Quantifying heat-related mortality attributable to human-induced climate change

Abstract

The impacts of anthropogenic climate change remain largely unquantified. Here we detail and address limitations in existing methods for attributing health impacts to climate change, including the representation of the climate-health relationship, choices in calculating counterfactual temperatures, assessment of long-term trends and individual events, and estimation of the effects of adaptation. Applying these methods, we found over 1,700 deaths attributable to anthropogenic temperature increases in the Canton of Zürich (Switzerland) over 50 years. Changing exposures and vulnerabilities to heat, including due to adaptation, avoided over 700 deaths. Heat-related deaths peak during heatwaves but also occur throughout summer months and the fraction of deaths attributable to climate change is higher outside heatwaves. Our approach supports targeted adaptation measures, and the analyses described here could be adapted and applied elsewhere to assess the effect of climate change on other health impacts or economic losses. Teaser Synthesis of climatological and epidemiology methods finds 1,700 heat deaths attributable to climate change in the Canton of Zürich (1969-2018).

The footprint of human-induced climate change on heat-related deaths in the summer of 2022 in Switzerland

“…Additionally, to calculate the contribution of human-induced climate change we applied the method recently developed in a health attribution study [ 8 ], that builds upon an established approach to derive health impact projections [ 44 ]. The methodology applied here has been extensively described and discussed in Stuart-Smith et al [ 45 ]. Specifically, we adapted the approach used in Vicedo-Cabrera et al (2021) for trend attribution analysis to our study setting which considers a shorter period of time, similar to the event-attribution setting.…”

“…Thus, the use of daily average mortality in that case, instead of the observed, would be more advisable, but it might lead to an underestimation of the mortality burden when the method is applied to short periods (i.e. events) [ 45 ].…”

Designing and describing climate change impact attribution studies: a guide to common approaches

“…One of the most common approaches to impact attribution directly combines climate trend attribution and impact simulations in a single analysis. For example, a number of studies have examined how long-term temperature trends have led to long-term changes in heat-related mortality (Chapman et al, 2022;Stuart-Smith et al, 2023;Vicedo-Cabrera et al, 2021) and morbidity (Puvvula et al, 2022), as well as a number of issues related to child health, including preterm births (Zhang et al, 2022), low birth weight (Zhu et al, 2023), and childhood malaria (Carlson et al, 2023). Other studies have examined the contribution of human-caused climate change to global trends in poverty (Callahan & Mankin, 2022;Diffenbaugh & Burke, 2019) and food systems vulnerability (Dasgupta & Robinson, 2022;Ortiz-Bobea et al, 2021).…”

“…Trend-to-trend impact attribution studies may or may not use probabilistic framings, depending on their aims. For example, one recent study found two-to-one odds that long-term warming trends have had a positive effect on childhood malaria in Africa (Carlson et al, 2023); another study took what they called an "intensity-based" approach, focusing only on the cumulative number of excess heat-related deaths in Switzerland (Stuart-Smith et al, 2023). Both studies used an ensemble of climate models to simulate health impacts, and so generated a statistical distribution of simulated outcomes.…”

“…Evidence for adaptation is often analyzed by testing for a reduction in impact-climate relationships through time (Oudin Åström et al, 2013); researchers can then cross-project these relationships to estimate how adaptation contributed to attributable impacts. For example, one trend-to-trend attribution study estimated that adaptation to extreme heat prevented 738 deaths in Switzerland between 2004 and 2018; to do so, they estimated contemporary mortality based on the historical mortality-temperature relationship observed between 1986(Stuart-Smith et al, 2023. Other studies may directly account for adaptation in the impact model itself; for example, one recent event-to-event attribution study used hydrological models to test how changes in streamflow during the 'Day Zero' drought in South Africa would have been mediated by invasive alien tree clearing (Holden et al, 2022).…”