Abstract:
Pollutant first flush in urban stormwater runoff is an important phenomenon influenced by a range of
rainfall and catchment related variables. Even though numerous studies have been undertaken to
mathematically define the first flush and the influential variables of first flush, limited research have
been carried out to rank such variables in terms of their level of importance in generating first flush.
Identifying the degree of importance of the variables is critical for accurate predictions of first flush
occurrence and understanding the main drivers of first flush. This research study undertook a
comprehensive analysis of the variables influencing the predictions of first flush occurrence and their
relative importance. The study results are expected to contribute to more accurate predictions of first
flush by affording greater importance to the highly ranked factors and their impacts. The study outcomes
confirmed that total rainfall depth was the most important variable influencing the prediction of first
flush events while the maximum intensity was the second. Rain duration, runoff depth, runoff peak and
average intensity were the next four most important variables. Antecedent dry period and effective
impervious area fraction had relatively low ranking while the time of concentration and the event mean
concentration were found to be the least important variables. Furthermore, the study outcomes highlight
that the use of a combination of variables and due consideration of their interactions can yield better
results than considering their individual roles.
