A heat wave in a city isn't just one temperature — it's a thousand different temperatures, depending on where you stand. A tall building can block a cooling breeze, while a wide street in full sun can bake like an oven. Scientists have long known that cities trap heat, but they struggled to measure exactly how each neighborhood feels it. Now a new dataset called HUMID-Atlanta is changing that, and it could help save lives across the United States.
Heat stress is a leading cause of weather-related deaths, and cities are especially dangerous places for it because of something called the urban heat island effect. Buildings, roads and dense layouts make cities hotter than the surrounding countryside. But the number on a thermometer only tells part of the story. Wind and shade can make the same temperature feel cooler or hotter, which means scientists need more than just degrees to understand how residents truly suffer.
A team from the U.S. National Science Foundation National Center for Atmospheric Research (NSF NCAR), Emory University and the University of North Carolina at Chapel Hill set out to build that deeper picture, starting with the city of Atlanta. The result is HUMID-Atlanta: High-resolution Urban Meteorology for Impacts Dataset for Atlanta Metropolitan Region. The scientists started with an earlier dataset called HUMID, which covered 1980–2018, and then supercharged it by pairing it with a powerful weather computer model called WRF. That model can add details like building height and see how wind swerves around structures to create cooler pockets.
The result covers the years 2010–2023 and lets researchers cross-reference things like hospital records with detailed neighborhood weather data from the same period. That means they can pinpoint exactly what mix of weather and city design leads to dangerous spikes in heat stress — and where to focus help before people get sick.
HUMID-Atlanta is already being put to work on a serious health problem: acute kidney injury (AKI), a condition where the kidneys suddenly stop working well. Heat can trigger it, but previous studies treated all neighborhoods the same. Andrew Newman, an NSF NCAR senior scientist, is now leading a team using the dataset to see how heat waves harm kidneys differently in city centers, suburbs and rural areas. "Not only is acute kidney injury a serious health concern, but the economic burden of treating those with it is substantial," he said.
The dataset is free for anyone to use, and the team has already stretched it through 2025. They're talking with other big cities about building versions for them, with hopes of covering the entire U.S. and even the whole globe. "HUMID-Atlanta is just the first dataset of its kind," said lead author Tzu-Shun Lin. In the future, this method could forecast heat waves before they arrive and trigger early warnings on dangerous days. The more precisely we can feel a city's heat, the better we can protect the people who live in it.