Extreme summer temperatures have a significant impact on Australasia, and these are expected to intensify with global warming. The simulation of these extremes, particularly near coasts, is limited by the modest resolution of global climate models such as the CSIRO Mk3.0. We examine the temperatures simulated by the regional model CCAM, at a resolution of around 0.5°, downscaled from Mk3.0 simulations. Extremes for three periods, from monthly to decadal, are presented. The distribution of daily temperatures in summer is characterised by a large variation of standard deviation (SD) across the region, peaking along the southern Australian coast. The distribution is positively skewed along most coasts, and negatively skewed in the interior. There are strong correlations with daily winds. The projected mean warming over the 21st century for the A1B greenhouse gas scenario reaches 3°C in the Australian interior, but barely half that on the south coast, as anticipated from the smaller rate of warming of the surrounding seas. There is rather little moderation of warming on the east and west coasts, however. Warming over New Zealand and New Guinea is typically 2°C, for both means and extremes, some 50 per cent larger than that of the oceans. The SD of temperature anomalies increases over much of Australia, while skewness increases on some coasts. Both SD and skewness decrease in the southeast. Changes in the extremes, relative to the means, relate to those in both measures. This is demonstrated using a beta distribution fit to local daily temperature distributions. At many locations, changes in extremes cannot be accurately estimated from a simple shift of the distribution. Much of the pattern of change is related to the effect of wind variability over the enhanced land-sea contrasts, as shown using a simple model for the influence of wind.
Watterson et al. (Mon,) studied this question.