Background

Despite ongoing efforts to reduce air pollution, the complex relationship between air pollution and climate change presents additional multifaceted challenges. The spatiotemporal co-occurrence of extreme temperatures and air pollution episodes remains understudied. Furthermore, current studies typically employ uniform temperature thresholds across broad areas, overlooking regional differences in health vulnerability. We aim to present a comprehensive assessment of extreme temperatures and air pollution, and to incorporate location-specific risk thresholds.

Methods

In this time series modelling study we analysed the changes in extreme-risk temperature (ERT) days for heat and cold in Europe from Jan 1, 2003 to Dec 31, 2020, considering time-varying temperature–mortality relationships based on regional mortality data (from Eurostat) from 35 European countries (543 million people). We used daily estimates of PM2·5, PM10, NO2, and O3 concentrations from quantile machine learning estimations at 0·1-degree to identify of heat-compound and cold-compound days co-occurring with air pollution levels exceeding WHO guidelines.

Findings

Cold-related mortality risk decreased over the study period across Europe, and adaptation to heat-related mortality was less pronounced. Between 2003 and 2020, annual cold-ERT days had decreased by 20·7 days per decade and annual heat-ERT days increased by 2·8 days per decade. Southeastern Europe had higher frequencies of both heat-ERT and cold-ERT days. Heat-O3 events were the only increasing heat-compound episodes, with 2·6 more days per decade. Conversely, cold-compound episodes decreased by 15·2 days per decade with cold-PM2·5 events remaining the predominant threat. Around 349 million Europeans were exposed to at least 1 cold-compound day annually, and around 295 million experienced at least one heat-compound day, including nearly 235 million affected by heat-O3 compound episodes.

Interpretation

By identifying region-specific risk thresholds, our study reveals spatial disparities and changes in ERT events, particularly when coupled with air pollution. These findings are essential for developing targeted adaptation strategies, facilitating subsequent health assessments, and implementing effective measures to safeguard public health.

Funding

European Research Council and Ministerio de Ciencia e Innovación