Based on the microwave radiometer and radiosonde data from January 2018 to December 2019 in the central and southern Hebei Province, this paper makes point-to-point matching of the data obtained by the two kinds of detection equipment in time and space. A total of 187 clear sky profiles, 1176 cloud sky profiles and 12 drizzle sky profiles are screened. The correlation and error of the temperature, relative humidity, and water vapour density profiles are quantitatively analysed from the two kinds of atmospheric detection equipment under the clear sky, cloud sky, and drizzle weather conditions in each altitude layer. The results show that temperature and water vapour density correlation between microwave radiometer and radiosonde is desirable. The distributions of atmospheric parameters observed by ground-based microwave radiometer are different. Still, the trends of microwave radiometers and radiosonde are consistent. The correlation between the atmospheric parameters observed by radiosonde and the temperature, relative humidity and water vapour density retrieved by microwave radiometer shows that the results in the lower atmosphere is better than those in the upper atmosphere; the temperature correlation is the best, the water vapour density is the second and the relative humidity is the lowest. By comparing the fitting degree of each altitude layer, it is found that the accuracy of atmospheric parameters below 3.5 km is more reliable. For the atmospheric physical quantities that need to be calculated indirectly, such as K-index, effective potential energy, integral water content, etc., the accuracy of atmospheric physical quantities calculated by using low-level data is better. The comparison of atmospheric physical parameters retrieved by ground-based microwave radiometers and radiosonde observation in this paper has reference significance for improving the detection accuracy of microwave radiometers in Atmospheric Physics and weather modification researches.