Utility of Optimal Reflectivity-Rain Rate (Z-R) Relationships for Improved Precipitation Estimates

Use of appropriate reflectivity (Z)-rainfall rate (R) relationships is crucial for accurate estimation of precipitation amounts using radar. The spatial and temporal variability of several storm patterns combined with availability of several variants of Z-R relationships makes this task very difficult. This study evaluates the use of optimization models for optimizing the traditional Z-R functional relationships with constants and coefficients for different storm types and seasons. Optimization model formulations using nonlinear programming methods are investigated and developed in this study. The Z-R relationships will be evaluated for optimized coefficients and exponents based on train and test data. The train data will be used to develop the optimal values of coefficients and constants and the test data will be used for assessment. In order to evaluate the optimal relationships developed as a part of the study, reflectivity data collected from NCDC and rain gage data are analyzed for a region in South Florida. Exhaustive evaluation of Z-R relationships in improving precipitation estimates with and without optimization formulations will be attempted in this study.