Use of a Generalized Radar Simulator for the Validation of Precipitation Estimation from GPM Dual-frequency Precipitation Radar Measurements

Precipitation estimation from space-borne radar measurements provided by the Global Precipitation Measuring (GPM) mission offers great opportunities not only to monitor and understand microphysical and thermodynamical features in more detail especially over the ocean, but also to improve climate models and numerical weather prediction (NWP) models. In this context, the validation of retrieval algorithms of precipitation estimation already in operation is important, and one of approaches is to use a numerical radar simulator based on theoretical radiative transfer formulations.

In this study, a physically-based simulator has been designed using the Mishchenko's T-matrix code and emulators already proposed (e.g. Jung et al. 2014). The simulator uses the output of the Weather Research and Forecasting (WRF) model as the input atmospheric data. The simulator can deal with Rayleigh/non-Rayleigh scattering from non-spherical precipitation particles. Ice particle shape is modeled as oblate and prolate spheroid. The effects of attenuation and multiple scattering are optionally accounted for shorter wavelengths. To account for effects of varying the density and dielectric constants due to melting of solid hydrometeors, that does not explicitly modeled in bulk microphysical schemes in the WRF model, a simple melting layer model is also included as previous studies. This radar simulator can be applied to any weather radar wavelengths, so that it is useful for a forward operator to link polarimetric radar observations and NWP models.

We will apply the radar simulator for calculating radar observables from a Ku-/Ka-band (13.6/35.55 GHz) dual frequency precipitation radar (DPR) onboard the core GPM plathome as well as for deriving polarimetric parameters obtained from (C- and X-band) ground-based radars in a framework of observing system simulation experiments. A squall-line system that includes both deep convection and stratiform regions is chosen as an event to be studied. The Milbrandt-Yau double-moment scheme is employed in the WRF model simulation. Sensitivities are investigated in terms of drop size distribution, the use of a single-moment scheme, and the effect of multiple scattering.

The radar simulator is evaluated by comparing simulated observables from DPR with ones from ground-based radars and/or model microphysical variables. The structure of bright-band of enhanced radar reflectivity is reasonably retrieved in stratiform regions, and the results confirm basic characteristics that the assumption of non-Rayleigh effect is crucial for space-borne radar frequencies, and that the effects of attenuation are also significant. Observations from the scattering of Ka-band electromagnetic wave can be easily contaminated by multiple-scattering effects that would be considered in precipitation estimation. Although considerable validations are needed, the results suggest that the simulator has a good reliability. The validation of DPR precipitation estimation will be also discussed using simulated ground-based polarimetric radar data from the viewpoint of hydrometeor type classification.