Global drivers of minimum mortality temperatures in cities

• Published in: The Science of the total environment 2019-12, Vol.695, p.133560-133560, Article 133560
• Main Authors: Krummenauer, Linda, Prahl, Boris F., Costa, Luís, Holsten, Anne, Walther, Carsten, Kropp, Jürgen P.
• Format: Article
• Language: English
• Subjects: City population; Climate; Minimum mortality temperature; Sigmoid model; Socio-economy; Topography
• ISSN: 0048-9697; 1879-1026
• DOI: 10.1016/j.scitotenv.2019.07.366

Human mortality shows a pronounced temperature dependence. The minimum mortality temperature (MMT) as a characteristic point of the temperature-mortality relationship is influenced by many factors. As MMT estimates are based on case studies, they are sporadic, limited to data-rich regions, and their drivers have not yet been clearly identified across case studies. This impedes the elaboration of spatially comprehensive impact studies on heat-related mortality and hampers the temporal transfer required to assess climate change impacts. Using 400 MMTs from cities, we systematically establish a generalised model that is able to estimate MMTs (in daily apparent temperature) for cities, based on a set of climatic, topographic and socio-economic drivers. A sigmoid model prevailed against alternative model setups due to having the lowest Akaike Information Criterion (AICc) and the smallest RMSE. We find the long-term climate, the elevation, and the socio-economy to be relevant drivers of our MMT sample within the non-linear parametric regression model. A first model application estimated MMTs for 599 European cities (>100 000 inhabitants) and reveals a pronounced decrease in MMTs (27.8–16 °C) from southern to northern cities. Disruptions of this pattern across regions of similar mean temperatures can be explained by socio-economic standards as noted for central eastern Europe. Our alternative method allows to approximate MMTs independently from the availability of daily mortality records. For the first time, a quantification of climatic and non-climatic MMT drivers has been achieved, which allows to consider changes in socio-economic conditions and climate. This work contributes to the comparability among MMTs beyond location-specific and regional limits and, hence, towards a spatially comprehensive impact assessment for heat-related mortality. [Display omitted] •Limited global coverage of temperature-mortality relationships at the city scale•Creation of global model for minimum mortality temperatures of urban populations•Comparison of linear, sigmoid and segmented models across all variable combinations•Climate, topography and socio-economy drive urban minimum mortality temperatures.•Model accounts for physiological constraints and incomplete adaptive capacity.