MODELLING DRY AND WET EXTREMES OVER THE CANADIAN PRAIRIE PROVINCES BASED ON THE DYNAMICAL DOWNSCALING AND MULTIVARIATE FREQUENCY ANALYSIS APPROACHES

Abstract

The primary tools to assess climate change are the Atmosphere–Ocean General Circulation Model (AOGCM) or Regional Climate Model (RCM) transient climate change simulations. Currently, RCMs offer higher spatial resolution than AOGCMs and therefore are preferred for assessing impact of climate change on different components of the hydrological cycle at regional domains of interest. The overall purpose of this research was to evaluate the impact of climate change on dry and wet climate extremes over the Canadian Prairie Provinces of Alberta, Saskatchewan and Manitoba using a multi–RCM ensemble from the North American Regional Climate Change Assessment Program (NARCCAP). This region of Canada is characterized by highly variable hydroclimate, with recurrent droughts and floods and localized summer convective storm activity often resulting in heavy precipitation events and thus poses many challenges for water managers. At first, the Saskatchewan River Basin, the largest river in the study area, was evaluated and drought vulnerable parts of the basin were identified based on historical data and multivariate frequency analysis approaches. For the development of projected changes to drought characteristics, the research effort was extended over the entire study area and changes to various return levels of drought severity, duration and maximum severity were developed based on NARCCAP RCM simulations and multivariate frequency analysis approaches. It was found that the southern and south-western parts of the study area will experience increased drought severity in the future. Based on the projected bi- and trivariate joint occurrence probabilities of drought characteristics, southern parts along with the central parts of the study area were found to be highly drought vulnerable, whereas the southwestern and southeastern parts were found less vulnerable. Though producing reliable estimates of changes in precipitation extremes remains an important challenge under climate change, this study attempted to develop projected changes to April–October short- and long-duration precipitation extremes based on the NARCCAP RCM simulations and regional frequency analysis approach. Projected changes to selected regional return levels of precipitation extremes were found mostly statistically significant, with relatively larger changes noted for the southeastern regions and smaller for the southwestern and western regions of the study area.