In order to best assess the expected climate change impacts on a species, ecosystem or natural resource in a region, climate variables and climate change scenarios must be developed on a regional or even site-specific scale (Wilby et al, 2002). To provide these values, projections of climate variables must be ‘downscaled' from the GCM results, utilizing either dynamical or statistical methods (IPCC, 2001).

In this study, three climate variables (maximum temperature, minimum temperature and precipitation) were statistically downscaled, utilizing the output from two general circulation models, the Canadian Coupled Global Climate Model (CGCM2) and the Hadley Center HadCM3 and running the Statistical Downscaling Model (SDSM2), developed by WiIby and Dawson. Two major urban centers (Shearwater, NS (as a proxy for Halifax) and St John's, NL) in Atlantic Canada were the focus of this study. Analyses were performed at each of the two sites comparing the different output from the models as well as giving future scenarios for climate in each tri-decade (2010-2039, 2040-2069, 2070-2099). Shifts in the distributions for temperature were examined to identify changes in mean and variability. Extreme climate indices were calculated for each site to represent local weather extremes; in this case, heat wave occurrence and duration (temperature meeting or exceeding 30C) and extreme daily precipitation amounts.

Results from both sites validated well when the GCM model results were compared to the downscaled values, providing confidence in the downscaling approach. Values developed from the CGCM2 showed an increased mean maximum temperature (3C in the 2070-2099 period) but the variability decreased. In comparison, the HadCM3 results showed the mean maximum temperature increased by 4.73 degrees for the same period and the variability increased. Both sites exhibited more frequent and longer heat waves and higher daily extreme precipitation amounts. Return periods for extreme precipitation events reduced by at least a factor of 2.