Dangerous heat is spreading into spring and fall

Clinical takeaway: Expect heat-related presentations beyond the usual season, particularly in late spring and  early fall. 

Clinicians often brace for heat-related illnesses in July. Cooling centers open and heat alerts also typically run on a summer timeline. But that calendar is becoming less reliable. Extreme heat is arriving in months when patients have not previously acclimated to high temperatures, or when they have already endured a full hot season. 

Research on warming seasons has concentrated on average temperatures, documenting summers that start earlier and run longer. But extreme heat events, the days that drive emergency visits and deaths, do not necessarily follow the averages. A new 45-year analysis of worldwide climate records has mapped extreme heat days, and finds the extreme heat season is not just growing but growing unevenly. 

In the US, the two shoulder seasons are associated with different heat types. Dry heat extremes shifted significantly later in the year across the Southwest and Southeast, pushing into fall, while humid heat extremes along the Gulf Coast moved earlier, expanding into spring. Humid heat carries the greater risk to patients because it limits sweat-based cooling.

Globally, extreme dry heat seasons grew significantly across 50% of the world's land area between the 1980s and the decade ending in 2024, and extreme humid heat seasons across 48%. The expansion was lopsided rather than symmetric. The western US, eastern China, northern Africa, and eastern Europe saw extreme heat extend into the months after the traditional season, while western Europe, southern Africa, and northwestern India saw it arrive before. 

Phoenix, for example, recorded 183 extreme dry heat days in the 1980s and 338 in the decade through 2024. The share falling after the traditional season's end rose from zero to 6% of the annual total, and the median date of dry heat extremes moved 10 days later in the year. Rising average temperatures alone do not account for the pattern: in roughly 80% of land areas, the observed asymmetry differed significantly from what an intensification of existing seasonality would predict, pointing to regional drivers such as circulation, land use, and moisture trends that the analysis was not designed to isolate. 

Researchers defined local extreme heat seasons as the three months holding the highest share of days in the hottest 5% of a 1980–1989 baseline, calculated separately for dry heat and for wet bulb globe temperature. They then compared each region's baseline distribution against 2015–2024 using the MERRA-2 global reanalysis. Findings were validated against the European ERA5 dataset, with broad agreement between the two. 

The researchers' next test is whether climate models reproduce the observed asymmetry; if simulations agree with the observational record, that strengthens the case that the pattern is a durable feature. Regional analyses of circulation patterns, soil moisture, land use are also queued. For US clinicians and health systems, the nearer-term implication is operational. The findings argue for extending heat surveillance and alert infrastructure past the historical season's end, particularly across the West and Southeast, where extreme heat increasingly overlaps wildfire and hurricane seasons. 

"There are very clear asymmetries in how extreme heat seasons are expanding in different parts of the world," says Catherine Ivanovich, lead author and climatologist at NASA Goddard Institute for Space Studies. "Extreme heat is starting to become something different in a lot of these regions." 

Source: Ivanovich CC, et al. (2026 Sep 10) AGU Adv. Extreme Dry and Humid Heat Seasons Are Changing Asymmetrically