When a heat wave is coming, who is most at risk?
As heat waves become more dangerous, public health officials need to know where the threat is greatest.
One street may have trees, shade, and a clinic nearby. A few streets away, pavement may hold heat late into the night, and residents may be older, poorer, farther from medical care, or more likely to work outdoors.
Vast heat, health, and demographic datasets already exist, collected from satellites, sensors, forecasts, and censuses. But for many public health teams and heat officers, those data are too large, scattered, and technical to turn into information they can use in their emergency response.
A new tool from CrisisReady, a Harvard-based research initiative focused on data and disaster response, does that translation. The Heat Risk Data API – an application programming interface that allows one computer system to request and automatically receive information from another – lets users define the area they care about, such as neighborhoods, census tracts, districts, or counties. They can then pull processed, localized heat-risk data into their own research, dashboards, websites, or planning systems.
“Advances in AI allow us to leverage large datasets and protect people in public health emergencies. And yet those communities most at risk do not have access to the technology to take advantage of these data,” said Satchit Balsari, associate professor of emergency medicine at Harvard Medical School and co-director of Crisis Ready.
During a heat wave complicated by wildfire smoke, officials may need to know what happens next, and where, Balsari added. “It advances neighborhood-scale understanding of risk and preparedness, moving away from national or global averages. If I’m a local clinician, what kind of surge should I expect in my health system? What are asthma patients in my neighborhood likely to experience?”
More than a weather map
CrisisReady draws on public data sources that include forecasts, land-surface temperature, population and age estimates, tree canopy, land cover, health facilities, nighttime lights (as a proxy for electricity access), and air quality.
Together, said CrisisReady co-director Andrew Schroeder, “the system gives you a curated set of indicators that put risks in context.”
Users can view different heat measures, including air temperature, heat index, wet-bulb globe temperature and the Universal Thermal Climate Index (UTCI) – a measure of how the human body responds to the environment. There is no single heat-risk number that works everywhere: Humidity, asphalt, nighttime cooling, electricity access, shade and medical care all shape what danger looks like in different places.
Where local data are scarce
Schroeder pointed to Kano state in Nigeria, where CrisisReady has built a demonstration dashboard for a place with severe, under-addressed heat risk: “Kano is in real need of attention to mitigate heat effects, and the integration of these data contributes toward the thinking and planning.”
“There’s no geographic limit, there’s no spatial limit, there’s no kind of temporal limit to where this can be applied. From the Global North throughout the Global South, it removes a whole range of barriers to effective heat and health analysis at whatever unit is desired,” said Nishant Kishore, who built the tool.
Knowing what the map can’t show
The tool also shows where the data are strong and where data should be handled carefully because measurements may be incomplete or changing. Satellite observations may be missing because of clouds. Some places have richer demographic or health data than others. The API includes information about coverage and resolution, a warning against treating every number on a map as equally precise.
The Heat Risk Data API will not plant trees or check on vulnerable residents. But it can help health officials decide where a cooling center, a warning, a door-knock, or other resources may matter most.
-As told to David Trilling
