Scale Based Evaluation of IMERG Rainfall Products in Hydrologic Modeling

In this study we investigate the performance of 2 recent satellite rainfall products, the evaluation will be in a hydrologic modeling basis. The two products are IMERG early and late run products. The products are provided in different latencies, 6 hours for the early product and 18 hours for the late run product. The late run product is enhanced by data from extra Microwave satellites and rain gauges. For this study we are using our distributed hydrologic model developed in the Iowa Flood Center (IFC). The enhancement of the late run product boosted the performance of the product (in terms of resulting discharges from the IFC hydrologic model) significantly during the wet season of the year 2015 in our early results. Also, the results show that the late run product shows promising performance in larger scales. We are going to expand the study to follow the dynamical evaluation method described in Quintero et al. 2016 to better evaluate the performance of the products. This will also help us identify the proper scale which is able to minimize the errors in the rainfall estimates through the accumulation over the stream network (rainfall only no discharge included). Our study area is an average sized ( 17,000 sqm) basin in eastern Iowa called the Cedar River Basin. It is monitored by approximately 12 USGS stream gauges. Reference Quintero, F., W. Krajewski, R. Mantilla, S. Small, and B. Seo, 2016: A SpatialDynamical Framework for Evaluation of Satellite Rainfall Products for Flood Prediction. J. Hydrometeor. doi:10.1175/JHM-D-15-0195.1, in press