High resolution estimation of specific differential phase and backscatter differential phase for polarimetric X-band weather radars

Wide interest and considerable effort have been given to estimate the specific differential phase (K_dp) and more recently the backscatter differential phase (d_hv) because of their independence to radar miscalibration and attenuation. However, their typical estimations require a substantial amount of smoothing processes. This might lead to undesirable spatial resolutions, significant biased estimation, and therefore unreliable weather radar observations. In this work, an advanced method to estimate K_dp and d_hv from pure rain at X-band frequencies is proposed. The method aims to obtain high spatial resolution of K_dp and d_hv estimators while controlling their inherent bias-variance dilemma. The method consists of 4 processes: 1) translate theoretical relationships between polarimetric variables, 2) isolate d_hv from the total differential phase, 3) scale range derivatives of the propagation differential phase component, and 4) distribute them among range gates. In addition, the variance of K_dp and d_hv were mathematically formulated for quality control scheme. This method was applied to two storms events, stratiform and convective, using the horizontally scanning polarimetric X-band weather radar in the Netherlands. Results have shown that estimated K_dp and d_hv were able to retain the spatial variability of storms, few tens of meters, and produce a variance similar to or less than those of conventional approaches. It is foreseen that the proposed method would improve the quality of radar-based rainfall and raindrop size distribution retrieval as recommended by both, urban-hydrology and weather-forecast users.