An integrated modeling process to assess water quality for watersheds

An integrated modeling process has been developed that combines remote sensing, Geographic Information Systems (GIS), and the Agricultural NonPoint Source Pollution (AGNPS) hydrologic model to assess water quality of a watershed. Remotely sensed Landsat Thematic Mapper (TM) images were used to obtain various land cover information of a watershed including sub-classes of rangeland and wheat land based on the estimates of vegetative cover and crop residue, respectively. AGNPS model input parameters including Universal Soil Loss Equation's (USLE) cropping factors (C-factors) were assigned to the landcover classes. The AGNPS-ARC INFO interface was used to extract input parameters from several GIS layers for the AGNPS model during several selected storm events for the sub-watersheds. Measured surface water quantity and quality data for these storm events were obtained from U.S. Geological Survey (USGS) gaging stations. Base flow separation was done to remove the base flow fraction of water and total suspended sediment (TSS), total nitrogen (total-N), and total phosphorous (total-P) from the total stream flow. Continuous antecedent moisture content ratios were developed for the sub-watersheds during the storm events and were used to adjust the Soil Conservation Service-Curve Numbers (SCS-CN) of various landcovers. A relationship was developed between storm amounts and estimated energy intensity (EI) values using a probability method (Koelliker and Humbert, 1989), and the EI values were used in running the AGNPS model input files. Several model parameters were calibrated against the measured water quality data and then the model was run on different sub-watersheds to evaluate the modeling process. This modeling process was found to be effective for smaller sub-watersheds having adequate rainfall data. However, in the case of large sub-watersheds with substantial variations of rainfall and landcover, this process was less satisfactory. This integrated modeling process will be useful for the resource managers to assess the water quality of sub-watersheds and identify areas within a watershed for which to develop action plans for land use decisions such as adoption of best management practices (BMPs) to improve water quality.*; *This dissertation includes a CD that is compound (contains both a paper copy and a CD as part of the dissertation). The CD requires the following applications: Microsoft Excel; AGNPS ARC INFO (program included); NDVI Estimation (included) for ERDAS IMAGINE software (not included); Winzip; Real Jukebox.