ABSTRACT
The use of hydrological models in planning and management of water resources has become the norm, and a wide array of hydrological models (including freeware) is now available. The Soil Water Assessment Tool (SWAT) (Arnold et al. 1998; Neitsch et al. 2005; Gassman et al. 2007) is one such model. The main data sets required to formulate and run the model include a Digital Elevation Model (DEM), land use, soil, climatic and land use management data sets. The quality of these inputs has a significant impact on the model formulation process and on the results. Many studies have investigated the impact of the resolution of DEM, soil and land use data on the SWAT simulations (Chaplot 2005; Dixon and Earl 2009). Research has also been devoted to examine the impact of catchment subdivisions on the SWAT simulations (Jha et al. 2004; Tripathi et al. 2006). The studies on evaluating the impact of climatic data input on SWAT simulations (discussed below) are gaining increased attention, given the fact that climatic data are a major driver of hydrological and other processes simulated by the model. The current way of climatic data input in the SWAT is rather simplistic. Climatic data of a rain gauge located nearest to the centroid of a subcatchment are used for that subcatchment. This may not be accurate enough, particularly in regions where spatial heterogeneity is high (e.g., mountainous terrains), or where data are sparse but spatial variability of processes is not. This, in turn, has an impact on the model formulation process (e.g., parameterization) and quality of the simulated results (Oudin et al. 2006; Mul et al. 2009). For example, in response to over-predicted rainfall, the model parameterization process may tend to increase ET to match the observed and predicted streamflows. In many cases, finding the appropriate model structure and parameter sets may not work well in delivering acceptable model simulations if the input precipitation is inaccurate. Hence, improved precipitation input is very important to obtain good results (Oudin et al. 2006; Mul et al. 2009; Tobin and Bennet 2009).
