Calibration of WinSLAMM to stormwater volume and pollutant mass flux data in Ohio, USA: Informing pollutant loads in untreated stormwater from various urban land uses

Abstract

Runoff hydrologic and water quality models are often utilized to inform pollutant load reductions and guide stormwater management since field-based studies are limited by time, money, equipment failures, and the inability to extrapolate to future climate scenarios. Many empirically based models, such as the Source Loading and Management Model for Windows (WinSLAMM), are used without calibration, thus uncertainty exists as to whether they accurately predict untreated stormwater characteristics in a watershed of interest. Inappropriate use and implementation of such models can lead to over- or under- design of stormwater control measures, reduced land for development, increased cost to developers, and unreasonably set regulations such as total maximum daily loads. Thirteen watersheds representing a variety of urban land use and land covers were modeled in WinSLAMM, calibrated to field collected hydrologic and water quality data, and compared against those data to understand if WinSLAMM is a viable tool to guide stormwater management. Results showed that the Small Storm Hydrology method utilized in WinSLAMM produced suitable model results for runoff volume prediction when watersheds were more than 45% impervious. Despite attempting a variety of water quality calibration techniques, pollutant concentrations were almost always overestimated by WinSLAMM, ultimately resulting in overestimation of annual pollutant loads. Results suggest that empirical models created with decades old data may no longer be useful for present day and future water quality scenarios. This may require WinSLAMM and other empirical tools to be consistently updated with source data and to amend input parameters such as regionalization schemes and rainfall/snowfall characteristics (intensity, dry time, duration, and minimum inter-event time) as the climate changes.