[A-AS03] Extreme Events and Mesoscale Weather: Observations and Modeling

[AAS03-06] Development of a Dual-Polarization Weather Radar Simulator Based on a Physical Rainfall Model Considering Atmospheric Turbulence

*Fumitaka Hino¹, Daichi Kitahara¹, Yuuki Wada¹, Tomoaki Mega¹, Tomoo Ushio¹, Toshio Iguchi¹, Eiichi Yoshikawa², Hiroshi Kikuchi³ (1.Osaka University, 2.JAXA, 3.The University of Electro-Communications)

Keywords:Dual-Polarization Weather Radar, Atmospheric Turbulence, Probability Density Function, Power Spectral Density, Polarimetric Parameters

The number of disasters caused by locally heavy rain is increasing. For more accurate rainfall prediction in a shorter time, dual-polarization weather radars, which receive both horizontally and vertically polarized signals, have been developed and introduced. To reduce the time and cost of weather radar development, there is a need for a simulator that can generate received signals similar to actual ones. Various simulators have been proposed, and they are classified into two types. One type outputs polarimetric parameters, such as radar reflectivity factor and specific differential phase, from rainfall conditions, such as drop size distribution and wind speed. The other type outputs a time series of received signals, whose statistical properties match given polarimetric parameters, as a realization of a complex-valued random vector. However, since each simulator has been proposed independently, the probability density function of the received signal is not directly given from the rainfall conditions through a consistent physical model. In this study, to realize a more precise and user-friendly weather radar simulator, we incorporate atmospheric turbulence into a conventional physical rainfall model that considers flattening, scattering, and terminal velocity. The proposed physical model enables us to define the power spectral density of the received signal from the rainfall conditions. Then, the probability density function of the received signal can be derived as a circularly-symmetric complex Gaussian distribution whose covariance matrix is given from the inverse Fourier transform of the power spectral density. Based on user-selected models for the flattening, scattering, terminal velocity, and atmospheric turbulence, our simulator outputs a time series of the received signals with the polarimetric parameters from given radar specifications, drop size distribution, and wind speed. Indeed, we confirmed that dual-polarized signals consistent with appropriate polarimetric parameters were generated when using the Pruppacher–Beard flattening model, the Rayleigh–Gans scattering model, the Atlas–Srivastava–Sekhon terminal velocity model, and a simplified Yanovsky atmospheric turbulence model. In the future, we plan to introduce more realistic and complicated flattening, scattering, terminal velocity, and atmospheric turbulence models, and to support the generation of pulse-compressed signals.