Effective stormwater system management is becoming increasingly challenging, in part due to aging stormwater infrastructure and the desire for the stormwater system to function for decades. Complicating these challenges are the implications of climate change which may influence the intensity of precipitation events. Climate change may be changing precipitation intensities which the buried stormwater system is intended to handle. The presentation reviews some of the findings observed to date, regarding buried stormwater system design, including:
i) Insights into why the use of the Gumbel Extreme Value distribution is reasonable as a probably distribution;
ii) Assessment of the observed intensities of precipitation records indicate that at some locations, there have been increases of precipitation
iii) intensity (e.g. at Waterloo, the 5-year recurrence storm of 1 hour duration has increased at a rate of 2.24%/year over the history of record) while at others (e.g. at Pearson airport, the 5 year return period storm of 1 hour duration has decreased at 1.2%/year). The implications for design for infrastructure intended to serve for 40 years are substantial;
iv) The presentation will describe the differences between, and ultimately, the merit of using, the partial duration series rather than the annual maxima precipitation series in relation to assignments of precipitation rates for infrastructure design; and,
v) The length of record needed, to achieve a 95% confidence interval as small as +/- 10% of the prediction, is assessed. The findings indicate that there is a considerable challenge given available records which indicate, for example, at Kingston, one needs 45 years of record for the 5 year storm, and a 70 year record for estimating within +/-10% for the 25 year storm. Confidence of precipitation intensities for design storms represents a major challenge.