Prediction of Catchment-Scale Efficiency of Stormwater Control Measures in an Urban Watershed using a Process-Based Modelling Approach

Stormwater control measures (SCM)s are practices aiming at reducing the volume and peak of surface runoff and pollutant loads through increased infiltration, evaporation, filtration or delayed release to the traditional sewer systems or receiving surface waters. To evaluate the effectiveness of SCMs, it is important to consider their effects on surface runoff as a result of not only their direct effect on capturing overland flow but also through their impacts on infiltration, inter-flow, groundwater recharge and base-flow into the streams. In this study, the application of a process-based model to predict the long-term impacts of SCMs on the hydrologic response of a highly urbanized watershed in the suburbs of Washington DC, is demonstrated. For this purpose a model representing pre-retrofit conditions of the watershed is constructed and calibrated based on observed flow data in the main stream. The watershed system is represented using a compartment-based model where each compartment represents one of overland flow, unsaturated soil, groundwater or segments of the stream network or different components of the SCMs. The model allows incorporating groundwater-surface water interaction which plays an important role on the baseflow in the stream. After model calibration several scenarios of SCM implementation has been added to the model and the long-term impact of the SCMs on the changes of the some hydrographs' long aggregate characteristics, specifically the stream-flow rates corresponding to given exceedance probabilities have been evaluated. The SCMs studied include simple stormwater ponds and SCMs with storage or engineered soil layers with and without underdrain pipes. The results help quantifying the long-term impacts of various level of SCM coverage on the base and peak flow in the main stream. It was found that a high coverage implementation of SCMs (50\%-90\%) can substantially reduce the peak stream flows as measured by the corresponding streams flows to 1\% and 0.1\% exceedance probabilities. It was also found that presence of under-drain systems decreases the peak flows substantially. It was also found that the spatial distribution of the SCMs over the watershed area can have a significant effect on their effectiveness in reducing the peak-flow.