Abstract: One of the important application of polarimetric Doppler weather radar is quantitative precipitation estimation (QPE). Polarimetric radar detects reflectivity (Z_H), differential reflectivity (Z_DR) and specific differential phase (K_dp), all of which are closely related to precipitation particles. Four commonly used QPE methods include Z_H-based R(Z_H), Z_H-Z_DR-based R(Z_H, Z_DR), K_dp-based R(K_dp) and K_dp-Z_DR-based R(K_dp, Z_DR) methods. Based on the observations of Shenzhen S-band and X-band polarimetric radars as well as the high-resolution elevation data and the disdrometer data, the QPE method based on polarimetric variables is presented. First, the blockage information of the two radars is analyzed, and the hybrid tilts of the two radars are generated based on the elevation data and locations of the radars. The parameters for the above 4 QPE methods suitable for Shenzhen are then calculated with disdrometer observations using the T-matrix method. Finally, the blend method of QPE is proposed to use different QPE methods based on the intensity of polarimetric signals (i.e. K_dp and Z_DR). The performance suggests the accuracy and stability of the blend method is better than any of the single QPE method. The QPE products of Shenzhen S-band and X-band radars are generated with the QPE method proposed in this paper, and their performances are evaluated against operational QPE products. Results suggest that the QPE products generated by the method proposed in this paper are better than operational products both in accuracy and stability. Besides, the performance of this method is slightly better for X-band radar than for S-band radar, suggesting that the utilization of high-resolution X-band radar can improve the accuracy of QPE. However, due to the uncertain relationship between polarimetric signals and rain rate in the melting layer and ice region in the radar scan domain, the QPE method proposed here can only be applied to liquid rainfall region of the radar scan domain.