Analysis of polarimetric weather radar data quality in the region of large normalized spectrum width

The data quality of polarimetric weather radar is affected by the spectrum width (SW), especially for the alternate transmission and alternate reception (ATAR) mode weather radars, where the interval between co-polarization signals is twice the pulse repetition time (PRT). Therefore, with the increase of the SW, the correlation of the signals decreases rapidly, leading to the degradation of the data quality. Moreover, for weather radars using the dual pulse repetition frequency (PRF) technique, different PRFs of adjacent radials have different impacts on the performance of radar data, resulting in differences in the estimates for adjacent radials. In this paper, we analyze the bias and standard deviation (SD) of radar data for different normalized SWs and PRF ratios using generated weather-like time-series I/Q data by signal simulation. The results show that when the normalized SW (NSW) is larger than a specific value, the values of bias and SD of radar data (except SNR) increase. For the weather radar using DPRF, the estimated SW from a ratio of lag-0 to lag-2 autocorrelations (SW02 estimator) and the correlation coefficient (CC) are most likely to show stripes. When the NSW is less than 0.3, the striated data can be avoided. Finally, the measured Ku-band weather radar data is used for verification, when the SW is greater than 2.5 m/s (i.e., NSW>0.32), the estimated SWs using the SW02 estimator and CCs have obvious stripes, which is consistent with the theoretical analysis results.