Abstract: Based on Chinas provincial panel data, this paper calculates the total carbon emission from cultivated land use in China, analyzes its temporal and spatial evolution characteristics, explores the relationship between carbon emission from cultivated land use and grain yield using cointegration analysis and Granger causality test, and reveals the driving factors of carbon emission from cultivated land use using the STIRPAT model and geographic weighted regression model (GWR). The study found that Chinas carbon emissions from cultivated land use generally showed an upward trend but entered a negative growth stage in 2017. The proportions of the four types of carbon emissions were waste treatment carbon emissions (48.165%), production factor input carbon emissions (46.924%), cultivated land planting carbon emissions (3.513%), and natural source carbon emissions (1.416%). At the same time, the spatial correlation test showed that the carbon emission from cultivated land use in China had significant regional aggregation characteristics. The carbon emission structure of cultivated land use in China was dominated by the carbon emission of production factor input and waste, and there were great differences in the types of carbon emission of cultivated land use among provinces. There was a longterm equilibrium relationship between the total carbon emission of cultivated land use and grain yield in China, and there was a oneway causal relationship between the carbon emission of cultivated land use to the growth of grain yield. The total rural population, the disposable income of rural residents, and the level of agricultural machinery all had a positive impact on the carbon emission of cultivated land use, and the influence coefficients are 0.624, 0.163, and 0.337, respectively. The regression coefficients of various influencing factors showed an obvious step distribution in space. The results showed that there were obvious temporal and spatial differences in carbon emissions from cultivated land use in China, and Chinas agricultural production still has the characteristics of high input, high output, and high pollution. Differentiated carbon reduction policies should be implemented according to local conditions based on the actual situation of all provinces (cities).

Key words: cultivated land utilization, carbon emissions, temporal and spatial differences, cointegration analysis, GWR model

摘要： 基于中国省级面板数据，对中国耕地利用碳排放总量进行核算，分析其时空演化特征，运用协整分析、Granger因果关系检验探究耕地利用碳排放与粮食产量间关系，运用STIRPAT模型及地理加权回归模型(GWR)揭示耕地利用碳排放的驱动因素。研究发现：中国耕地利用碳排放总体呈上升趋势，但自2017年起进入负增长阶段，四种碳排放类型所占比重从大到小依次为废弃物处理碳排放(48.165%)、生产要素投入碳排放(46.924%)、耕地种植碳排放(3.513%)、自然源碳排放(1.416%)；与此同时，空间相关性检验表明中国耕地利用碳排放具有显著的区域聚集特征；中国耕地利用碳排放结构上以生产要素投入碳排放和废弃物碳排放为主，且省际间的耕地利用碳排放类型存在较大差异；中国耕地利用碳排放总量与粮食产量之间存在长期均衡关系，同时存在着从耕地利用碳排放到粮食产量增长的单向因果关系；农村总人口数、农村居民可支配收入、农业机械水平均对耕地利用碳排放均有正向影响，影响系数分别为0.624、0.163、0.337，且各影响因素的回归系数在空间上呈现出明显的阶梯分布。研究结果表明：中国耕地利用碳排放时空差异特征明显，且我国农业生产仍存在高投入、高产出、高污染的特点，应基于各省(市)的实际情况，因地制宜地实施差异化的降碳政策。

关键词: 耕地利用, 碳排放, 时空差异, 协整分析, GWR模型