With extreme heat now a public health crisis, local data can save lives

Eric Mackres is senior manager of urban analytics for the WRI Ross Center for Sustainable Cities and attended London Climate Action Week during the June 2026 heatwave. Usama Bilal is an associate professor of epidemiology and co-director of the Urban Health Collaborative at Drexel University. As thousands gathered in London for one of the year’s largest climate gatherings last week, Western Europe faced its most severe heatwave ever recorded. The irony was not lost. Across Europe, over a dozen countries issued urgent heat warnings and Spain registered significant deaths. In London, where air conditioning is rare in buildings and on trains and buses, temperatures soared past 36 degrees Celsius (97F) and schools closed early. The mayor announced the city’s first heat action plan – an important step. Extreme heat is now a public health crisis for many of the world’s cities, as the urban heat island effect intensifies dangerous temperatures – and it’s growing worse. Around 500,000 people die from extreme heat every year. As global temperatures rise, and with a severe El Niño getting underway, even more people will die and be hospitalised unless cities act soon. But most cities are still taking a far too one-sized-fits-all approach to tackling heat, looking only at temperatures and not its local effects on people and their health. People experience heat differently How extreme heat affects people’s health can vary widely across a country and city, depending on their environment and demographics. Cities can save far more lives and prevent more hospitalisations by taking a tailored approach, using data to understand who’s most vulnerable and directing solutions toward them. The good news: better data now exists that enable cities to pinpoint who’s most at risk. And that data can inform customised adaptation strategies to save lives. Indeed, the future of cities will hinge on their ability to deliver solutions to extreme heat tailored to at-risk people and neigh