The Old Woman Creek watershed, located in the south-western basin of Lake Erie is one of its concerns for the amounts of sediments, nutrients, and other chemical discharges into Lake Erie. BASINS, a watershed-scale model, was used along with its internal models, HSPF and PLOAD, to simulate flow and pollutant loads from the Old Woman Creek watershed. Physical properties of soil and land use, meteorological data and, observed flow data were collected for a 6-year period from 2001 to 2006 and are used in the model development and validation. The model was calibrated for four years (2000-01 to 2003-04) and validated for another two years (2004-05 and 2005-06). For the calibration period, the correlation between the observed and simulated daily runoff was strongly accurate, as shown by the coefficient of determination value of 0.77. The coefficient of determination was 0.81 for the validation period. The Nash-Sutcliffe coefficients obtained were 75.5% and 79.7% for the calibration and validation period, respectively.
The model was run and calibrated by adjusting lower zone evapotranspiration parameter (LZETP), fraction of groundwater inflow to deep recharge (DEEPFR), lower zone nominal soil moisture storage (LZSN), index to mean soil infiltration rate (INFILT), groundwater recession rate (AGWRC), interflow coefficient (INTFW) and, interflow recession coefficient (IRC). Calibration of these parameters improved the model simulation of flow as compared to the initial model run. The calibrated model was applied to the validation data set.
The developed watershed model predicts the flow and pollutant loads and concentration. The model had a commendable success in the prediction of flow, which was calibrated and validated. The pollutant loads and concentration could not be validated because of the lack of observed data for the time series. The developed model was a strong success since all the model performance statistical parameters displayed strong accuracy comparable to the observed flow data. The model validation was a success too with the calibrated model. The daily observed values of stream flows were averaged to obtain monthly values which were then compared with the predicted values. The predicted values conformed well to the average monthly observed values, thus leading to the strong performance of the watershed model.
Data availability and its accuracy play an important role in the development of any watershed model. It can be inferred that instead of using nationally derived datasets, study area specific datasets should be obtained to be used in the modeling process. Studies conducted by Earthtech and Strand Associates Inc. in Rock River Basin, Wisconsin (2000) pointed out discrepancies in data obtained from USGS gauge stations. It can be concluded that the BASINS model with the availability of actual weather data as well as actual field data for modeling the Old Woman Creek watershed.
It was concluded that the model performance can be greatly improved by adopting a more comprehensive method for simulating flow representing all the hydrological components and to simulate various management scenarios to solve hydrologic problems in the Old Woman Creek watershed.