Main content area

Calibrating WEPP Model Parameters for Erosion Prediction on Construction Sites

Moore, A.D., McLaughlin, R.A., Mitasova, H., Line, D.E.
Transactions of the ASABE 2007 v.50 no.2 pp. 507
Water Erosion Prediction Project, buildings, water erosion, runoff, prediction, watersheds, sediment yield, North Carolina
Soil erosion on construction sites can be many times greater than on agricultural fields, yet there has been little modeling done for construction conditions. The objective of our study was to calibrate management and soil parameters in the agriculturally based model Water Erosion Prediction Project (WEPP) for construction and post-construction site conditions. Data from a 4 ha watershed at various stages of construction in Wake County, North Carolina, were used to compare model results with measured runoff volume and sediment yields. Model simulations were performed in GeoWEPP, a geospatial interface designed for WEPP that operates within ArcView GIS. Model parameters were adjusted from WEPP default parameters as supported by the literature and site observations. Predicted values were regressed against field data for Nash-Sutcliffe model efficiency (NSE), with NSE > 0.50 regarded as satisfactory performance. We were able to generate runoff and sediment yields comparable to observed values by replacing soil surface properties with subsoil properties, in conjunction with the cutslope management parameter file in WEPP. We found a similar agreement between predicted and observed values for stabilized conditions by increasing critical shear stress from 0.3 to 10 Pa for the soil input layer. In addition, changes to the model source code to reduce the lower limit of effective hydraulic conductivity (Kef) for impermeable surfaces resulted in improved runoff NSE, but consequently increased sediment yield on areas with higher Kef values. WEPP has great potential for modeling applications on construction sites; however, more validation studies are needed to confirm and expand upon our findings.