CMIP6 MultiModel Evaluation of Present-Day Heatwave Attributes

Heatwave frequency, duration and intensity has increased since the 1950s, a trend likely to continue with anthropogenic climate change. Although climate model projections are an important source of information on how heatwaves and other climate extremes may evolve with climate change it is necessary to understand whether climate models are able to adequately reproduce historical statistics of these events. We provide the first global evaluation of CMIP6 models to characterize heatwave characteristics between 1950 and 2014. Further, we also include a global heatwave evaluation of the CMIP5 models, as no study currently exists. Our results demonstrate that both ensembles underestimate mean heatwave frequency with region-dependent biases for mean heatwave duration, intensity, and cumulative heat. Comparisons between CMIP5 and CMIP6 show that improvements in skill for the heatwave metrics evaluated here are marginal, suggesting that future improvements in simulating heatwaves may only be possible with significant advances in climate modeling capabilities. Plain Language Summary Heatwaves are increasing in response to climate change associated with human activity, a trend that is likely to continue. In order to build confidence in projected changes in heatwaves it is necessary to first understand whether the models producing these projections can represent historical heatwave characteristics. We present the first global evaluation of the current and previous suite of climate models that inform the Intergovernmental Panel on Climate Change Assessment Reports. Our analysis considers four different heatwave characteristics that measure the frequency, duration, average intensity, and cumulative intensity over the period 1950-2014. Our results show that most climate models underestimate historical heatwave frequency with regional under- and overestimation of heatwave duration, average intensity and cumulative intensity. Our analyses are useful for decision makers and other users of climate model projections as they show which regions and heatwave characteristics have higher skill over the present-day period.