January 27, 2014

A Comparison of Alternative Techniques for Deriving Extreme Rainfall Statistics in the Context of Climate Change


In Ontario, rainfall intensity-duration-frequency (IDF) statistics are used extensively in water management. Given the anticipated impacts of climate change on extreme rainfall, there is a great deal of interest by municipalities, conservation authorities, provincial agencies, infrastructure proponents, and risk managers in developing rainfall IDF statistics that reflect anticipated future climate conditions, so that these can be reflected in design and analysis.

There are however, a number of challenges involved in the derivation of future IDF statistics. For example, there are a number of methods for applying climate model output to derive future extreme rainfall statistics. Within the existing literature on this topic, different studies have used varying combinations of global climate models, downscaling techniques, spatial and temporal scales, emission scenarios and curve-fitting methods. The result of all this work has been divergent or inconsistent results among future IDF datasets even for the same study area, making it difficult for practitioners to use this information in the design of infrastructure and management programs in the context of climate change. This raises the question as to how the level of uncertainty associated with future IDF curves influences their use in infrastructure design and risk management in the context of climate change adaptation.


The overall goal of this study is to develop information on the level of uncertainty associated with future IDF curves so practitioners can determine how data should, or should not, be used in design and management applications under a changing climate. More specifically, this study compares how the selection of different global climate models and downscaling techniques influences future IDF statistics in two study areas in Ontario: The Greater Toronto Area (GTA) and Essex Region.