Evaluating the Efficiency of Reanalysis and Remote-Sensing based Rainfall Data sets for Hydrological Modeling Using VIC-3L Large Scale Model

Currently, in most of the catchments, the lack of ground-based gauges is one of the most important problems for accurate hydrological modeling. According to the quick developments of satellite-based technologies and the computer’s computational power, several rainfall datasets have been developed with different spatial and temporal resolutions. These datasets usually are based on remote-sensing techniques or the combination of land surface models (LSMs) and general circulation models (GCMs). This research addressed the efficiency of ECMWF reanalysis dataset and PERSIAN for hydrological modeling using VIC-3L large-scale model over the SefidRood catchment. The results of statistical analyses at daily time scale indicated that the correlation coefficient (CC) between ECMWF, PERSIAN, and ground-observed dataset is about 0.83 and 0.48, respectively. In addition, at monthly time scale, the performances of both rainfall datasets approximately are the same and in most parts of the catchment, the value of CC is higher than 0.80. Hydrological analyses by VIC-3L model showed that despite having low efficiency in estimating rainfall, the PERSIAN dataset led to better simulation of runoff when it compared to ECMWF. For example, the Nash-Sutcliffe (NS) coefficient between daily and monthly simulated runoff using PERSIAN and observed runoff at the outlet of SefidRood catchment are about 0.80 and 0.88, respectively, while in the case of ECMWF these coefficients are about 0.67 and 0.72. Moreover, by using the PERSIAN dataset, the performance of the VIC model in simulating daily and monthly peak flows significantly increases.