The temperature difference between a built-up urban core and the surrounding countryside, measured at the same time on the same day, can be five to ten degrees Celsius higher in the city. The effect is well documented, has been understood for over a century, and is intensifying as the planet warms. The interaction between background warming and the urban heat-island effect is multiplicative, not additive: a heat wave in a city is a more severe event than the same atmospheric heat wave in the country.
The physical causes
The mechanisms are straightforward. Asphalt and dark rooftop surfaces absorb solar radiation more efficiently than vegetation and re-radiate it as heat. The geometry of urban “canyons” (streets bounded by tall buildings) traps the re-radiated heat. The reduction of evapotranspiration relative to vegetated land removes a significant cooling mechanism. Anthropogenic heat from air conditioners, vehicles, and industry adds a continuous local source.
The effect is most pronounced at night. Daytime urban-rural temperature differentials are sometimes modest; nighttime differentials are typically much larger because the heat absorbed during the day continues to radiate after sundown, suppressing the overnight cooling that human and animal bodies rely on for thermal recovery.
The mortality
Heat is now the deadliest weather-related public-health hazard in most temperate-zone cities. The 2003 European heat wave killed an estimated 70,000 people, predominantly elderly residents of cities. The 2021 Pacific Northwest dome killed at least several hundred in cities that had not historically required residential air conditioning. The 2022 European heat wave killed an estimated 60,000.
The mortality is concentrated in identifiable populations: people over 65, people with cardiovascular or respiratory conditions, people in low-income housing without effective cooling, and people in the most heat-impervious neighborhoods. These overlap. The geography of urban heat-related mortality is, in most cities, also the geography of historical disinvestment in tree cover, parks, and building-envelope quality.
What works
The interventions are well understood, modestly priced relative to the mortality they prevent, and politically slow to deploy. Tree canopy expansion, particularly in the warmest neighborhoods, produces measurable cooling. High-albedo (“cool”) roofs and pavements lower local temperatures meaningfully. Building-code improvements that require passive cooling features and adequate insulation reduce both mortality and the energy load of mechanical cooling.
Cooling-center networks, neighborhood social-contact interventions for elderly residents, and emergency communication systems reduce mortality during acute events. Each of these is a public-health investment whose returns accrue largely to people who are not the demographic most likely to vote in municipal elections.