从种植业和畜牧业两方面入手，采用排放因子法对甘肃省2000—2020年农业碳排放进行了估算，分析了其时空变化特征，基于STIRPAT（Stochastic impacts by regression on PAT）模型探析了甘肃省农业碳排放的影响因素，并提出了相应对策。结果表明：（1） 甘肃省2000—2020年CO₂-e排放量呈“升高-降低-升高”的趋势，2015年达到峰值，估算为2320.41×10⁴ t；从2018年开始又逐年增加，直至2020年增至2290.69×10⁴ t。（2） 甘肃省农业CO₂-e排放结构中，种植业占35%，畜牧业占65%。主要碳排放源中，畜禽胃肠道发酵对农业碳排放总量的贡献最大，其次是化肥和畜禽粪便管理。主要畜禽中，肉牛养殖对碳排放的贡献最大，其次是绵羊、山羊、奶牛和猪，家禽养殖的贡献最小。（3） 农村人口、农村居民人均GDP、农村居民人均可支配收入、农业机械总动力、农业增加值占全省生产总值比重、农村住户固定资产投资额、农业科技成果应用数量、农业科技投入是影响甘肃省农业碳排放的主要因素，影响力指数分别为-0.017、0.026、0.020、0.038、-0.025、0.031、-0.017、0.016。为有效控制农业碳排放，建议在5个方面采取相应策略：努力提高种植业资源利用效率和土壤碳汇能力；强化畜牧业源头减量、过程控制和末端处理；努力降低农业机械对石油的依赖；有效推动农村清洁能源利用；加大农业低碳技术的研发与应用。

This study aimed to investigate aspects of the plant production and animal husbandry by employing the emission factor method to estimate agricultural carbon emissions and analyze their temporal and spatial variation features in Gansu Province, China, from 2000 to 2020. The factors influencing agricultural carbon emissions were investigated based on the stochastic impacts by regression on PAT (STIRPAT) model. Subsequently, corresponding countermeasures were proposed in this study. The results indicated the following: (1) The CO₂-e emissions from the agricultural sector of Gansu Province from 2000 to 2020 showed an increasing-decreasing-increasing trend; the emissions reached a peak in 2015, which was estimated to be 2320.41×10⁴ t. Additionally, from 2018 onward, the emissions increased annually to 2290.69×10⁴ t in 2020. (2) In Gansu Province’s agricultural CO₂-e emission structure, the plant production accounted for 35% and animal husbandry accounted for 65% of the total emissions. Among the various carbon sources, livestock and poultry gastrointestinal fermentation contributed the most to total agricultural carbon emissions, followed by fertilizer and animal manure management. Among the major livestock and poultry, beef cattle farming contributed the most to agricultural carbon emissions. And then were sheep, goats, cows, and pigs farming, and poultry farming contributed the least to agricultural carbon emissions. (3) Rural population, per capita GDP, per capita disposable income of rural residents, total power of agricultural machinery, proportion of agricultural added value in the province’s GDP, investment in fixed assets of rural households, application of agricultural scientific and technological achievements, and investment in agricultural science and technology were the main factors affecting agricultural carbon emissions in Gansu Province. The influence indices of these factors for agricultural carbon emissions were -0.017, 0.026, 0.020, 0.038, -0.025, 0.031, -0.017, and 0.016, respectively. To effectively control agricultural carbon emissions, appropriate strategies regarding the following should be adopted: improving the resource utilization efficiency of the plant production and soil carbon sequestration capacity industriously; strengthening the source reduction, process control, and end treatment of animal husbandry; reducing the dependence of farm machinery on oil to the maximum possible extent; promoting the use of clean energy in rural areas effectively; and increasing the research, development, and application of low-carbon agricultural technologies.