Impact of aerosol forcing on heat extreme event attribution results

Heatwaves represent some of the most impactful extreme weather events, with profound implications for ecosystems, human health, and economies globally. Accurately attributing changes in their occurrence probabilities, intensity, and duration is crucial for effective climate change adaptation strategies. The intensity, frequency, and duration of heatwaves have increased globally, yet their attribution is not straightforward (Oldenborgh et al., 2022) because diverse factors influence regional heatwave trends.  A common practice for calculating heatwave return periods relies on extreme value statistics, where the Generalized Extreme Value distribution (GEV) shifts linearly with a covariate on global mean temperature (GMT) for the location parameter (the “standard method”). This approach is widely used in rapid event attribution studies. However, local temperature trends in recent decades have been influenced by anthropogenic aerosol emissions (AER), depending on the region. AER have a predominantly cooling effect on heatwaves via reflecting incoming solar radiation. AER trends can therefore counteract or amplify the warming effect of greenhouse gases (GHG), depending on emission trends. These trends may affect  regional climate dynamics and thermodynamic processes and, consequently, the return periods of extreme temperature events. In this study, we use state-of-the-art large ensemble and single forcing large ensemble climate model simulations from the Community Earth System Model 2 (CESM2). To examine the impact of AER on extreme event trends, we added an additional covariate on local aerosol optical depth (AOD) for the location parameter. We then compared this approach with the standard method. Our results indicate a substantial bias in the standard method during periods of strong regional AER trends. This bias is most pronounced in major industrial regions, where regional AER trends show the strongest deviation from the GMT covariate. In contrast, in some regions, AER trends have little or no impact. Adding AOD as an additional covariate reduces these biases and improves the goodness of the GEV fit. In regions such as North America, Central and Eastern Europe, and China, the GEV fit improves significantly for nearly all individual ensemble members with the addition of the covariate on AOD. For example, in Central Europe and the Midwest US, the standard method overestimates extreme temperatures by more than 1°C in the 1970s and 1980s, whereas this bias disappears when AOD is added as a covariate. This study underscores the importance of incorporating regional aerosol trends into attribution studies to improve the estimation of return periods, and thus attribution statements.