On July 15, temperatures along the East Coast were close to 100 degrees Fahrenheit, with the heat index in New York City (a measure of what the air temperature feels like to the human body when actual air temperature and relative humidity are combined) reached roughly 104 degrees. The city had already extended its heat emergency operations, keeping hundreds of cooling centers open into Thursday, July 16. At almost the same time, the smoke from more than 830 wildfires burning in Canada descended on the region, blanketing several states and forcing New York State to issue an Air Quality Health Advisory covering all its regions.
By the following day, the Air Quality Index was forecast to reach around 200—the city’s worst smoke event since June 2023, when the sky turned orange and the index peaked near 480. At the end of the same week, New York City Emergency Management activated its Flash Flood Emergency Plan as the National Weather Service issued a Flood Watch running from Saturday morning into Sunday, with isolated locations expected to receive two to four inches of rain.
Compounding effects can be very powerful. The damage from one event can be catalyzed or amplified by others, and the combination of the overall effects is larger than the sum of its parts. We see this with the Greenland Ice Sheet, where surface melt, albedo and firn dynamics interact in ways no single measurement can capture. Similarly, a person standing outdoors in West Harlem that mid-July week did not experience “a heat event” and then, separately, “an air-quality event.” These events happened all at once, multiplying their negative effects.
Extreme heat forces the cardiovascular system to work harder to shed warmth. Fine particulate matter from wildfire smoke (PM2.5) simultaneously inflames the lungs and stresses the heart, aggravating those same cardiac and respiratory conditions that heat already burdens. Officials told New Yorkers to stay indoors in air-conditioned spaces to escape both the heat and the smoke, while also urging them to wear an N95 mask if they had to go out. But air conditioning demands electricity, and demand peaks precisely when the grid is most fragile. The very act of protecting against two hazards concentrates stress on the city’s infrastructure and creates a failure risk that could have devastating consequences. And flooding adds an extra layer. When streets flood, the polluted waters are not only a safety challenge, but they can also have negative effects on people’s health.
Climate scientists call this a compound event: the combination of drivers and hazards that together produce a societal or environmental impact larger than any of them would alone. Traditional risk assessment typically considers one driver or hazard at a time and, in doing so, can significantly underestimate the actual risk by missing interactions among different hazards.
Let us now think back to what happened the week of July 15.
Our heat advisory system evaluates heat. Our air-quality system evaluates particulates. Our flood system evaluates rainfall and drainage. Each is a genuine achievement of public science. And each, by design, considers one hazard at a time. We have built a warning architecture that is underestimating the risk by default.
The research community is aware of this issue and has been working to address it. Over the past decade and a half, compound-event research has grown into a substantial field, motivated by the failure of conventional assessments to account for compounding drivers. Because of this, we now have global maps identifying hotspots where multiple hazards co-occur—and North America is among them.
The science is not the bottleneck. The bottleneck is that the science is not being translated into the operational tools that cities and states can actually use to manage emergencies, allocate resources and protect people.
Right now, a New Yorker facing a heat index of 104, an AQI of 200 and a flood watch has to read three separate warnings and is left to integrate them personally, in real time, with no guidance on how a “moderate” reading of one hazard interacts with a “severe” reading of another. There is no accepted metric that says: “together, this is worse than any single number suggests.”
We also need to remember that the hazards do not fall evenly. Those neighborhoods with the least tree canopy and the most heat-retaining surfaces are frequently the same ones with the oldest housing stock, the least air conditioning, the highest rates of the respiratory and cardiovascular conditions that smoke aggravates, the aging drainage most likely to fail in a cloudburst, and are home to a highly socially and economically vulnerable population. The compounding is not only meteorological. It is social.
That week did not deliver its hazards simultaneously but rather in a rolling series—heat, then smoke thickening over the heat, then storms arriving as the smoke lingered. The physiological and psychological debt of a brutal Wednesday is still being carried into a stormy Saturday. Our systems treat each day’s alert as if it begins from zero. Bodies do not.
As a reminder: a warmer atmosphere holds more moisture and delivers it in more intense bursts; a fire regime in the boreal forests to our north is lengthening and intensifying; and heat that arrives earlier stays longer and reaches higher. These are expressions of climate change, and the probability that they overlap is trending upward faster.
This is where adaptation policy has to catch up with climate science. We plan, budget and warn in silos because that is how the hazards were historically studied and how agencies were historically organized. When the same week can require a person to stay inside for the heat, seal the windows against the smoke and prepare to evacuate a flooding basement, the siloed plan is not just incomplete; it can issue contradictory instructions to the same person on the same afternoon.
The work ahead means building indices that integrate across hazards, coupling them to the social data that determine who actually gets hurt, and giving emergency managers a way to see the compound picture as clearly as they now see the single-hazard one. It means the climate science of compound events finding its way into a firehouse and a mayor’s dashboard.
Marco Tedesco is a research professor at the Lamont-Doherty Earth Observatory, which is part of the Columbia Climate School, and an adjunct scientist at the NASA Goddard Institute for Space Studies.
David Sathuluri is founder and co-director of Lamont’s Laboratory of AI, Climate and Society.
Views and opinions expressed here are those of the authors, and do not necessarily reflect the official position of the Columbia Climate School, Earth Institute or Columbia University.
