This thesis presents the computer model FORWP that is capable of predicting attenuation and cross polarization at 10-30 GHz from dual-polarization radar data with a high degree of accuracy. FORWP uses a rigorous backscattering computer model BSCAT to infer rain drop size distribution along the slant path from radar measured reflectivity and differential reflectivity data collected along the path. Then, two semi-empirical models are developed which predict attenuation from radar measured reflectivity and differential reflectivity. These two semi-empirical models are used to evaluate FORWP. Finally, attenuation predictions of FORWP are compared to the two semi-empirical prediction models and measured link attenuation at 11.4 GHz for a rain event in southwest Virginia. / M.S.