New Insights into the Quality of Urban Storm Water in South Eastern Australia

Abstract

Quantifying the quality of urban storm water is an important prerequisite to the effective management of urban runoff, which is recognized as the major nonpoint source of pollution in urban areas. Although data on urban storm-water quality are widely available, they are often based on relatively limited data sets, usually containing few samples per event and/or few events per catchment. This paper reports on a large scale monitoring of the key storm-water pollutants found in urban discharges during both wet and dry weather from seven urban catchments in South Eastern Australia. The catchments are all separately sewered (with wholly piped systems) with varying sizes and land uses. Using the same monitoring technique, between 16 and 52 pollutographs were captured at each site for total suspended solid (TSS), total phosphorus, and total nitrogen (TN), while event mean concentrations (EMCs) of heavy metals and major ions, as well as species of N and P, were recorded at a subset of sites. It was found that EMCs of TSS were around 50% less than have been typically reported in earlier literature. During wet weather, nutrients were similar to previously reported, as were most metals concentrations. However, zinc concentrations were significantly higher than previously reported. EMCs of TSS were higher during storm flows than in baseflow, while TN concentrations were consistently higher during baseflow. EMCs of all pollutants monitored were poor with simple hydrological parameters (e.g., event rainfall depth); however, event pollution loads correlated very well with the rainfall intensity to a power, summed over the event duration. It was not possible to distinguish an impact of land use on pollutant concentrations. The first-flush effect was found not to be significant at all sites except the smallest catchment with the simplest drainage layout (the roof of a large building). All these findings have significant implication for treatment strategies with the significantly lower than previously observed TSS requiring consideration in future modeling and treatment design.