Hottest Days Landing in Wrong Months Across Half the Planet, Study Warns

Phoenix baked through 113 straight days above 100 degrees Fahrenheit in the summer of 2024. That alone might not shock anyone who has spent a July in Arizona. What came next did: 21 consecutive days of record-breaking high temperatures in October, a month most residents associate with cooling off, not sheltering from the sun. The city had to keep its cooling centers open weeks longer than usual.

That stretch of unexpected fall heat wasn’t a fluke confined to the desert Southwest. A new study published in AGU Advances finds that extreme heat seasons have expanded across roughly half of the world’s land area over the last 45 years, and they’re expanding lopsidedly. Some regions are seeing dangerous heat creep in earlier than expected, in the spring. Others, like the western United States, are seeing it push later into the fall. Scientists behind the research say this uneven spread cannot simply be chalked up to the planet getting warmer overall, which means something else is driving heat into new, unprepared corners of the calendar.

Those timing shifts are already visible in public health records. Maricopa County, home to Phoenix, reported that only 46% of its 608 heat-related deaths in 2024 happened in July, historically the hottest month, compared to 64% the year before. That’s a meaningful shift in when people are dying from heat, and it lines up with the idea that danger is spreading beyond the months when people expect it and know how to protect themselves.

How Scientists Tracked Extreme Heat Across Decades

Researchers from NASA’s Goddard Institute for Space Studies, Columbia University’s Lamont-Doherty Earth Observatory, and New York University set out to answer a basic but oddly unstudied question: when exactly does extreme heat happen around the world, and is that timing changing?

Past research had looked at how the broader warm season is lengthening, but that’s a blunter measure built around average temperatures across whole months. This team wanted to zero in on true extremes, the hottest and most oppressive days on record, and track exactly when they occur.

To do that, they pulled decades of global weather data covering temperature, humidity, wind, and other conditions going back to 1980. They looked at two different kinds of dangerous heat: dry heat, measured by simple air temperature, and humid heat, measured by a combined score of heat and moisture that shows how hard it is for the body to cool itself by sweating. Dry heat tends to hurt crops and ecosystems more, while humid heat is more dangerous for human health because it blocks the body’s natural cooling system.

For each location on Earth, the team identified the hottest days on record, the top 5%, during the 1980s, establishing a fixed baseline that later decades would be measured against. Then they found the three-month stretch when those extreme days piled up most often, and called that the local “extreme heat season.” This three-month window often didn’t match up with calendar summer, even in tropical regions where temperatures barely change year-round. They also marked two-month “shoulder seasons” right before and after that peak window, when early warning signs of an expanding heat season first show up.

From there, the researchers compared how often extreme heat showed up in those shoulder seasons during the 1980s versus the most recent decade, 2015 through 2024. Phoenix served as a detailed case study: in the 1980s, the city’s driest, hottest days were almost entirely confined to June through August, with zero extreme dry-heat days landing in the post-season. In the past decade, extreme dry heat pushed noticeably into that once-quiet stretch after the traditional season ended.

A Global Pattern in Extreme Heat, and a Puzzle

Zooming out to the whole planet, the pattern held. Extreme dry heat seasons expanded over about 50% of global land area, and extreme humid heat seasons expanded over about 48%, using the same 45-year comparison. But this growth wasn’t balanced. Places including the western United States, eastern China, northern Africa, and eastern Europe saw a bigger jump in extreme heat events happening after their traditional heat season ended. Meanwhile, western Europe, southern Africa, and northwestern India saw the opposite: a bigger increase in extreme heat creeping in before the season typically began.

That’s where the study takes its most surprising turn. It’s reasonable to assume this lopsided spread is just what happens when the whole planet warms up: months that were already warmer on average would naturally cross the extreme heat threshold more often, while cooler months would lag behind. The research team tested that idea by building a simplified model that asked what would happen if the only thing that changed since the 1980s was a higher average temperature, with the same old seasonal patterns simply shifted upward.

That simple explanation fell short. For roughly 80% of the world’s land area, rising average temperatures alone could not explain the lopsided way heat is spreading into the shoulder seasons. That simple explanation fell short. For roughly 80% of the world’s land area, rising average temperatures alone could not explain the lopsided way heat is spreading into the shoulder seasons. In many regions, the mismatch even ran backward from what warming would predict. Higher average temperatures should have nudged more heat into the spring in the southwestern United States, northern Canada, and eastern Europe, yet those areas instead gained the most heat in the fall. Western Australia and northern Russia did the opposite. Something region-specific, not just a warming planet, is steering where and when these extremes are landing. Something region-specific, not just a warming planet, is steering where and when these extremes are landing.

Why the Timing of Extreme Heat Matters as Much as the Heat

Dry and humid heat also aren’t moving in lockstep. In Phoenix, the study found the dry heat season pushing later while the humid heat season shifted earlier, meaning a resident trying to plan around “typical” hot weather now faces two moving targets rather than one predictable window.

That distinction carries real consequences. Heat arriving earlier can catch people before they’ve acclimated or prepared cooling strategies, like tuning up the air conditioning or knowing where the nearest cooling center is. Heat arriving after the expected season piles onto a body and a community already worn down by a long summer. And when an expanding heat season starts overlapping with other seasonal hazards, like wildfire season in the western United States or hurricane season in the Southeast, the risks compound in ways that are harder to manage than either hazard alone.

Heat has always been dangerous. This research makes clear that the danger no longer stays within the calendar boundaries people have learned to expect, and simply pointing to a warming planet doesn’t tell the whole story. Until scientists pin down the local forces pushing extreme heat into new seasons in each region, communities from Phoenix to eastern China are left managing a moving target with an outdated map.