Atmospheric temperature is an important parameter for studying the process of atmospheric dynamics and photochemistry, and is a significant sign of the atmospheric vertical stratification structure. It is a challenge in atmospheric science research to obtain temperature profiles with high space-time resolution all the time. MST (mesosphere-stratosphere-troposphere) radar, which is the modern large-scale ground-based radio remote sensing equipment, can measure 3D atmospheric winds with high space-time resolution and unattended 24 hours a day. This paper uses the time series of vertical wind observed by MST radar to make spectral analysis and calculate B-V (Brunt-Väisälä) frequency. Then this paper builds the discrete model of temperature inversion based on the relationship between B-V frequency and temperature. Compared with actual measurement of temperature from radiosondes, the agreement between the radiosonde profile and the profile from MST radar data is quite good. Furthermore, the Richardson number can also be obtained by using the B-V frequency and horizontal winds, which can judge the atmospheric stability, quantize many of the dynamic characteristics, and explain the wave phenomenon. So the acquisition of B-V frequency is MST radar's another outstanding contribution to atmospheric dynamics research. It can accurately calculate the atmospheric temperature profiles and attain dynamic stability parameters.