Abstract: In order to comprehensively evaluate the temperature and humidity detection performance of the ground microwave radiometer (referred to as radiometer, the same below) and improve the accuracy of temperature and relative humidity profiles, by using the ground microwave radiometer data, radiosonde data, and ERA5 reanalysis data of Beijing from March to August 2021, this paper analyzes the temperature and humidity detection performance of the radiometer under sunny, cloudy and rainy days. In the meanwhile, this paper uses radiosonde data and corrected ERA5 data to verify temperature and humidity profiles of radiometer and correct temperature and humidity data of radiometer based on the linear deviation correction method. The results show that: (1) The temperature of microwave radiometer has good consistency with radiosonde temperature under sunny and cloudy days, but it is higher than the radiosonde temperature in rainy days. The relative humidity of microwave radiometer is higher than that of radiosonde in sunny days and lower in cloudy and rainy days. The linear deviation correction effect of radiosonde on ground-based microwave radiometers in rainy days is better than that in sunny and cloudy days. (2) Above 4 km, ERA5 relative humidity data has obvious deviation with radiosonde data, while the temperature correlation between ERA5 and radiosonde is higher than relative humidity correlation. (3) The ERA5 data after radiosonde correction can improve the temperature and humidity profile accuracy of the Microwave radiometer, and the effect is better in rainy days than in sunny days. In vertical direction, with the increase of height, the correction effect of radiometer temperature and humidity data is more obvious. (4) The two true values of radiosonde data and corrected ERA5 data were used to analyze the temperature and humidity data of the ground-based microwave radiometer corrected by ERA5, and the results showed that the vertical distribution characteristics of the error (mean error BIAS and root mean square error RMSE) profiles of radiometer and radiosonde, radiometer and corrected ERA5 data were similar under sunny and cloud conditions. Under rainy conditions, there were obvious differences between the two in the middle and upper height (7-10 km), among which the temperature RMSE differed by 4 K and the relative humidity RMSE differed by 20%.