Modeling impact of Land Use/ Cover changes on Water Quality and Quantity of Fish River Watershed

Abstract

Hydrologic and water quality models are widely used for assessing the impacts of land use/cover (LULC) changes on water quality and quantity. However, their credibility for predicting LULC change impacts has not been tested with observed water quality. The main objective of this study was to test the Soil and Water Assessment Tool (SWAT), in predicting variations in water quality/quantity due to alterations in LULC over time in the Fish River watershed. Fish River is the main freshwater source of the Weeks Bay, which is one of the only three designated Outstanding National Resource Waters in the state of Alabama. A water quality model was setup for the Fish River watershed using the SWAT model. The model was first calibrated and validated for hydrology at a USGS flow monitoring station on the Fish River for the period 1990 to 1998 using the 1992 National Land Cover Dataset as the LULC input. The SWAT model was then calibrated and validated for water quality (NO3-, Org-P and TSS) using data collected during 1994-1998 at six tributaries of the Fish River. Model performance was best for flow, and weakest for NO3-. The calibrated SWAT model was later fed by 2008 LULC data and model simulations of flow and water quality were compared to their observed counterparts from the period 2008–2009 (post-validation). Post-validation results closely followed the calibration and validation trends. This study showed that SWAT is a reliable tool in predicting the impact of LULC changes on flow and water quality. Often, it is not adequate to know that changes in LULC will cause a certain amount of increase or decrease of a particular or group of water quality constituents. In case of undesired increases, locating the critical source areas (CSA) of pollutants has practical implications from management perspective. Distributed watershed models are frequently used for identification of CSAs of pollutions in watersheds. One of the main inputs to these models is the spatially-explicit soils data. The second objective of this study was to evaluate if the use of two commonly used soil datasets, the State Soil Geographic (STATSGO) and the Soil Survey Geographic (SSURGO) data, can lead to differences in location of CSAs of sediment. The use of STATSGO soil data resulted in higher soil erodibility factor and surface runoff. As a result, higher sediment yield was obtained from the use of the STATSGO data as compared to the sediment yield obtained from the use of the SSURGO data. Therefore, for accurate identification of CSAs of sediment (and potentially other pollutants) and for effective implementation of economically-feasible best management practices (BMPs), it is important to use the most detailed spatial dataset available.