Evaluating hydrologic responses to soil characteristics using SWAT model in a paired-watersheds in the Upper Blue Nile Basin

Abstract

Watershed responses are affected by the watershed characteristics and rainfall events. The characteristics of soil layers are among the fundamental characteristics of a watershed and they are input to hydrologic modeling similar to topography and land use/cover. Although the roles of soils have been perceived, there are limited studies that quantify the role of soil characteristics on watershed runoff responses due to the lack of field datasets. Using two adjacent watersheds (Ribb and Gumara) which have a significant different runoff response with a similar characterstics except geological settings (including soil characteristics), we studied the effects of soil characteristics on runoff and water balance. The Soil and Water Assessment Tool (SWAT) was used to simulate the surface runoff response at the outlet of the watershed and the optimal model parameters distribution was tested with a non-parametric test for similarity. Results indicated that SWAT model captured the observed flow very well with a Nash-Sutcliffe Efficiency (NSE) of greater than 0.74 and with a PBIAS of less than 10% for both calibration and validation period. The comparison of the optimal model parameter distributions of the SWAT model showed that the watershed characteristics could be uniquely defined and represented by a hydrologic model due to the differences in the soils. Using field observations and modeling experiments, this study demonstrates how sensitive watershed hydrology is to soils, emphasizing the importance of accurate soil information in hydrological modeling. We conclude that due emphasis should be given to soil information in hydrologic analysis.