This study investigates evolving methodologies for multisensor precipitation estimation to determine their influence on the flow predictions of a distributed hydrologic model. These methods include the National Mosaic and Quantitative Precipitation algorithm package (NMQ) under development at the National Severe Storms Laboratory, and the Multisensor Precipitation Estimator (MPE) and High-Resolution Precipitation Estimator (HPE) suites currently operational at National Weather Service field offices. The goal of the study is to determine which combination of algorithm features offer the greatest benefit toward operational hydrologic forecasting. These features include automated radar quality control, range correction, automated Z-R selection and bright-band identification, bias correction, multiple-radar data compositing, and gauge-radar merging, all of which vary among NMQ, MPE, and HPE.

The algorithms described above have been applied to driving high-resolution (4-km hourly) gridded precipitation estimates over and near the Tar River Basin of North Carolina. In these comparisons, the NMQ and MPE/HPE algorithms are driven by identical WSR-88D reflectivity and rain gauge input. Our previous study for a series of precipitation events during the period December 2004-January 2005 indicates that algorithm features such as automated Z-R selection and bright-band identification lead to improved accuracy in rainfall estimates and streamflow simulations. Later phases of the study include warm season convective events and a tropical cyclone (Hurricane Isabel in September 2003). All results are currently being compared to an independent set of hourly and daily rain gauge reports; the various precipitation grids will also be input to the NWS Hydrology Laboratory - Research Distributed Hydrologic Model, to determine impacts on the quality of its discharge simulations at several gauged points on the Tar River and its tributaries. These and additional results will be reported in our extended abstract.