Performance and Evaluation of Multisensor Precipitation Estimation Algorithm Using a High Density Rain Gauge Network and Hydrologic Simulation

Abstract

Abstract: A rain gauge network (28 rain gauges) was installed in western Puerto Rico (PR) within a 4km x 4km GOES satellite pixel. Located within the pixel is a well monitored sub-watershed of 3.55 km2, referred to here as the testbed subwatershed (TBSW). The rain gauge network was established to evaluate the performance of the GOES-based Hydro-Estimator (HE) rain rate algorithm, and estimated rain rates from NEXRAD radar and the Center for Collaborative Adaptive Sensing of the Atmosphere (CASA) radar network, which has a high spatial resolution (approximated 200 m). Furthermore, the rain gauge network will provide a high temporal and spatial resolution rainfall dataset to be input into a distributed hydrologic model in the TBSW. The focus of this work is to evaluate the performance of the Multisensor Precipitation Estimation (MPE) product at 1hour and 1day temporal resolution within the 4km x 4km HE pixel and at watershed level for 2007. The MPE product is popular within the hydrologic modeling community due to its resolution and mean field bias correction computations in its coverage. Results for 2007 indicate that the highest rainfall measured by the rain gauges within the HE pixel area were September with an average and standard deviation of 241.75 mm and 73.3 mm, respectively; and August with 223.7 mm and 64.66 mm, respectively. While for the same months the Multisensor Precipitation Estimation, produced a total monthly rainfall accumulation and standard deviation of 247.36 mm and 64.4 mm for September, respectively, and 233.68 mm and 36.54 mm for August, respectively. The mean and standard deviation daily field bias for these months were 1.08 and 1.5 for September, respectively, and 0.93 and 1.6 for August, respectively. The bias changed, when considering an hourly analysis, to 1.98 average and 5.45 standard deviation for August and 1.49 average and 3.01 standard deviation for September.