| dc.description.abstract | Accurate and reliable Quantitative Precipitation Estimation (QPE) plays important role in hydrological and meteorological applications. QPEs are vastly derived from radar measurements; however, the inherent noise in radar polarization parameters and the variation of drop size distributions (DSD) in different precipitation events limit the accuracy. Therefore, this study utilized the variational technique to estimate the rainfall rate and compared with the radar-based QPE, to evaluate the performance of variational QPE in different precipitation events.
Three radar observed parameters (differential reflectivity (Z_dr), differential phase shift (∅_dp), and specific differential phase (K_dp)) of collocated dual-polarized S- (RCWF) and C-band (RCMD) radars were used. Mei-yu and Typhoon cases in 2017 were selected, and eight different experiments were conducted according to (1) either used three radar observed parameters (Z_dr, ∅_dp, and K_dp) or two radar parameters (Z_dr and ∅_dp); (2) fixed or changed the observation errors; and (3) two different resolutions (raw (R1) and modified-low (R2)).
The observed radar parameters and variational results showed in good agreements. The variational results have the noise reduced and become smoother, variational derived ∅_dp and K_dp are all positive values. Generally, the two cases of variational QPE of RCWF showed that using three parameters are better than two parameters, and resolution R2 are better than resolution R1. However, for the comparison of the fixed and changed observation errors, fixed observation errors in Mei-yu case perform better than Typhoon case. Besides, the variational QPE of RCWF perform better than RCMD in the Mei-yu case, but for the RCMD, the variational QPE are better than the radar-based QPE. | en_US |