Effects of rain gage density on correcting radar precipitation and its influence on hydrologic model simulation

The NEXRAD multisensor precipitation estimator (MPE) data provided by National Weather Service River Forecast Centers uses rain gauges to correct biases in the radar precipitation estimates. However, limited number of rain gauges is used in the MPE analysis, which may not completely eradicate biases in the radar data, and may lead to poor prediction of streamflows from hydrologic models. This study investigates the role of rain gauge density on bias correction and its effects on hydrologic model simulations. By using Kalman Filter, data from independent rain gauges that are not used in the MPE analysis are assimilated into the MPE data. The performance of an in-house event based distributed hydrologic model, GIS and Hydrologic Information System Modeling Object (GHISMO), developed for Cedar Creek in northeast Indiana is evaluated on the original MPE data and MPE data filtered using increased rain gauge density. The results suggest that increasing rain gauge density improves radar precipitation data leading to better hydrologic model predictions.