| In order to resolve the spatial component of the design of a water quality monitoring network, a methodology has been developed to identify the critical sampling locations within a watershed. The methodology, Critical Sampling Points (CSP), focuses on a single contaminant, namely total phosphorus, and is applicable to small, upland, predominantly agricultural-forested watersheds. The methodology incorporates a geographical information system (GIS) for spatial analysis and data manipulation purposes, a hydrologic/water quality simulation model for estimating the total phosphorus loads, and an artificial intelligence technology, known as fuzzy logic, for improved input data representation. The input data that are considered vital in the determination of the critical sampling locations include a number of hydrologic, topographic, soil, vegetative, and land use factors. The model also includes an economic and logistics component in order to estimate the number of sampling points required for a given budget and to screen accessible sample stream reaches in the analysis, respectively. The CSP methodology was translated into a model, denominated as WQMSA (W&barbelow;ater Q&barbelow;uality M&barbelow;onitoring S&barbelow;tation A&barbelow;nalysis), which can be used by watershed managers designing water quality monitoring networks. The WQMSA model was developed on a PC platform and was primarily implemented utilizing the geographic information system ArcView® as its interface and programming environment. An important goal in the development of the CSP methodology was to make it universally applicable to watersheds, regardless of their geographic position, and hence it is designed to have few input data requirements so that even remote watersheds with scarce data availability can still be evaluated.; In order to test the validity of the CSP methodology, a small rural, experimental watershed in Pennsylvania was selected, for which total phosphorus data from a number of single storm events were available for various sampling points within the watershed. A comparison of the ratios of observed to predicted TP loads at the four sampling locations within the watershed revealed that the WQMSA model's results approached the expected TP load ratios at high storm flows/high storm volumes, which are what greatly determine the annual total phosphorus loads of a watershed. However, since the largest storm events were not available for the watershed under study, this observation could not be conclusively proven. Nevertheless, the application demonstrated that the results obtained from the WQMSA model were promising. (Abstract shortened by UMI.)... |
