Abstract: A lot of researches about wind profile radar are based on analysis of echo power. So the estimation accuracy of echo power has direct effects on the use and expansion of wind profile radar data. At present, return signal power of wind profile radar is estimated by signal to noise ratio (SNR) and radar system noise power. Environmental noise may cause errors when estimating SNR in power spectrum of wind profile radar, affecting calibration accuracy. Another viable option is quantitatively testing the receiver and signal processor respectively using the same signal source, and then the calibration curve of radar systems can be determined according to the measured signal processor linear curve. Using this method, the calibration curve of CFL 03 wind profile radar is determined, and probe data of the radar in July and August, weather radar data and surface rain gauge data are compared. The results show that the weak signal of wind profile radar can be calibrated effectively, on the basis of measurement of the coherent accumulation and pulse compression accumulation using intermediate frequency source. The method proves feasible through the actual test. And through field experiments it can be seen that calibrated echo intensity of wind profile radar and SA weather radar echo intensity are basically the same with an average error less than 2 dB, within the scope of weather radar calibration error. The errors between calibrated echo intensity of wind profile radar and ground rain gauge data is less than 2 dB, indicating that the calibration results are credible.