Impact of topographic conditions on the modelling performance of various global precipitation products in a mountainous basin

Abstract

This study aims to conduct a comparative analysis of global precipitation products and ground station data using a hydrological model. The effects of different gridded precipitation datasets and topographic model inputs, such as subbasin delineation and elevation band details, on streamflows were investigated. The study focused on the mountainous Nilufer Basin in Turkiye. Ground station data (GSD) and three different global precipitation datasets -Climate Forecast System Reanalysis (CFSR), Climate Hazards Group InfraRed Precipitation with Station (CHIRPS), and National Aeronautics and Space Administration-Prediction of Worldwide Energy Resources (NASA-POWER)- were used. The Soil and Water Assessment Tool (SWAT) was employed for hydrological modelling and Nash-Sutcliffe Efficiency (NSE) was utilized as the performance criterion for model calibration. The results showed that GSD, CHIRPS, and NASA-POWER achieved reasonable NSE levels (>0.5) without calibration, whereas CFSR performed poorly (NSE<0.2). After calibration, all models indicated successful results (NSE>0.70), with a notable improvement in CFSR (NSE increased from 0.12 to 0.71). Increasing the number of subbasins slightly improved the results, with the highest change in NSE of 0.09. Generating too many subbasins, though, lead to longer processing times without further improvements. However, introducing elevation bands significantly enhanced model performance (NSE increased by 0.21-0.27 across all datasets). An increase in the number of bands yielded only slight improvements, with NSE increasing by 0.03 at most.