How reliable are satellite precipitation estimates for driving hydrological models: A verification study over the Mediterranean area

Abstract Satellite-based rainfall products (SRPs) are nowadays available at ever increasing accuracy and higher spatial and temporal resolution with respect to the past. Despite this, they are scarcely used in hydrological modeling. The main reasons may be related to: 1) the large bias characterizing satellite precipitation estimates, which is dependent on rainfall intensity and the season, 2) the relatively large spatial/temporal resolution with respect to the applications, 3) the timeliness, which is often insufficient for operational purposes, and 4) a general (often not justified) skepticism of the hydrological community in the use of satellite products for land applications. The objective of this study is to explore the feasibility of using SRPs to force a lumped hydrologic model – MISDc – over 15 basins in the Mediterranean area with different sizes and physiographic characteristics. Specifically, TMPA 3B42-RT, CMORPH, PERSIANN and a new soil moisture-derived rainfall datasets, SM2RAIN CCI , obtained through the application of SM2RAIN algorithm ( Brocca et al., 2014 ) to ESA CCI soil moisture product are used in the analysis. The performances obtained with the selected SRPs are compared with those obtained from a ground-based rainfall product (E-OBS). In addition, the performance obtained by an integration of the SRPs is investigated to see whether merged rainfall observations are able to improve flood simulations. The results highlight that SRPs provide relatively low performances when used to force MISDc model with respect to the case where ground observations are used. Worst results are obtained in smaller basins ( 2 ) that, however, represent the main target for flood modelling in the Mediterranean area. Relatively better performances are obtained when independent SRPs are integrated each other, and particularly for the case where TMPA 3B42-RT is integrated with SM2RAIN CCI . In accordance with previous studies, this suggests that the exploitation (via integration) of independent sources of rainfall offers a potential way to reduce rainfall-induced streamflow errors, even in challenging areas like the Mediterranean.