Objective The aim was to promote the application of climate reanalysis data in the study of ground temperature variation in Inner Mongolia, solving the problem of severe lack of historical ground temperature observation data, and providing a scientific basis for the study of climate change in Inner Mongolia.

Method Taylor diagrams and characteristic stations were used to analyze the applicability of ERA5 reanalysis ground temperature data at five different depths of soil layers from 1961 to 2022 in the Inner Mongolia region, and Sen trend analysis and M-K test were used to study the ground temperature change trend.

Result The consistency between ERA5 ground temperature and meteorological station observed ground temperature decreased with the increasing of soil depth, but all correlation coefficients passed the significance test at the 0.01 level. The correlation between ERA5 land temperature at each layer and the observed ground temperature of stations with deeper soil depth in the vicinity were higher. Compared with the station observations, ERA5 ground temperature generally showed a phenomenon of lower temperatures in the warm season and higher temperatures in the cold season. The annual average ground temperature of each layer in Inner Mongolia ranges from 4.6 to 6.5℃, with the surface temperature being the lowest and the shallow layer temperature near the surface being the highest. Spatially, ground temperatures at each layer in Inner Mongolia showed an increasing distribution from northeast to southwest. During the study period, the ground temperatures at each layer in Inner Mongolia showed a warming trend, with an average change rate of 0.038 to 0.026℃ per year, with the surface temperature warming the fastest and the shallow layer temperature near the surface warming the slowest. Differences in changing rates by month decreased with the increasing of soil depth at the same time. The months with the most obvious warming also showed a lag phenomenon as the depth increases.

Conclusion ERA5 ground temperature data had good applicability in Inner Mongolia region.