<p>Land surface temperature (LST) plays a dominant factor in surface energy exchange and carbon cycle of the Earth system. This paper uses LST data from 2009 to 2022 provided by MODIS the products to analyze the temporal and spatial variations in China. The spatial pattern of LST in China is consistent with climate regionalization and the Heihe-Tengchong line. We used the curve fitting method to extract temporary LST variations and the K-means method to extract the spatial pattern of LST variations. The North China Plain, Xinjiang region, and Inner Mongolia region have an obvious warming trend in spring and winter. Clustering results of long-term changes indicated that regional LST variations are caused by the joint effects of human activity intensity and natural factors such as climate, terrain, and vegetation. The Qinghai-Tibet Plateau and Northeastern China are the most sensitive areas for LST changes, where abnormal high temperatures can be detected in El Niño years. Through the collaborative analysis, we found that LST and CO₂ are significantly positively correlated in space. In terms of long-term time series, more significant correlations are observed in the North China Plain and Northeastern China. These results provide a basis for further research on the feedback mechanism between surface temperature and the carbon cycle in different regions of China.</p>