Modelling the washoff of pollutants in various forms from an urban catchment

Abstract

The exponential washoff model was originally developed based on observations of particulate pollutants, however, its applicability when applied to different forms of pollutants is not well understood. Data from a previous study of 6 stormwater pollutants from 126 events at 12 sites in Singapore was used for event based model parameter calibration using a Monte Carlo technique. The accuracy of the calibrated exponential washoff model was clearly best for particulate pollutant total suspended solids (TSS), and worst for dissolved pollutants Ortho-Phosphate (PO4), nitrate (NO3) and ammonium-nitrogen (NH4). Model accuracy for mixed forms of pollutants total Phosphorus (TP) and total Nitrogen (TN) were in between these two extremes. Relationships between model parameters with rainfall and flow characteristics were also investigated. Statistically significant relationships could only be found for TSS, where the total rainfall depth was identified as being the most significant variable to explain model parameter behaviour. Antecedent dry period (ADP) was shown to have little or no importance across all land uses and pollutant forms. The results showed that the model parameter behaviour could be explained only for particulate pollutants and small (≤10 ha) sub-catchments, and that replicating washoff of mixed or dissolved forms of pollutants as a fraction of solids is likely to lead to misleading results.