This study employs Moran's I index and cold-hot spot analysis to characterize the spatiotemporal dynamics of carbon emissions in Yunnan Province from 2000 to 2021. Additionally, a random forest model is used to identify the key socioeconomic factors influencing carbon emission of 16 prefectures in Yunnan Province. The study finds that there are no significantly low carbon emission areas in Yunnan Province, with emission values generally close to the average and evenly distributed spatially. The hot spot regions remained stable across time, exhibiting a clear spatial clustering effect. Further analysis reveals that industrial added value, energy consumption, population size, and GDP are the main factors affecting carbon emissions. Our findings can offer useful guidance in formulating regional carbon neutrality roadmaps, implementing differentiated carbon reduction strategies, and promoting low-carbon green development.