Spatial-Temporal Characteristics of Agricultural Greenhouse Gases Emissions of the Main Stream Area of the Yellow River Basin in Gansu, China

: In 2021, The People’s Republic of China proposed goals for peaking carbon dioxide emissions before 2030 and carbon neutrality before 2060, in the 15 counties (districts) of the Main Stream Area of the Yellow River Basin in Gansu that plays an important role in ecological protection and green development. Next the CO 2 equivalents were converted according to the IPCC2 standard, the total agricultural GHG emissions was calculated, the relationship with the agricultural output value was analyzed, and the discretization of the space was analyzed by the coefﬁcient of variation and standard deviation. Firstly, the total agricultural GHG emissions in 15 counties (districts) of the Main Stream Area of the Yellow River Basin increased 55.54% in 2000–2019, and 2.35% annually, roughly divided into three stages: the rapid growth period (2000–2008), the slow decline period (2009–2014) and the rapid decline period (2015–2019). The economic efﬁciency is signiﬁcantly improved, with an average annual decline of 6.49%, roughly divided into three stages: the slow-descent stage (2000–2004), the period of slow-growth stage (2005–2008) and the period of fast-decline (2009–2019). Secondly, based on the characteristics of the total GHG emissions, Maqu County has the largest GHG emissions increase, from 26.8842 kt in 2000 to 38.9603 kt, in 2019, an increase of 44.92%, while the smallest GHG emissions, in Anning District, decreased 87.33% from 111 t in 2000 to 14.1 t in 2019; In the rate of increase in the total GHG emissions, Dongxiang County had the largest rate of increase from 2000 to 2019, an increase of 160.28% and an average annual increase of 4.90%. The smallest rate of decrease in GHG emissions was seen in Chengguan District, where they decreased 92.11% from 2000 to 2019, an average annual decrease of 11.93%. The characteristics of agricultural GHG emissions intensity signiﬁcant efﬁciency improved. Finally, to provide a basis for the formulation of differentiated agricultural energy conservation and emissions reduction policies, reduce agricultural GHG emissions intensity and reduce the use efﬁciency of resources by formulating differentiated emission targets, tasks and incentive measures.


Introduction
Agricultural production is the world's second-largest greenhouse gas (collectively referred to in this article simply as GHG) emissions source [1]. In 2021, The People's Republic of China proposed the goals for peaking carbon dioxide emissions before 2030 and carbon neutrality before 2060. China is a traditional agricultural country, still in the stage of extensive development. The GHG produced by agricultural activities account

Emission Quantity and Intensity Measurement Method
This paper refers to the calculation and research results of Zhang et al. [26], Chen et al. [24], Shang et al. [27], Li et al. [28], Kang et al. [25] and West [29]. Kumar et al. [30] in particular reviewed GHG measurements from Chinese freshwater bodies and the GHG emissions calculation methods and factors of emissions in the 2011 Provincial GHG List Preparation Guide (Trial) and calculated the agricultural GHG emissions in the 15 counties (districts), mainly including the CO 2 , CH 4 , N 2 O: (1) The calculation of agricultural GHG CO 2 is primarily the CO 2 emissions from fertilizer use, agricultural machinery and electricity during agricultural production in the region: where CO 2 is the total GHG emissions of agricultural CO 2 , F is chemical fertilizer use, Area is agricultural planting area, Machine is the total agricultural machinery power, Electric is agricultural production electricity consumption, formula is k 1 = 857.54, k 2 = 16.47, k 3 = 0.18, k 4 = 0.18.
(2) Agricultural GHG CH 4 , which mainly includes emissions of CH 4 from paddy fields and CH 4 from storage animals: CH 4−paddy fields = EF paddy fields × Area paddy fields where CH 4−paddy fields is the CH 4 emissions of rice field, which is mainly single-season rice in northwest China, while EF paddy fields is 231.2 kg/hm 2 , Area paddy fields is the annual rice planting area. CH 4−animals is the CH 4 emissions of animals, and α 1 , β 1 is the CH 4 emissions coefficient of animal gut and fecal fermentation respectively. The coefficient of GHG emissions is mainly calculated according to the correlation coefficient of the provincial level.
animals animals are the animal N 2 O emissions coefficient, T i is the feeding amount of class i, and γ i is the N 2 O coefficient of fecal fermentation of class i. Due to the difficulty of nitrogen application and the large area of fertilizer use, this paper calculates the N 2 O calculation excluding farmland soil. N 2 O agricultural cultivation is the N 2 O emissions of farmland cultivation, T i is the yield of Class i agriculture, S i is the straw coefficient of Class i agricultural crops, Straw conversion coefficient was determined according to the United Nations Food and Agricultural Organization (FAO); assuming that all straw is returned to the field through agricultural production and livestock breeding.
(4) G is the total CO 2 emissions of CO 2 , mainly the total overall regional carbon emissions: According to the IPCC's National GHG Inventory Guide 2006, a global warming submersible meter radiation forces a ton of GHG on a ton of carbon dioxide for some time. Therefore, were all used, 1 t CH 4 = 25 t CO 2 , 1 t N 2 O = 298 t CO 2 .
Based on the basis of relevant research results, in order to further measure the relationship between agricultural GHG emissions and agricultural economic development in the main stream region of the Yellow River Basin and 15 counties, the relationship between the emissions and the agricultural economic level is measured by the way of GHG emissions intensity. Agricultural GHG emissions intensity was calculated as: where G is the agricultural GHG emissions (10,000 t); A is the total agricultural output value (10,000 yuan).

Standard Deviation and Coefficient of Variation
To illustrate the differences seen in the counties of agricultural GHG emissions in 15 counties of the Main Stream region of the Yellow River Basin, the standard deviation (S) and coefficient of variation (V) usually measure relative and absolute differences, with larger values indicating greater difference: where G i is the emissions (or emissions intensity) of the i county/district unit in the main stream of the Yellow River Basin; n is the number of county units; and G is the average of discharge of n county units.

Spatial Variability Analysis
The Moran' I statistic is a very widely used spatial autocorrelation statistic, in its specific form as follows: where, x i represents the observant i at the i-th spatial position, x = 1 n n ∑ i=1 x i , w ij is the element of the spatial weight matrix W (n × n) representing the topological relationship between spatial units and S 0 is the sum of all elements of the spatial weight matrix W.
Reflecting is the degree of similarity of emissions regional cell attribute values in spatial or spatial proximity. This article mainly uses the global Moran 'I and spatial distribution characteristics of total agricultural GHG emissions and GHG emissions intensity in the main stream of 15 counties in the Main Stream of the Yellow River Basin in the Gansu section.  "G" represents the total GHG emissions and "CI" represents the intensity of agricultural GHG in Figure 2. It shows the relationship of change between total carbon emissions and economic development.

Different Types of Agricultural GHG Analysis
The total emissions in different types of agricultural GHG gas analysis, it's mainly presents the following characteristics:  an inter significant downward trend in the 15 counties in the Main Stream area of the Yellow River Basin. With an average annual decline of 6.49%, indicating that its economic efficiency is significantly improved. It can mainly be divided into the following three stages ( Figure 2): (1) During the period of slow-descent stage (2000)(2001)(2002)(2003)(2004), agricultural GHG emissions intensity from 0.016 to 0.011, down by 31.25% in this period, with an average annual decline of 9%. It shows that the agricultural production efficiency shows a higher growth trend.

Different Types of Agricultural GHG Analysis
The total emissions in different types of agricultural GHG gas analysis, it's mainly presents the following characteristics:  Figure 3). Except for Maqu County, the total carbon emissions in the remaining 15 counties tend to spatial agglomeration.

Analysis of Total GHG Emissions
From the characteristics of the total GHG emissions in the Main Stream part of the Gansu section of the Yellow River Basin, Maqu County had the largest total carbon emissions, Anning District was the smallest in the total carbon emissions, Dongxiang County had the largest increase in the total GHG emissions, and Chengguan District decreased the most in the total GHG emissions. Details are shown below: In the total GHG emissions, Maqu County has the largest GHG emissions increase, from 26.8842 kt in 2000 to 38.9603 kt, in 2019, an increased by 44.92%, and an average annual increase of 1.87%, far below the average growth rate of 15 counties/districts (55.54%), and a growth rate of 2.23% in 15 counties/districts. The smallest GHG emissions in Anning District decreased 87.33% from 111 t in 2000 to 14.1 t in 2019, and an average annual decrease of 9.81%, far below the average growth rate of 15 counties/districts, and a growth rate of 2.23% in 15 counties/districts.
In rate of increase in the total GHG emissions, Dongxiang County has the largest rate of increase in GHG emissions from 3.4106 kt in 2000 to 8.8771 kt, in 2019, increased by 160.28%, and an average annual increase of 4.90%, far over the average growth rate of 15 counties/districts, and a growth rate of 2.23% in 15 counties/districts. The smallest rate of decrease in GHG emissions in Chengguan District decreased 92.11% from 0.4893 kt in 2000 to 0.0390 kt in 2019, and an average annual decrease of 11.93%, far below the average growth rate of 15 counties/districts, and a growth rate of 2.23% in 15 Figure 5).    Figure 5). "C/D" represent "county (district)". The units in the table is "kt", it represents "kiloton".

Analysis of GHG Emissions Intensity
The characteristics of agricultural GHG emissions intensity in the Main Stream of the Yellow River Basin in Gansu section is a significant decline trending, agricultural production efficiency has been significantly improved. Maqu County was the largest in the agricultural GHG emissions intensity, Anning District was the smallest in the agricultural GHG emissions intensity, Chengguan District had the largest decrease in the agricultural GHG emissions intensity, and Anning District decreased the least in the agricultural GHG emissions intensity. See below for details: In  .85%, and an average annual decrease of 7.85%, 7.42%, 7.37%, 6.68%, 6.37%, 6.15%, 6.05%, 5.78%, 4.67%, 4.03%, 3.28%, 3.20%. The increase was relatively small in Baiyin District, Xigu District, Yuzhong District, and Jishishan County, but the agricultural GHG emissions in Gaolan County were the smallest, from 1.5467 thousand tons in 2000 to 1.5434 thousand tons, a decrease by 0.22%, with an annual decline of 0.01%. (Table 3 and Figure 5). Further analysis of the growth rate finds that the agricultural production efficiency has improved significantly, and total GHG emissions have been greatly improved. From the growth rate of total agricultural GHG in the Main Stream area of the Yellow River Basin in the Gansu section from 2000 to 2019, the growth rate is above the average rate in Linxia County, Yongjing County, Dongxiang County, Qilihe District, Jingtai County, Pingchuan District and Jingyuan County. These counties (districts) averaged a 98.75% increase, while other counties (districts) are below the average growth rate of 55.54%. From the growth rate of the agricultural GHG emissions intensity in the Main Stream region of the Yellow River Basin in the Gansu section from 2000 to 2019, the growth rate is above the average rate in Yongjing County, Dongxiang County, Qilihe District, Anning District, Xigu District, Jingtai County, Pingchuan District, Baiyin District. These counties (districts) averaged a −58.43% increase, Other counties (districts) are below the average growth rate of −72.71% ( Figure 6). Further analysis of the growth rate finds that the agricultural production efficiency has improved significantly, and total GHG emissions have been greatly improved. From the growth rate of total agricultural GHG in the Main Stream area of the Yellow River Basin in the Gansu section from 2000 to 2019, the growth rate is above the average rate in Linxia County, Yongjing County, Dongxiang County, Qilihe District, Jingtai County, Pingchuan District and Jingyuan County. These counties (districts) averaged a 98.75% increase, while other counties (districts) are below the average growth rate of 55.54%. From the growth rate of the agricultural GHG emissions intensity in the Main Stream region of the Yellow River Basin in the Gansu section from 2000 to 2019, the growth rate is above the average rate in Yongjing County, Dongxiang County, Qilihe District, Anning District, Xigu District, Jingtai County, Pingchuan District, Baiyin District. These counties (districts) averaged a −58.43% increase, Other counties (districts) are below the average growth rate of −72.71% ( Figure 6).  Further analysis of the growth rate finds that the agricultural production efficiency has improved significantly, and total GHG emissions have been greatly improved. From the growth rate of total agricultural GHG in the Main Stream area of the Yellow River Basin in the Gansu section from 2000 to 2019, the growth rate is above the average rate in Linxia County, Yongjing County, Dongxiang County, Qilihe District, Jingtai County, Pingchuan District and Jingyuan County. These counties (districts) averaged a 98.75% increase, while other counties (districts) are below the average growth rate of 55.54%. From the growth rate of the agricultural GHG emissions intensity in the Main Stream region of the Yellow River Basin in the Gansu section from 2000 to 2019, the growth rate is above the average rate in Yongjing County, Dongxiang County, Qilihe District, Anning District, Xigu District, Jingtai County, Pingchuan District, Baiyin District. These counties (districts) averaged a −58.43% increase, Other counties (districts) are below the average growth rate of −72.71% ( Figure 6).

Spatial Difference Analysis of Total GHG Emissions and GHG Emissions Intensity
The relative gap in total GHG emissions gradually widened, while the absolute gap gradually narrowed and basically stable. absolute gap between counties. (4) From 2014 to 2017 was a stable transition period, and the standard deviation and coefficient of variation remained basically unchanged, indicating that the relative and absolute gap between counties remained basically stable. (5) From 2018 to 2019 was a rapid decline period, with the standard deviation and coefficient of variation decreasing 7.40% and 7.14%, respectively, indicating a gradual narrowing of the relative and absolute gap between counties.
The trend of GHG emissions intensity and absolute gap between counties showed a narrowing trend, but the absolute gap is relatively stable (Figure 7).The emission standard deviation between counties decreased from 2000-2019 to 83.45% from 0.0597 to 0.0099 and the coefficient of variation to 22.8217 from 2.3374 to 22.06%.It can be divided into four stages: from 2000-2007, for the fluctuation period, the standard deviation reached the highest value in 2001, at 0.0624; from 2008-2019, the gap between the two gradually expanded, indicating that the relative gap and absolute gap between counties are getting smaller and smaller, the relative gap is smaller than the absolute gap, but the absolute gap shows a relatively stable change trend. The trend of GHG emissions intensity and absolute gap between counties showed a narrowing trend, but the absolute gap is relatively stable (Figure 7).The emission standard deviation between counties decreased from 2000-2019 to 83.45% from 0.0597 to 0.0099 and the coefficient of variation to 22.8217 from 2.3374 to 22.06%.It can be divided into four stages: from 2000-2007, for the fluctuation period, the standard deviation reached the highest value in 2001, at 0.0624; from 2008-2019, the gap between the two gradually expanded, indicating that the relative gap and absolute gap between counties are getting smaller and smaller, the relative gap is smaller than the absolute gap, but the absolute gap shows a relatively stable change trend.

Discussion
Taking 15 counties in the Main Stream area of the Yellow River Basin in Gansu as the research subject, we analyze the spatial and temporal differences of agricultural GHG emissions by calculating the total GHG emissions, emissions intensity, Moran index, standard deviation and coefficient of variation. The total agricultural GHG emissions in the region are constantly increasing, while the intensity of agricultural GHG emissions continues to decline, and the agricultural production efficiency is significantly improved, the absolute gap gradually narrowed and basically stable. The trend of GHG emissions intensity and absolute gap between counties showed a narrowing trend, but the absolute gap is relatively stable. The main research conclusions are as follows: Firstly, from the perspective of the total agricultural GHG emissions in the study area, it occupies a small proportion in Gansu Province and represents a very small proportion in western China. The total agricultural GHG emissions in 15 counties in the Main Stream area of the Yellow River Basin increased significantly from 62.5773 kt in 2000 to 97.3355 kt, in 2019, and an average annual increase of 2.35%, for example, in 2017, the total agricultural GHG emissions in the study area was 100.2805 kt was 3.34% of 300.17 × 10 4 t [31] in Gansu Province, representing only 0.12% of the 8108.65 × 10 4 t [32] in Western China.
Similarly, compared with Jiangsu Province in eastern China, the values were far lower than the average annual growth rate of Jiangsu Province (4.37%) [24], which can be roughly divided into three stages: rapid growth period (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008), slow decline period (2009)(2010)(2011)(2012)(2013)(2014) and rapid decline period (2015-2019). The interannual change of emission intensity in Gansu section of the Yellow River Basin showed a significant downward trend, with an average annual decline of 6.49%, higher than an average annual decline of 4.26% in Jiangsu Province [24], it is also highly coupled to the fact that "western regions show a trend of increase before gradient reduction" [33], indicating that its economic efficiency is significantly improved. This has a similar trend to the single peak change of "the GHG footprint of chemical fertilizer application in China presents a single peak change of 'first increasing and then decreasing' in 2013" [34]. However, the total increase of agricultural GHG emissions from Gansu section in the Main Stream of the Yellow River Basin has decreased since 2008, and did not show an obvious trend of hump development. Considering the total emissions in the different types of agricultural GHG analysis, they mainly present the following characteristics: firstly, agricultural emissions of CO 2 show a smaller increase; secondly, emissions of CO 2 /CH 4 in paddy fields are the smallest in agricultural GHG, drop for 53.49 percent, the trend shows a smaller decline. The emissions of N 2 O in agricultural cultivation shows a smaller increase; thirdly, the emissions of CH 4 and N 2 O maintain a basically consistent change trend, but the average annual growth rate of the emissions of CH 4 is higher than that of N 2 O.
Secondly, the spatial differences and characteristics of the Gansu section of the Yellow River Basin are obvious. Maqu County was the largest in the total GHG emissions, from 26 County, Pingchuan District and Yongjing County was relatively large (increases by 160.28%, 117.53%, 114.99% and 106.45%). The agricultural GHG emissions in Gaolan County were the smallest, with an increase by 0.22%, and an annual decline of 0.01%. This is mainly related to the agricultural production conditions in Gansu Province [35]. The characteristics of agricultural GHG emissions intensity in the Main Stream of the Yellow River Basin in Gansu section showed an overall decline by 68.29%. The most occurred in Chengguan District, down by 84.25%. The least was in the Anning District, a decline by 40.40%. The agricultural production efficiency of 12 counties/districts (such as Maqu, Linxia, and others) has been significantly improved, but the improvement of agricultural production efficiency is only obvious in Qilihe District, Xigu Districts and Baiyin District. This is mainly related to agricultural land use efficiency level of around 0.7 and other factors have important relations [36].
Thirdly, from the perspective of regional GHG emissions intensity characteristics, the GHG emissions intensity in 15 counties (districts) of the Yellow River Basin, with an average decrease of 68.29%, and the relative gap in total GHG emissions gradually widened, while the absolute gap gradually narrowed and basically stable. Agricultural GHG emissions and agricultural economic development present three different types of elastic characteristics of negative, weak and strong decoupling [37], the path of decoupling between economic and social development and agricultural GHG emissions should be further considered, and the change of "the critical period of transformation from traditional farming mode to green and low-carbon farming mode should be properly handled" [38], and the GHG emissions intensity in Chengguan District decreased the most, by 84.25%; the GHG emissions intensity in Maqu County declined by 83.15%. Anning District's GHG emissions intensity fell the least, down 40.40%. The relative gap in total GHG emissions gradually widened, while the absolute gap gradually narrowed and was basically stable. The GHG emissions standard deviation increased from 657.97 to 963.38, up 46.42% and an annual increase of 3.89%; the coefficient of variation (absolute) decreased from 1.58 to 1.48, with an annual increase of 1.70%. It can be divided in five phases: first decline ( Fourthly, on the basis of analyzing the total emissions, emissions intensity and spatial differences of agricultural GHG emissions in 15 counties in the Main Stream area of the Yellow River Basin. It will provide a basis for the formulation of differentiated agricultural energy conservation and emission reduction policies. Firstly, the reduction of total GHG emissions is effectively realized through the adjustment of agricultural structure [39]. We can formulate different emission targets, tasks and incentive measures, determine different carbon reduction tasks and economic development indicators through GHG emissions intensity, explore regional agricultural carbon trading policies, and increase financial support and policy preference. To formulate differentiated ideas for modern agricultural development, Maqu County, mainly a single animal husbandry region, can enhance the agricultural production value and reduce the agricultural GHG emissions intensity by enhancing the attached value of animal husbandry production, Urban areas such as Chengguan District and Anning District can explore the "zero emission" mode of agricultural GHG emissions in the process of urbanization, The region combining urban areas and agricultural areas, mainly in Xigu and Baiyin District, will explore the balanced development mode of "carbon source-carbon sink" of agricultural GHG emissions and economic decoupling in the process of urbanization, Yuzhong County, Gaolan County and other planting counties mainly enhance the added value of planting industry, adjust the emissions reduction of agricultural machinery [40] and reduce the intensity of agricultural GHG emissions, improving the use of resources and reducing total GHG emissions.

Conclusions
This paper aims to research the GHG emissions and GHG emissions intensity in the Main Stream Area of the Yellow River Basin in Gansu, which includes CH 4 , CO 2 , and N 2 O emissions. Our results show that the total GHG emissions in this region increased between 2000−2019, but showed significant agricultural production efficiency and spatial differences. The study area represents a small proportion in Gansu Province and a very small proportion in Western China. The total agricultural GHG emissions in 15 counties in the Main Stream Area of the Yellow River Basin increased significantly from 62.5773 kt in 2000 to 97.3355 kt, in 2019, and an average annual increase of 2.35%. The relative gap in total GHG emissions gradually widened, while the absolute gap gradually narrowed and was basically stable. The spatial differences and characteristics of the Gansu section of the Yellow River Basin are obvious, Maqu County was the largest in the total GHG emissions, Anning District was the smallest in the total GHG emissions, Dongxiang County had the largest increase in the total GHG emissions, and Chengguan District decreased the largest in the total GHG emissions. However, the characteristics of agricultural GHG emissions intensity is a significant decline trending, agricultural production efficiency has been significantly improved. Maqu County was the largest in the agricultural GHG emissions intensity, Anning District was the smallest in the agricultural GHG emissions intensity, Chengguan District had the largest decrease in the agricultural GHG emissions intensity, and Anning District decreased the leaest in the agricultural GHG emissions intensity.
Calculated mainly based on IPCC's National GHG Inventory Guide 2006 and the 2011 Provincial GHG List Preparation Guide (Trial) to provide a typical case area for research in this field, and to propose countermeasures and suggestions for agricultural low-carbon emissions and economic sustainable development. However, in the process of research, statistical data are mainly used. If one needs to do more in-depth research, new technical means such as field research data should be used to ensure the accuracy and scientific nature of the results.