206, respectively.These variables represent the driving effect of agricultural transformation and upgrading on agricultural methane reduction from multiple aspects.In addition, the explanatory power of residents' living conditions, urbanization level, environmental regulation, and agricultural machinery input are all above 0.03, which cannot be ignored even though it is not the most critical factor influencing regional differences during the investigation period.Agricultural science and technology R&D investment has an impact on the spatial distribution of the three regions.It provides important technical support for methane emission reduction, and continuously strengthening the government's investment in agricultural science and technology is an important driver to achieve green development.The key influencing factors of main producing area, main sales area, and balance area are different(Table 5).Specifically, for the main producing area, the industrial structure is the most critical factor, and its explanatory power can reach 0.549.Main producing areas undertake the due obligations to ensure food security.Under the safety bottom line, adjusting the proportion between agriculture, forestry, animal husbandry, and fishery, especially the reasonable industrial structure adjustment process under the guidance of nutrition, will play an important role in methane emission reduction.The explanatory power of industrialization level and agricultural machinery input is also above 0.08, which is the main influencing factor of the spatial distribution of agricultural methane emission in main producing areas.For the main sales area, agricultural development level is the most important factor, and its explanatory power reaches 0.292.The level of agricultural development in main sales area will affect the income of residents and their demand for a better life.The better the agricultural development level, the stronger the demand for green and lowcarbon products, which will reduce the methane emissions.The explanatory power of residents' living standards and agricultural mechanization input is also higher than 0.1, indicating that these factors are also important factors influencing the spatial distribution of agricultural methane.
For the balance area, the industrial structure is the main influencing factor with the explanatory power of 0.479, and the industrialization level is also the main influencing factor with the explanatory power of 0.545.The evolution of industrial structure is that the transformation from planting to the integration of primary, secondary, and tertiary industries, which will effectively reduce agricultural methane emissions.The acceleration of industrialization will shift the rural labor force and release a greater demand to replace chemical fertilizers with organic fertilizer.Agricultural development level and investment in agricultural science and technology R&D are also key factors affecting the spatial distribution of balance area.

Discussion
Agriculture is both a source of greenhouse gas emissions and one of the most sensitive sectors to the impact of temperature change.In recent years, agricultural methane emission reduction has achieved some success, but the impact of greenhouse gas emissions caused by agricultural activities on climate change cannot be ignored.The year 2019 was the fifth warmest year since 1951, the national average temperature was 0.79 • C higher than usual, the average number of high temperature days was 11.8 days, with 4.1 days more than usual, compared with the second warmest year.Methane is the second largest greenhouse gas, and the recyclability of methane shows that controlling agricultural methane emission is an effective way to achieve the dual carbon goal.From the mechanism of agricultural methane emission, the formation ways of agricultural methane include paddy field emissions from the planting industry, livestock and poultry manure emission from the breeding industry, and livestock and poultry intestinal fermentation emission.Among them, the emission of livestock and poultry intestinal fermentation is the main emission source, accounting for 51.55-60.71%.Therefore, it is important to focus on controlling the emission of livestock and poultry intestinal fermentation to control agricultural methane emission, adjust the livestock breeding structure according to the actual situation of each region, and promote standardized breeding is an effective means.Paddy field emissions from planting industry is the second largest emission source after livestock and poultry intestinal fermentation, accounting for 26.45-35.54%.As the largest paddy producing country, China should choose appropriate paddy varieties and cultivation methods, promote large-scale planting, improve the utilization capacity of straw returning to the field, and other measures to reduce the emission coefficient of paddies.Compared with the previous literature, this paper adds the analysis of grain functional areas, uses nuclear density to investigate the differences and changes, utilizes geographic detectors to analyze the influencing factors, and then puts forward targeted emission reduction measures and countermeasures.
On the basis of the above analysis, the countermeasures to reduce agriculture methane emissions are provided as follows.
Firstly, it is important to carry out spatial cooperation and explore the spatiotemporal absorptive mechanism of agricultural methane emission reduction.The administrative boundary should be broken, the spatial association network should be constructed, and the free flow of elements (manpower, capital, etc.) in space should be realized.In addition, the temporal and spatial benefit protection and incentive policy support system needs to be constructed.Initially, it is necessary to clarify the spatial constraint characteristics of agricultural development, especially the differences between natural resources endowments.From the national level, legal, economic, and social tools should be adopted to provide institutional guarantee for the coordination mechanism of cross-regional green development of agriculture.Then, a national cross-regional cooperation platform should be established.Consistent with the development of the national carbon market, the platform of methane emission reduction also needs to be built.In terms of agricultural methane emission reduction, attention should be paid to reducing the differences between regions, such as setting up a regional agricultural methane emission reduction fund, to form a platform for cooperation and exchange between regions.
Secondly, it is crucial to strengthen the research and development of green and lowcarbon agricultural technologies.The investment in agricultural science and technology has a significant impact on the spatial distribution of agricultural methane in the three grain functional areas; therefore, increasing investment in agricultural science and technology research and development is the most effective measure to form China's unique development pattern of high-value agriculture and low-carbon agriculture.In the input, the role of the promising government and the effective market shall be highlighted, social capital shall be guided into basic research and development, and an integrated guidance mechanism of breeding, reproduction, and promotion shall be established.In the output, the enthusiasm of various agricultural business entities needs to be stimulated, and a new pattern of green development should be formed by combining key technologies such as big data and smart agriculture.In addition, methane can be further utilized by recycling.The liquid storage process can be reduced for manure with high emission potential, and methane can be recovered through anaerobic fermentation to reduce greenhouse gas emissions.Methane can be recycled through the construction of biogas projects, and emissions can be reduced by changing the wet manure into dry manure and changing the way of manure is stored by covering.
Thirdly, the mechanism of reducing agricultural methane emission by time, land, and classification should be formed.For main producing area and balance area, emphasis should be placed on the optimization of industrial structure, especially on the premise of ensuring food security to promote the green transformation of industrial structure.The adjustment of industrial structure should be carried out from the aspects of spatial distribution, product structure, and industrial chain upgrading.The spatial layout should mainly highlight local characteristics, optimize the spatial layout, take into account variety improvement and quality improvement in product structure, improve regional quality agricultural products, solve the problem of "high quality but not good price", upgrade the industrial chain to achieve vertical extension, and promote advanced planting and breeding, advanced planting and fishing, ecological circular agriculture development, and realize the integrated development of primary, secondary, and tertiary industries.For the main grain sales area, the leading function of agricultural green transformation should be brought into play, and residents' awareness of green agricultural products should be initiated to drive the improvement and promotion of the entire agricultural industry chain.
This paper selects the emission coefficient published by IPCC for calculation, which is a widely used research method; however, due to the different actual conditions in various regions of the country, there is a problem of low accuracy in the face of the change of emission coefficient.At the same time, this paper lacks specific analysis at the provincial level, which is the direction of improvement in the future.

Conclusions
On the basis of the construction of the agricultural methane measurement method, the total amount and structure of the agricultural methane emissions in China is calculated and the spatiotemporal differentiation features of the agricultural methane emission through the grain production areas are analyzed.Then, through the interpretation of the kernel density function curve, distribution location, sample state, and kurtosis extension-the key factors affecting the spatial differentiation of agricultural methane emission-are analyzed by using geographic detectors.The main conclusions are as follows.
Firstly, agricultural methane emissions showed a trend of increasing first and then decreasing from 2000 to 2019; agricultural methane emissions showed a downward trend in 2006 and have been stable since then.During the 13th Five-Year Plan, the greening level was further improved, and methane emissions declined steadily.Agricultural methane emission reduction is closely related to the green development policy.
Secondly, the characteristics of agricultural methane emissions in the three major grain areas are basically the same as those in the whole country, but the changes in the main producing area, main sales area, and balance area have their own characteristics.From 2000 to 2019, the main reason for the reduction of agricultural methane emissions in the main producing area was the decline of intestinal fermentation emissions of livestock and poultry, which exceeded the increase of emissions from paddy fields.The paddy field, livestock and poultry intestinal fermentation, and manure management emissions in the main sales area and balance area fell, which generally reflects that the overall effect of green development in China is gradually emerging.At the provincial level, emphasis should be placed on reducing agricultural methane emissions in major producing provinces such as Hunan, Sichuan, Inner Mongolia, Heilongjiang, and Jiangxi.
Thirdly, the spatial patterns of agricultural methane emissions in China will remain unchanged in the short term, and it is difficult to achieve convergence.
Fourthly, the production heterogeneity of grain production areas has a decisive impact on the difference in agricultural methane emissions.Industrial structure, the level of industrialization, the investment in agricultural science, and technology R&D are the key influencing factors.Agricultural development level, residents' living conditions, urbanization level, environmental regulation, and agricultural machinery input are secondary influencing factors.In particular, the investment in agricultural science and technology R&D is influential in the agricultural methane emission pattern in the three main grain areas.For the main grain sales area, the agricultural development level is the core driving factor for the main sales area, and the industrialization level and industrial structure are the main driving factors for the balance area.

1 .
Measurement of Agricultural Methane Emission in China 3.1.1.Overall Agricultural Methane Sequential Variation Agricultural methane emissions showed a trend of first increasing and then decreasing.From 2000 to 2019 (Figure 1), agricultural methane emissions decreased from 21.46 million tons to 17.69 million tons, with an average annual drop of 0.88%.Overall, China's green development in agriculture has been significantly improved.From 2000 to 2005, agricultural methane emissions continued to increase, with a significant decline in 2006.

3. 1 . 1 .
Overall Agricultural Methane Sequential Variation Agricultural methane emissions showed a trend of first increasing and then decrea ing.From 2000 to 2019 (Figure 1), agricultural methane emissions decreased from 21. million tons to 17.69 million tons, with an average annual drop of 0.88%.Overall, Chin green development in agriculture has been significantly improved.From 2000 to 200 agricultural methane emissions continued to increase, with a significant decline in 2006
3.2.1.Model Building of Influencing FactorsReferring to relevant references, the factors affecting agricultural methane emissions include agricultural development level, residents' living conditions, industrial structure, urbanization level, industrialization level, environmental regulation, agricultural machinery input, and agricultural science and technology R&D investment, etc. Agriculture 2022, 12, x FOR PEER REVIEW 10 of 18 effectively reduce methane emissions.The agricultural methane emissions of Tianjin, Beijing, Shanghai, and other Eastern developed provinces were relatively small and remained at the bottom of the emission proportion in 2000, 2010, and 2019.The reason lay in the composition of industrial structure.For example, in 2019, the proportion of the first industry in the GDP of the three provinces was 1.3%, 0.3%, and 0.3%, respectively.The number of livestock and poultry in Ningxia and Hainan was low; in 2019, the number of livestock and poultry was 20.707 million and 63.913 million, respectively, far lower than the national average of 233.2885 million, so the methane emission was low.Liaoning and Jilin, also major grain-producing regions in Northeast China, had relatively low emissions, ranking lower (20th and 21st, respectively, in 2019).Compared with Heilongjiang, the two provinces had a lower paddy sown area, with only 22% and 13%, respectively, than Heilongjiang in 2019.In addition, among the main producing areas, Hebei accounted for 4.05% of emissions in 2000, and the number of livestock and poultry decreased by 37.26% in 2010 compared with 2000, which was the main reason for its low emissions, ranking 18th in 2019.From the perspective of emission sources, in 2019, the agricultural methane emission in 12 of the 31 provinces was dominated by paddy field emissions, represented by Jiangsu (82.73%), and the remaining 19 provinces were dominated by intestinal fermentation emissions, represented by Qinghai, Gansu, Tibet, Xinjiang, and Inner Mongolia, which accounted for more than 90%, and the intestinal fermentation emissions caused by them reached 18.57% of the national total emissions.(a)

Table 1 .
Methane emission factors from paddy fields.

Table 2 .
Methane emission factors from manure management.

Table 3 .
Methane emission factor from intestinal fermentation of livestock and poultry.

Table 4 .
Detection results of influencing factors of agricultural methane emissions in China.

Table 5 .
Factor detection results of methane emission in different production areas.