Technical guidelines for future intensity-duration-frequency curve estimation in Canada

Intensity-duration-frequency (IDF) curves need to be reliable to serve as a relevant tool in preventing or reducing potential damage to society. This often requires accounting for the effect of climate change due to evidence of its effect on extreme precipitation. There is a large number of approaches for IDF curve estimation under climate change in the literature; however, a general framework with practical guidelines for facilitating their application by practitioners is not readily available. The aim of the present study is to provide practical guidelines and recommendations for helping federal and provincial agencies, as well as others who might produce practice guidelines, to develop standardized procedures for the estimation of future IDF curves in Canada that can then be used by practitioners in infrastructure design, management and risk assessment. This is done by gathering and summarizing findings in rainfall frequency analysis from the Canadian FloodNet Research Group, under a practically oriented perspective. Technical recommendations are presented within a methodological framework to facilitate understanding; decision-making procedure-specific flowcharts are provided to facilitate their application. The proposed methodological framework is based on the use of pooled frequency analysis for reasonable estimation of extreme rainfall intensities, and on the estimation of gridded relative changes for IDF updates under climate change following different approaches depending on rain gauge network density. In particular, three methods that do not compete against each other but rather are different methodologies to be applied depending on the case study are proposed: pooled estimation of extreme rainfall in Canada (Method I), gridded relative changes in 24-h extreme rainfall intensities in Canada (Method II) and gridded relative changes in 24-h and sub-daily extreme rainfall intensities in regions of Canada with a relatively high station density (Method III).