Assessing Digital Soil Inventories for Predicting Streamflow in the Headwaters of the Blue Nile

Abstract

Comprehensive spatially referenced soil data are a crucial input in predicting biophysical and hydrological landscape processes. In most developing countries, these detailed soil data are not yet available. The objective of this study was, therefore, to evaluate the detail needed in soil resource inventories to predict the hydrologic response of watersheds. Using three distinctively different digital soil inventories, the widely used and tested soil and water assessment tool (SWAT) was selected to predict the discharge in two watersheds in the headwaters of the Blue Nile: the 1316 km2 Rib watershed and the nested 3.59 km2 Gomit watershed. The soil digital soil inventories employed were in increasing specificity: the global Food and Agricultural Organization (FAO), the Africa Soil Information Service (AfSIS) and the Amhara Design and Supervision Works Enterprise (ADSWE). Hydrologic simulations before model calibration were poor for all three soil inventories used as input. After model calibration, the streamflow predictions improved with monthly Nash–Sutcliffe efficiencies greater than 0.68. Predictions were statistically similar for the three soil databases justifying the use of the global FAO soil map in data-scarce regions for watershed discharge predictions.