[1] An optimization algorithm linked with a nonpoint source (NPS) pollution model can be used to optimize NPS pollution control strategies on a field-by-field basis in a watershed by maximizing NPS pollution reduction and net monetary return. In this paper a methodology is described which integrated a genetic algorithm (GA) (an optimization algorithm) with a continuous simulation, watershed-scale, NPS pollution model, Annualized Agricultural Non-Point Source Pollution model (AnnAGNPS) to optimize the selection of best management practices (BMP) on a field-by-field basis for an entire watershed. To test the methodology, optimization analysis was performed for a U.S. Department of Agriculture experimental watershed in Pennsylvania to identify BMPs that minimized long-term (over a 4-year period) water quality degradation and maximized net farm return on an annual basis. Results indicate that the GA was able to identify BMP schemes that reduced pollutant load by as much as 56% and increased net annual return by 109%.

The Soil and Water Assessment Tool (SWAT) model, designed for use on rural ungaged basins and incorporating a GRASS GIS interface, was used to model the hydrologic response of the Ariel Creek watershed of northeastern Pennsylvania. Model evaluation of daily flow prior to calibration revealed a deviation of runoff volumes (Dy) of 68.3 percent and a Nash-Sutcliffe coefficient of -0.03. Model performance was affected by unusually large observed snowmelt events and the inability of the model to accurately simulate baseflow, which was influenced by the presence of fragipans. Seventy-five percent of the soils in the watershed contain fragipans. Model calibration yielded a D of 39.9 percent and a Nash-Sutciffe coefficient of 0.04, when compared on a daily basis. Monthly comparisons yielded a Nash-Sutcliffe coefficient of 0.14.Snowmelt events in the springs of 1993 and 1994, which were unusually severe, were not adequately simulated. Neglecting these severe events, which produced the largest and third largest measured flows for the period of record, a D of 4.1 percent and Nash-Sutcliffe coefficient of 0.20 were calculated on a daily comparison, while on a monthly basis the Nash-Sutciffe coefficient was 0.55. These results suggest that the SWAT model is better suited to longer period simulations of hydrologic yields. Baseflow volumes were accurately simulated after calibration (D = -0.2 percent).Refmements made to the algorithms controlling subsurface hydrology and snowmelt, to better represent the presence of fragipans and snowmelt events, would likely improve model performance.

Abstract:A 40 m ð 20 m mowed, grass hillslope adjacent to a headwater stream within a 26-ha watershed in east-central Pennsylvania, USA, was instrumented to identify and map the extent and dynamics of surface saturation (areas with the water table at the surface) and surface runoff source areas. Rainfall, stream flow and surface runoff from the hillslope were recorded at 5-min intervals from

Abstract“Traditional” statistical analyses based on the assumption of independent observations are being replaced by spatial analyses that take account of correlations between neighboring observations. Geostatistics is one approach; it characterizes the spatial relationships of data via the variogram, which is in turn used for kriging (optimal, unbiased linear interpolation). Exploratory data analysis techniques, relying on resistant measures, graphical tools, and robustness ideas, can be used to help “model” the spatial structure of data. Data should fulfill certain stationarity conditions before computing and interpreting the semivariogram. Data of soil‐water pressure potential are analyzed by straight‐forward techniques that assure the data meet the implicit assumptions of stationarity (of the mean and the variance) and at least symmetry. Stem‐and‐leaf plots and plots of mean vs. variance (or for a more resistant analysis, median vs. interquartile range squared) are used to assess the variance stationarity and data distributions. Median‐based techniques (rather than polynomial modeling and generalized least‐squares fitting of drift) are used to remove drift along both grid directions. Then the spatial structure is exposed through computing and interpreting semivariograms of the modified data.

No abstract