Abstract: The greenhouse gas emissions from agricultural sources in Xinjiang from 2009 to 2019 were estimated with reference to the methods recommended in the Provincial Greenhouse Gas Guidelines. The temporal characteristics and spatial differences of greenhouse gas emissions from agricultural sources in Xinjiang were explored. The decoupling characteristics and influencing factors of greenhouse gas emissions from agricultural sources in Xinjiang were analyzed using Tapio model and LMDI decomposition method. Finally, the emissions of greenhouse gas from agricultural sources in Xinjiang in 2030 and 2050 were predicted by constructing different scenarios. The results indicated that the emission of greenhouse gases from agricultural sources in Xinjiang showed an overall upward trend from 2009 to 2014, which was divided into two stages： fluctuating rise and steady rise, taking 2013 to 2014 as the boundary; The regional emissions of greenhouse gases from agricultural sources in Xinjiang were significantly different, and the Ili Kazakh Autonomous Prefecture and Altay region belonged to the “double high” region; the decoupling state of agricultural greenhouse gases in Xinjiang was not ideal, mainly weak decoupling; the most important factor to curb agricultural greenhouse gas emissions in Xinjiang was agricultural production efficiency, followed by agricultural structure. The most important factor to promote agricultural greenhouse gas emissions was agricultural economic level, followed by agricultural population size; the prediction results showed that the agricultural greenhouse gas emissions in Xinjiang would not reach the peak in 2030 and 2050, and would continue to rise. Facing the severe emission reduction situation in Xinjiang, agricultural production efficiency should be improved the industrial structure should be adjusted and optimized emission reduction indicators should be quantified, and emission reduction tasks should be implemented.

Key words: greenhouse gases from agricultural sources, space time characteristics, decoupling characteristics, scenario prediction, Xinjiang