Accounting for the Spatio-Temporal Variability of Pollutant Processes in Stormwater TSS Modeling Based on Stochastic Approaches

Abstract

Stormwater quality modeling remains one of the most challenging issues in urban hydrology today. The processes involved in contaminant generation and transport are very complex, with many associated uncertainties, including uncertainty arising from process variability. In this study, the spatio-temporal variability of build-up/wash-off processes in a heterogeneous urban catchment within the Parisian region is assessed based on three stochastic modeling approaches integrated into the physically based distributed hydrological model, the Urban Runoff Branching Structure (URBS) model. Results demonstrate that accounting for process variability at the scale of a hydrological element is important for analyzing the contamination recorded at the catchment outlet. The intra-event dynamics of total suspended solids (TSS) were most accurately selected for the stochastic exponential SWMM model, as this model succeeded not only in simulating the general trend of TSS concentrations fluctuations but also in replicating multiple peaks observed in pollutographs. The advantage of this approach is that it captures the stochastic nature of the processes with minimal prior knowledge and without extensive calibration, though further enhancement is necessary for it to become a useful tool to support decision making.