Abstract
Climate change will modify the spatio-temporal variation of hydrological variables worldwide, potentially leading to more extreme events and hydraulic infrastructure failures. Changes in extreme sub-daily precipitation must be accounted for in urban infrastructure planning and management; hence in intensity-duration-frequency (IDF) curves which represent the primary climatic input to urban water infrastructure design. The IDF_CC tool is a computerized tool for developing IDF curves under a changing climate that is widely used in adaptation studies, particularly in Canada. Earlier versions of the tool generate IDF curves using equidistant quantile mapping (EQM), while the latest version uses quantile delta mapping (QDM). In this study, we critically assess both algorithms to propose new and improved versions and compare their performance to the original algorithms at four locations in Eastern Ontario, Canada. In total, 3,600 simulated precipitation intensity series of the control (1976-2005) and future (2011-2100) periods under the RCP 4.5 and RCP 8.5 climate scenarios derived from four global climate models (GCMs) and four regional climate models (RCMs) were analyzed. The original EQM method was found to be insensitive to the climate change signal whereas the original QDM used an inadequate mapping function between quantiles of observed and RCM-simulated precipitation time series, leading to misrepresentations of future climate conditions. In general, the new versions outperformed the original versions and showed consistent projections, hence providing better inputs for adaptation assessments.