Extreme monthly rainfall
Extreme monthly rainfall in a changing climate
The new CSIRO and BOM (2015) climate projections for Australia are largely based on simulations of the coming century from an ensemble of some 40 CMIP5 global climate models. By 2080-2099, with strong global warming under the RCP8.5 forcing scenario, the median projection for mean annual rainfall is a decrease of around 10%, aside from little change in northern Australia, however the 10 to 90 percentile range is typically 30% about the median. Projections for extreme daily rainfall within a year, based on the models, are mostly positive, indicating a broadening of the distribution of daily rainfall amounts. We focus on the distribution of monthly mean amounts, a time scale with important implications for agricultural and pastoral activity and for large-scale flood events. Averages of the top 10 percent of observational (ERA-Interim) monthly amounts, partitioned by season, at most locations over Australia are typically two to three times the mean monthly rainfall for the season. The extremes are relatively smaller in the south in winter months.
Initial CMIP5 results, from the ACCESS1.3 and CESM1-CAM5 models, show good agreement with this. Composite maps, for the extreme rainfall months for selected locations, show that rainfall anomalies extend over much of Australia, and are associated with cooler daily maximum temperatures. There are circulation anomalies consistent with regional moisture flux convergence, often associated with a large-scale La Niña pattern. Regional rainfall and temperature anomalies show some persistence, over following months. Under strong global warming, there is a similar character to the anomalies of both rainfall and associated circulations. However, there is some indication of a broadening of the rainfall distributions. Using multiple models, projections of the high monthly rainfall amounts can be made, with changes attributed to a combination of changes in atmospheric saturation humidity and in large-scale circulation patterns.