The representation of health-relevant heatwave characteristics in a Regional Climate Model ensemble.

Heatwaves have been linked to increased rates of human mortality and morbidity. Due to these adverse health impacts, it is essential to understand how these extreme events might change in the future. Health impacts studies and adaptation planning benefit from having this information at a local scale. While Global Climate Models can provide continental-scale projections of future changes in these events, Regional Climate Models, such as those used in the New South Wales/Australian Capital Territory Regional Climate Modelling (NARCliM) project, provide simulations at a finer scale more appropriate for regional assessments. This paper uses the NARCliM simulations for New South Wales and the Australian Capital Territory to investigate the ability of a Regional Climate Model ensemble to represent heatwave characteristics through the Excess Heat Factor, an index believed to be relevant to human health. Both uncorrected and bias-corrected model output was evaluated against observationally-derived heatwave characteristics for the period 1990-2009. The effect of bias- correction on future changes in heatwave characteristics was also assessed. Overall, while the simulations were able to provide a good representation of the recent climate, bias-corrected simulations did not greatly change model output. Some regions were more affected than others, with bias-correction being most beneficial for coastal regions. We emphasise that these results may not apply to all indices measuring extreme heat and we demonstrate that results for a fixed-threshold index are substantially affected when bias-correction is applied. While supporting bias-correction, this study demonstrates that bias-corrected climate model output is not necessarily required when evaluating a relative measure such as the Excess Heat Factor.