Sensitivity analysis of a GIS-based model: A case study of a large semi-arid catchment

Water resource and hydrologic modeling studies are intrinsically related to spatial processes of hydrologic cycle. Due to generally sparse data, and high rainfall variability, the accurate prediction of water availability in complex semi-arid catchment depends to a great extent on how well spatial input data describe realistically the relevant characteristics. The Geographic Information System (GIS) provides the framework within which spatially distributed data are collected and used to prepare model input files. Despite significant recent developments in distributed hydrologic modeling, the over-parameterization is usually a critical issue that can complicate calibration process. Sensitivity analysis methods reducing the number of parameters to be adjusted during calibration are important for simplifying the use of these models. The objective of this paper is to perform a sensitivity analysis for flow in a semi-arid catchment (1,491 km2), located in northwestern of Tunisia, using the Soil and Water Assessment Tool (SWAT) model. The simulation results revealed that among eight selected parameters, curve number (CN2), soil evaporation compensation factor (ESCO), soil available water capacity (SOL_AWC) and threshold depth of water in the shallow aquifer required for return flow (GWQMN) were found to be the most sensitive parameters. Calibration of hydrology, facilitated by the sensitivity analysis, was performed for the period 2001 through 2003. Results of calibration showed that the model accurately predict runoff and performed well with a monthly Nash Sutcliffe efficiency (NSE) of 0,78, a coefficient of determination (R²) of 0,85 and a percent of bias (PBIAS) equal to −13,22 %.