Advanced Polarimetric Doppler Weather Radar Simulator S. Lischi * , A. Lupidi † , M. Martorella ‡ Department of Information Engineering, University of Pisa Via G. Caruso, 16 - 56122 Pisa - Italy Email: * stefano.lischi@for.unipi.it { † alberto.lupidi, ‡ m.martorella}@iet.unpi.it Telephone: +39 050 2217673 F. Cuccoli * , L. Facheris † CNIT-RaSS Via di S. Marta, 3 50139 Firenze - Italy Email: * fabrizio.cuccoli@cnit.it † luca.facheris@unifi.it Telephone: +39 055 4796382 L. Baldini CNR, ISAC, ARTOV Via F. del Cavaliere, 100 00133 Roma - Italy Email: l.baldini@isac.cnr.it Telephone: +39 06 49934325 Abstract—The Advanced Polarimetric Doppler Weather Radar Simulator (APDWRS) capable of generating I&Q time series voltages for a ground based fully-polarimetric Doppler weather radar is described in this paper. The Weather Re- search and Forecast (WRF) model is used in conjunction with the T-Matrix code to generate radar signals according to the propagation-modified covariance matrix. Mixtures of rain, hail, graupel and snow hydrometeors relative to a realistic weather phenomenon are considered. The polarimetric and Doppler observables estimated from the simulated radar signals showed a good agreement with those found in previous observations. Keywords—Weather radar, radar polarimetry, radar simulation, T-Matrix, WRF. I. I NTRODUCTION Tools able to simulate realistic weather radar measurements from a known meteorological scenario is an important tool to develop signal and data processing algorithms. A method based on the propagation modified ensemble-averaged covariance matrix, obtained from the Weather Research and Forecast model with the T-matrix numerical electromagnetic model has been developed to simulate X-band polarimetric Doppler weather radar received voltage [1], [2], [3]. In literature we can find other simulators which adopt a similar approach. Li et al. [4] designed a microphysical based avionic weather radar observables simulator and Augros et al [5] designed instead a ground based weather radar simulator. These works are aimed to simulate a set of moment measurement remapped onto a radar grip using the output from a weather model. Our approach intends to develop a full weather radar simulator that starting from a reference ”true” scenario, obtained by a weather model, is able to generate (in reasonable computing time) real- istic I&Q time series at the receiver output from which typical weather radar moment measurements are obtained. In such a way, each processing step of the weather radar processing chain (e.g. signal processing algorithms, estimation of radar measurements and meteorological products as well) can be tested in a simulated environment. The method proposed in [1] is extended and refined in this work and applied to the case of a ground based fully-polarimetric C-band Doppler weather radar. The polarimetric and Doppler observables are then estimated from the simulated radar voltages and presented as a Plan Position Indicator (PPI). True fields of the corresponding radar variables are also produced on the same radar grid to easily assess the performance of signal processing algorithm and effects related to radar sampling as well. This paper is organized as follows. The system geometry and the radar voltages simulation method are introduced in Section II. The microphysics and electromagnetic models for the different type of hydrometeors are summarized in section III. Section IV describes the case of study considered in this work. The estimated polarimetric and Doppler observables and true fields are presented and discussed in Section V. The conclusion in Section VI ends this paper. II. WEATHER RADAR SIMULATION METHOD In this work we follow the simulation method based on the propagation-modified ensemble-averaged polarimetric covari- ance matrix presented in [1]. This method is herewith refined and extended to simulate a complete radar scan performed by a ground based fully-polarimetric Doppler weather radar. A. Simulation geometry The transmitted waveform is characterized by the pulse width T 0 and the repetition period T s . The extent of the Radar Resolution Volume (RRV), which is proportional to the distance from the radar, is determined by the range resolution Δr = cT 0 /2 and the antenna aperture angles [6]. Since a weather phenomenon can not be assumed homogeneous within a RRV especially at far distance from the radar, we generate the radar signals as the superposition of radar echos from a collection of P sub-RRVs contained in the current RRV. The reference systems and the geometry are shown in Fig.1. The weather phenomenon is referred to the reference system (x, y, z ) with an arbitrary origin (x 0 ,y 0 ,z 0 ) and parallel to the Earth’s surface. The radar reference system (x ′ ,y ′ ,z ′ ) is centered on the radar position (x a ,y a ,z a ) and is characterized by the elevation angle of the antenna θ el with respect to Earth’s surface. The p th sub-RRV is identified by the polar vector  r p ≡ (r p ,θ p ,φ p ) with respect to the antenna reference system. If each sub-RRV is sufficiently small compared with the dimensions of the weather phenomenon, the spatial distribution of the particles can be assumed to be homogeneous. Under this assumption, the radar echo from each sub-RRV, and on each polarimetric channel, can be modeled as a stationary Gaussian distributed random process with a Gaussian shaped auto-correlation function [6].