Analysis of the Accuracy of Radar Forward Operators in Simulation of a Supercell Storm

It is well recognized that the storm microphysical properties are very important for successfully prediction of severe thunderstorms, but are not well understood. Radar data observations and numerical modeling are two complementary tools to solve this problem. To cross-validate both observations and numerical models with different microphysics schemes, forward operators that convert model variables into radar observations are needed. Various forward operators have been developed during past years. For example, the simple Ferrier operator converts model variables rain, snow and hail mixing ratios into radar reflectivity (Z) using the Rayleigh approximation. A relatively more complex dual-pol simulator developed by Center for Analysis and Prediction of Storms (CAPS) links model hydrometer variables into radar dual-pol observations including Z, differential reflectivity (Z_dr), correlation coefficient (ρ_hv), and specific differential phase (K_dp), by using a rigorously calculated t-matrix and allowing for Mie scattering. The present study analyzes the accuracy of the different forward operators with an idealized supercell storm and a real tornadic supercell storm – May 20, 2013 Moore/Oklahoma tornadic supercell case. Forecast sensitivity experiments to different microphysics schemes will also be performed to determine the most appropriate scheme that could be used for storm prediction consistent with the forward radar simulator. Adjustments to some of the assumptions made by the Ferrier operator were made to test the effects of changing the modeled dielectric factors in the reflectivity equations. Another reflectivity category was also added to differentiate wet and dry hail. These adjustments led to an overall decrease in ice-phase reflectivity values that brought the simulated Z closer to what is calculated by the more rigorous dual-pol simulator.