Effects of using different Stage III techniques on radar-derived rainfall accumulations

Previous studies have shown that the WSR-88D and Stage III underestimate rainfall. This is especially true during winter when stratiform rainfall is most common. Stage III hourly rainfall composites had been created at the Lower Mississippi River Forecast Center using an averaging technique when radar coverages overlap. This technique was believed to account for a large portion of the total underestimation and as a result, the LMRFC switched to using the maximum rainfall in overlapping coverages last winter. This study documents changes in rainfall accumulations that result from switching to using a maximum rainfall amount during the Stage III process. Specifically, the winter of 2000 is compared to the winters of 1998 and 1999 for several watersheds across the southern U.S. using Stage III and mean areal precipitation derived from rain gages.