Based on the 2-m air temperature data of 1 519 meteorological observation stations in Shandong from March 2021 to February 2022, the applicability of the 2-m temperature daily statistics (mean, maximum, and minimum temperature) from HRCLDAS (High Resolution China Meteorological Administration Land Data Assimilation System) and ERA5-Land (ECMWF Reanalysis v5-Land) in Shandong is compared and evaluated. The results are as follows. (1) The RMSE (root mean square error) of mean, maximum, and minimum temperature from HRCLDAS/ERA5-Land daily statistics is 0.1/1.2 ℃, 0.6/1.9 ℃, and 0.4/1.7 ℃, respectively, which means that HRCLDAS is more accurate, and HRCLDAS in different geographical areas and at different altitudes performs better than ERA5-Land. The deviation of HRCLDAS in most areas (-0.5 to 0.5 ℃) is much lower than that of ERA5-Land (-2.0 to 2.0 ℃). (2) The comparison and evaluation of the two datasets on the monitoring ability of high temperature and cold wave show that HRCLDAS can capture most of the high temperature and cold wave processes, and the spatial distribution of the number of high temperature and cold wave days is similar to the observation, but it somewhat underestimates the influence range. However, ERA5-Land can only monitor part of the high temperature and cold wave processes, and the number of high temperature and cold wave days is seriously underestimated.