A Comprehensive Comparison of Seven Surface Air Temperature Reanalysis Products over Mainland China from Multiple Spatiotemporal and Geographical Perspectives

Reanalysis surface air temperatures (SATs) offer significant advantages for monitoring climate change, but their performance has not been systematically evaluated across diverse perspectives, limiting studies on surface–atmosphere interactions. In this study, observations from approximately 2400 stations across China (2013–2018) were used as a reference to assess seven types of reanalysis SATs (CMA-RA/Land, ERA-Interim, ERA5, ERA5-Land, GLDAS, JRA55, MERRA2) from different perspectives, and the main factors affecting the evaluation results were also discussed. The results indicate that (1) the overall RMSEs of the seven SATs range from 2.06 to 2.85 K, with ERA5 presenting the highest accuracy (RMSE = 2.06 K, R2 = 0.97) and JRA55 performing the worst (RMSE = 2.85 K, R2 = 0.95); (2) overall, their performances vary across multiple time scales, spatial scales, and weather conditions, exhibiting poor performance in western China, in winter, in high-altitude areas, over forest land, and under cloudy skies; (3) by analyzing the influencing factors, we found that terrain correction and station quality are key sources of assessment errors, while the impact of interpolation methods can be disregarded. These results provide valuable references for the development and improvement of reanalysis data and can assist researchers to select appropriate SAT products in various geographical environment and weather conditions in China.