Spatial And Temporal Distribution Of XCO₂ And Analysis Of Its Influencing Factors In China

China is one of the major emitters of carbon dioxide.As of 2019,its emissions reached 10.175 billion tons.In the 14th Five-Year Plan,for the first time,my country has included the goals of"carbon peak"and"carbon neutrality"into the five-year plan.This paper studies the temporal and spatial distribution of CO₂ in China and the contribution of its influencing factors,hoping to provide a data reference for the monitoring and evaluation of China’s carbon emissions.In this paper,the column-averaged CO₂ dry air mole fraction（XCO₂）observation data（2014-2018）of the OCO-2 satellite is used to extract the first mode of XCO₂ space-time distribution in China,and its influencing factors and sensitivity are singled out.Factor influence analysis and multi-factor cross-effect analysis,including machine learning algorithms,global sensitivity and cumulative local effect analysis,realize regional CO₂ temporal and spatial distribution modeling and quantitative evaluation of the importance of various influencing factors.The main research work and results obtained in this paper are as followed:（1）XCO₂ spatio-temporal distribution:The first mode of China’s CO₂spatiotemporal distribution from 2014 to 2018 was extracted using the empirical orthogonal decomposition method,including the spatial distribution gradient,annual change trend,and seasonal pattern.It was found that China’s XCO₂ existed in the spatial distribution during the study period.The obvious east-west gradient difference shows an upward trend from west to east,with a gradient difference of 3.91 ppm.In terms of time distribution,there is an obvious trend of rising year by year and seasonal fluctuation:the annual rising trend is 2.56ppm/a,which is higher than the global average growth rate of 2.07ppm/a from 2014 to 2018;the average seasonal amplitude is 3.89ppm/a.（2）Single-factor impact analysis of CO₂ distribution in China,including the degree of impact,distribution range,and seasonal driving.The analysis of the influence of XCO₂ on vegetation,temperature and precipitation,biomass burning,and anthropogenic emissions shows that NDVI and XCO₂ are negatively correlated（r=-0.47,p<0.05）.Stronger vegetation coverage occurs in the northeast region.Negative correlation（r=-0.58）,and when vegetation is active（April-October）,the correlation is significantly enhanced,increasing by 17.5%;there is a significant negative correlation between precipitation and temperature and CO₂,and there is a certain time-lag effect:The national precipitation,temperature,and CO₂ delays fluctuate generally in 1-3months.The average precipitation delay is 1.6 months,and the temperature is 2.1months.Combustion emission factors are positively correlated with the spatial distribution of CO₂:The correlation coefficient r of material combustion in areas with strong emissions（such as Yunnan and Northeast）can reach 0.382;the correlation coefficient of anthropogenic emissions r=0.397,and the correlation coefficient can reach 0.791 in areas with strong anthropogenic emissions（emissions>10³t）.（3）XCO₂ growth trend forecast model of China:before the multi-factor cross-analysis,the national XCO2 growth trend is modeled using machine learning XGBoost.Take temperature,precipitation,NDVI,anthropogenic emissions,and biomass combustion emissions as model inputs to construct a national carbon growth model.The RMSE of the modeling result is 0.0685ppm,and the R² reaches 0.9618.The regions with large XCO₂ growth rate in the country are concentrated in eastern China,and the maximum is in Shanghai,at 4.859 ppm/a;the regions with small growth rate are concentrated in the Qinghai-Tibet Plateau.The minimum value is 1.848ppm/a.（4）Multi-factor cross-effect analysis that affects China’s CO₂ distribution:Based on the construction of China’s XCO₂ growth trend prediction model,Sobol sensitivity analysis is used to estimate the main effects of multiple influencing factors and the overall sensitivity index,and to quantitatively evaluate the importance of influencing factors.The results show that temperature,anthropogenic emissions and vegetation are the main influencing factors affecting the growth of XCO₂ in China.The main effects are 0.3967,0.3666 and 0.1387,and the overall sensitivity index is 0.4224,0.3938 and0.1549,and the overall contribution is 40.14%,respectively.37.43%and 14.73%.The cumulative local effect is used to quantify the synergistic changes between the main influencing factors and the CO₂ growth.The annual average temperature and the regional CO₂ growth rate show a synergistic upward trend,and before 3.4℃,the annual average temperature rise has little impact on the regional CO₂ change.When the temperature increases within the range of 3.4°C-13.9°C,the regional CO₂ rises rapidly;anthropogenic emissions also show a similar upward trend of synergy.When the emissions change within 0-2.4×106t,as the anthropogenic emissions increase,the regional CO₂ Shows a rapid upward trend;vegetation shows a negative correlation trend.After the regional average NDVI value is greater than 0.07,as the vegetation coverage increases,the regional CO₂ shows a clear downward trend.