Evaluation of Satellite Rainfall Products for Flood Prediction Using a Spatio-Dynamical Analysis Framework

Abstract:

Rainfall maps derived from satellite observations provide hydrologists with an unprecedented opportunity to forecast floods globally. However, the limitations of these precipitation estimates to produce reliable flood forecasts at multiple scales are not well-understood. Here is presented a data analysis framework to support the scientific and practical question of applicability of space-based rainfall products for global flood forecasting given the limits of space-time resolution and considerable uncertainty of these products. The authors used the hydro-meteorological data collected during the IFloodS campaign in Iowa between May and June 2013 to perform hydrologic simulation using two different rainfall-runoff model structures and five different precipitation products, two radar-based (Stage IV and IFC) and three satellite-based (PERSIANN-CCS, TMPA-RV and CMORPH 1.0). The framework allows studying hydrologic discrepancies at several spatio-temporal scales: spatially by considering all the sub-basins on a given catchment, and temporally by changing dynamically the temporal window of analysis. We use the framework to study discrepancies in rainfall accumulation and peak flow estimates resulting from using different rainfall products as external forcing to a hydrological model, with the well-established Stage IV radar product taken as a reference for comparison.