Evaluation of forward operators for polarimetric radars aiming for data assimilation

Abstract

In the preparation for polarimetric radar data assimilation, it is essential to examine the accuracy of forward operators based on different formulations. For this purpose, four forward operators that focus on warm rain conditions are compared both with each other and actual observations with respect to their performance for C-band dual polarimetric radars. These operators mutually consider radar beam broadening and climatological beam bending. The first operator derives polarimetric parameters assuming an exponential raindrop size distribution obtained by the models and is based on fitting functions against scattering amplitudes. The other three converters estimate the mixing ratio of rainwater from the measured polarimetric parameters. The second converter uses both the horizontal reflectivity (ZH) and the differential reflectivity (ZDR), the third uses the specific differential phase (KDP), and the fourth uses both KDP and ZDR, respectively. Comparisons with modeled measurements show that the accuracy of the third converter is superior to the other two. Another evaluation with actual observations shows that the first converter has slightly higher fractions skill scores than the other three. Considering the attenuation effect, the fitting function and the operator only with KDP are found to be the most suitable for data assimilation at C-band.