Historical and projected trends in temperature and precipitation extremes in Australia in observations and CMIP5

This study expands previous work on climate extremes in Australia by investigating the simulation of a large number of extremes indices in the CMIP5 multi-model dataset and comparing them to multiple observational datasets over a century of observed data using consistent methods. We calculate 24 indices representing extremes of temperature and precipitation from 1911 to 2010 over Australia and show that there have been significant observed trends in temperature extremes associated with warming while there have been few significant observed trends in precipitation extremes. We compare the observed indices calculated from two mostly independent datasets with 22 CMIP5 models to determine how well global climate models are able to simulate observed climatologies, variability and trends. We find that generally temperature extremes are reasonably well simulated (climatology, variability and trend patterns) although the models tend to overestimate minimum temperature extremes and underestimate maximum temperature extremes. Some models stand out as being outliers and we exclude one model (INMCM4) entirely from the multi-model analysis as it simulates unrealistic minimum temperature extremes over the historical period. There is more spread between models for precipitation than temperature extremes but in most cases the observations sit within the model spread. Exceptions are consecutive wet days (CWD) where nearly all models overestimate the actual number of annual wet days and simple daily intensity (SDII) and one day precipitation maxima (Rx1day) where the models tend to underestimate precipitation intensity. However, some of these differences likely lie in observational uncertainty. Most models including the multi-model mean indicate that precipitation intensity has increased over the last century but the two observational datasets analysed disagree on the sign of change of precipitation intensity, one of them indicating a significant decrease. We use the CMIP5 simulations for two future Representative Concentration Pathway (RCP) scenarios (RCP4.5 and RCP8.5) to project changes in temperature and precipitation extremes across Australia. By the end of the century the number of cold temperature extremes substantially reduces and the number of warm temperature extremes substantially increases; changes scaling relative to the strength of emissions scenario. Changes in temperature extremes are often greatest in the tropics. While the results for precipitation extremes are less marked, simulations for the end of the century compared to present day indicate more periods of dryness while the most intense precipitation extremes increase substantially, with a separation becoming clear between emissions scenarios.