SimRadar: A Radar Simulator to Investigate Dual-Pol Characteristics of Tornadic Debris

Tornadoes are one of the most hazardous natural disasters on the earth surface, and radar is the primary instrument to monitor, study, and predict the occurrence of tornadoes. In an attempt to understand some of the unique electrical characteristics of tornadic debris observed by dual-pol weather radars, a numerical radar simulation was developed. The electrical characteristics of interests are regions of low cross-pol correlation coefficient and negative differential reflectivity near the vortex of tornadoes. Such signatures have been observed numerously by weather radars, but the underlying root cause of the phenomenon are yet to be understood. While there are suggestions of common debris orientation due to the centrifuging effects or collections of large objects that dominate the scattering signals from a radar due to their high radar cross section, validations of the hypotheses are, in practice, prohibitive due to the extreme danger near tornadoes. Needless to mention that obtaining the 3D orientation of a particle is in itself a very challenging task. We elected to validate the hypotheses through a numerical scene construction. It is accomplished by using a physical weather background with realistic tornadic vortex and populating the simulation domain with debris objects to produce radar time-series data through a Monte Carlo method, then generating the radar products to verify if the same electrical characteristics can be observed. The novel aspects of this simulator include the use of realistic debris trajectory calculation, which is supported by wind-tunnel measurements, and the use of representative radar cross section of debris, which is supported by anechoic chamber measurements. Through this simulator, we have successfully constructed simple tornado configurations and produced radar measurements with the same electrical characteristics as evidenced by simulated dual-polarization signatures of negative ZDR and low correlation coefficient.  The detailed description of the simulator and some initial results will be presented.