The Application of Hydrometeor Classification Algorithm for China New Generation Doppler Radar

Abstract

This paper mainly improves the hydrometeor classification algorithm for China new generation doppler radar. Firstly, the algorithm uses the following parameters as input: horizontal reflectivity, differential reflectivity, correlation coefficient, specific differential phase, standard deviation of horizontal reflectivity [SD(Z h )] and standard deviation of differential phase [SD($\Phi_{DP}$)]. Secondly, the hydrometeor classification model is built through the processing of fuzzification, rule judgment, integration and deblurring. Due to errors in radar measurements and antenna beam broadening at longer distances, the result of the hydrometeor classification is inaccurate. To account for radar measurement errors and beam broadening, it makes some modifications based on the original algorithm. These modifications include an estimate of the confidence factors and the parameters of the melting layer. The confidence factors characterize the possible impacts of the radar measurement errors. The parameters of the melting layer are determined as functions of azimuth with polarimetric radar measurements. These additions provide greater flexibility in algorithm optimization while also improving the discrimination between liquid and frozen hydrometeors. Finally, the algorithm is verified using dual-polarization radar data collected by a S-band dual-polarization weather radar in Beijing. The results show that the new hydrometeor classification scheme identifies 10 different classes of radar echoes.