Abstract
One of the most significant anticipated consequences of global climate change is the increased frequency of hydrologic extremes. Predictions of climate change impacts on the regime of hydrologic extremes have traditionally been conducted using a top‐down approach. The top‐down approach involves a high degree of uncertainty associated with global circulation model (GCM) outputs and the choice of downscaling technique. This study attempts to explore an inverse approach to the modelling of hydrologic risk and vulnerability to changing climatic conditions. With a focus targeted at end‐users, the proposed approach first identifies critical hydrologic exposures that may lead to local failures of existing water resources systems. A hydrologic model is used to transform inversely the main hydrologic exposures, such as floods and droughts, into corresponding meteorological conditions. The frequency of critical meteorological situations is investigated under present and future climatic scenarios by means of a generic weather generator. The weather generator, linked with GCMs at the last step of the proposed methodology, allows the creation of an ensemble of different scenarios, as well as an easy updating, when new and improved GCM outputs become available. The technique has been applied in Ontario, Canada. The results show significant changes in the frequency of hydro‐climatic extremes under future climate scenarios in the study area. Copyright © 2007 John Wiley & Sons, Ltd.