Abstract: X-band radar has been now more and more widely used in rainfall estimation, while its attenuation is more serious than S- and C-band radar. This constitutes a major source of error for X-band radar rainfall estimation, in particular for intense precipitation. Thus, the X-band radar data need to be corrected before used. On the basis of previous studies, this paper proposes a correction method which uses the characteristics of meteorology object and specific differential propagation phase shift KDP. The copolar differential propagation phase shift is composed of two components, namely, differential propagation phase and differential backscatter phase shifts. To estimate specific differential phase KDP, these two components must first be separated when significant differential backscatter phase is present. The paper introduces the Kalman filter method that is an Optimal Recursive Data Processing Algorithm, which can separate these two components and filter out other random noises. The method is highly effective and practical and can also be used to process other radar signals. The results show that the reflectivity attenuation correction using KDP processed with the Kalman filter method is very stable.