Consider the wind drift effects in the radar-raingauge comparisons

Most radar-raingauge comparisons follow a major assumption that the precipitation observed aloft impacts the surface directly below the volume sampled by the radar. However, it is well known that rain may be advected laterally considerable distances by the horizontal wind. This is a serious problem when high resolution urban hydrology model are utilized as an input parameter. A new comparison method considering the horizontal displacement is proposed which is to find an optimal horizontal displacement of the radar data above the surface raingauges. After the horizontal adjustment, the radar estimated rain rates aloft will have maximum correlation with the rain rates of surface raingauges. The main assumption is that the precipitation observed by radar within the same sweep will fall down to the ground in the same time and shift with same horizontal displacement. Three convective events in Darwin, Australia were analyzed, the analysis shows that excellent results could be achieved using KDP-based rain-rate estimator by C-band polarimetric radar. The normalized error (normalized bias) of radar-raingauge comparisons for the traditional method is 43.9% (-14.1%) and for the optimal method is 27.0% (-10.7%). It's about 17% improvement of the normalized error when the optimal method is applied.