Overview       Research       Support Programs       Focus Catchments   


Urban Stormwater Quality: A Statistical Overview

Hugh Duncan

Publication Type:

Technical Report
This is a publication of the initial CRC for Catchment Hydrology

CRC Program:

Urban Hydrology (Previous CRC)

Publication Keywords:

Storm Sewage
Water Quality (Natural Waters)
Urban Areas
Pollution (Surface Water)
Pollution Sources
Statistical Analysis
Quality (Microbiological)

Abstract / Summary:


This report presents a statistical overview of urban stormwater quality, obtained by analysing the results of many investigations reported in the literature. The objective is to assess the broad scale behaviour of urban runoff quality, and its interactions with land use and other catchment characteristics. A data management system that permits consistent and objective comparison between studies is described.

Information on each water quality parameter includes a brief description of the contaminant, its likely sources and possible effects, the statistical distribution of concentration data, means and standard deviations for each land use with sufficient data, and significant relationships with catchment characteristics. Correlations between water quality parameters are also investigated.

Concentrations are approximately log-normally distributed for all water quality parameters investigated except pH, which is approximately normally distributed. Concentrations of suspended solids, total nitrogen, and total phosphorus are on average highest for agricultural catchments, intermediate for urban catchments, and lowest for forested catchments. Concentrations of total lead, BOD, COD, total coliforms, fecal coliforms, and fecal streptococci are higher on average from high urban catchments than from all low urban catchments taken as a single group.

Roads are a major source of most contaminants in urban runoff. This is due to their lower elevation as well as their vehicular traffic. Concentrations from roofs are substantially lower on average than concentrations from roads and all high urban zonings, for all parameters tested except zinc. Within urban areas, residential zonings tend to produce lower concentrations of metals and organic carbon, and higher concentrations of phosphorus and microbiological measures than the other urban zonings, but the explanatory power is low.

Urban sites with higher mean annual rainfall produce lower stormwater concentrations, on average, for most metal and non-metal parameters, but not for the microbiological measures. For suspended solids, which shows the strongest effect, increasing the mean annual rainfall by 500 mm approximately halves the most likely concentration in runoff. Sites with higher population density produce higher stormwater concentrations, on average, for total nitrogen, BOD, and fecal coliforms, and perhaps for COD and total coliforms, but not for metals.

Correlations between water quality parameters over many measurement sites are often low. As a result, there are only a few cases where one quality parameter can provide a good estimate of another parameter. No single quality parameter can provide a good estimate of a range of other parameters.

The explanatory power of all normally reported catchment characteristics is low, which implies that one or more important explanatory variables are yet to be recognised. Possible contenders include geological age of catchment rocks and soils, and short term rainfall intensity. A higher level of detail in modelling may also be helpful, but would be difficult to apply in practice.

Acrobat Icon 1 Download(s)

Centre Office:
CRC for Catchment Hydrology
Dept of Civil Engineering
Building 60
Monash University Vic 3800

Tel: +61 3 9905 2704
Fax: +61 3 9905 5033