For how we live.
Researchers modelled 854 European cities under the climate that exists and the one that would have existed without human warming. The attributable share rose from 45% to 64%.
An analysis of global temperature data from 1950 to 2024 finds the share of the world's population facing dangerous heat-stress days has climbed sharply since the 1970s, summarized this month in JAMA.
An analysis of nearly 30 years of confirmed heat deaths finds many occurred well under standard alert thresholds, and that humidity-based measures may track risk more consistently.
A 43-year analysis at hourly resolution finds a 3.3% rise in average heat stress exposure duration across districts, with the change varying by time of day, season and region.
Researchers tested heat-mortality forecasts against the European summers of 2022 and 2023. The skill pattern is not the one a warning system designer would have guessed.
A 47-prefecture analysis of a decade of heatstroke mortality finds the alert level set at 33C misses almost all of the deaths a warning could plausibly prevent. About 31C would halve them.
A 22-year US analysis links hotter months to more overdose deaths. A British Columbia case-control study finds antipsychotic dispensing associated with dying in the 2021 heat dome.
A high-resolution model of East Africa's most densely settled basin finds that two-thirds of the projected rise in exposure comes from temperature and population growth acting together.
A 50-year attribution study separates two trends that usually get modelled as one: rising temperatures and a population that has become less likely to die in them.
A nine-year case time-series across every municipality finds both ends of the temperature range raise mortality risk on a Caribbean island, with the burden spread unevenly across towns.
The causes recorded do not look like a textbook heat event: nonspecific findings, nervous system disease and mental disorders led the excess, and deaths at home rose 64%.
A 55-expert consensus sets out how to attribute health outcomes to climate change. Its sharpest recommendation is against the fraction-of-attributable-risk method that many studies still use.