Simulation of Runoff and Pollutant Loss in Urbanizing Watersheds
The effect of urbanization on previously agricultural watersheds is an increasingly important issue for watershed planners. Urbanization increases runoff and pollutant loadings to the watershed outlet. Watershed planners in areas that previously had little impervious cover must now consider the effects of new roads and buildings on hydrologic processes. The ANSWERS-2000 watershed model was modified to simulate watersheds with mixtures of agricultural and urban areas. In addition, components were added to simulate atmospheric deposition and urban management practices, including wet ponds, dry ponds, and infiltration trenches.
The modified model was evaluated on two watersheds in Blacksburg, Virginia, including a subwatershed of Stroubles Creek and a large parking lot on the Virginia Tech campus with a dry pond at its outlet. The model predicted the hydrology and pollutant losses for the year 1999 from the Stroubles Creek watershed within 50% of the observed values after calibration. Prediction errors were much higher for the parking lot and dry pond simulation of the period of time from August 1995 to February 1996. For the parking lot inflow to the dry pond, errors ranged from 0 to 100%. For the dry pond effluent, errors for runoff and sediment losses were -11.5 and 60.1%, respectively, and nutrient losses were poorly predicted (greater than 100% error). There was considerable uncertainty as to the quality of the observed data and this may account for some of the predicted sediment and nutrient loss errors. The modified model was applied to the Battlefield Green Watershed in Hanover County, Virginia to demonstrate the watershed response to development in that watershed. As simulated, sediment and nutrient losses were 30 to 50 times higher after development.
The model is intended for use on watersheds with an impervious cover of 30% or less, due to the increased difficulty in accurately quantifying the hydrology of highly urbanized watersheds and because of uncertainty in atmospheric deposition rates on such watersheds. The pond subroutines are very simplified, and limit simulation to ponds with simple geometries.