Performance of a Multi-Model Ensemble for the Simulation of Temperature Variability Over Ontario, Canada

Multi-model ensembles for climate modeling, which have generally proven to have a superior performance compared to individual models, are assessed for Ontario, Canada. Specifically, the performance of seven Global Climate Model (GCM) and Regional Climate Model (RCM) combinations are evaluated on twelve stations across Ontario, as well as for the entire domain. Two multi-model ensembles were produced, one using the mean of seven GCM and RCM combinations and the other using the median of the same seven GCM and RCM combinations. Three temperature variables (average surface temperature, maximum surface temperature, and minimum surface temperature) were used to evaluate the performance of the models, as well as twelve stations chosen within the domain. Data obtained from the North American Coordinated Regional Downscaling Experiment were compared with gridded data based on observations from the Climactic Research Unit’s TS v4.00 dataset, as well as observed station data from the Digital Archive of Canadian Climatological Data provided by Environment and Climate Change Canada. For all three climate variables, at each station, and over the whole domain of Ontario, the multi ensemble based on the mean generally outperformed the ensemble based on median and each of the individual models. Future predictions of the multi-model ensemble under the Representative Concentration Pathway 4.5 (RCP4.5) scenario are generated to provide bases for the climate change mitigation and adaptation in Ontario. The multi-model ensembles predict a 2.89 increase in annual mean temperature between 1951-2005 and 2040-2069