Extreme fire weather is increasingly hitting many regions at the same time

Two papers published a day apart this week approach the wildfire problem from opposite ends. One asks how often dangerous fire weather now shows up in several places at once. The other asks what the resulting smoke costs in lives, and what avoiding it is worth. Read together, they describe a health hazard whose defining feature is no longer local.

Fire weather is becoming synchronised

The first paper, in Science Advances, examines what the authors call synchronous fire weather (SFW) — periods when conditions favourable to large fires occur simultaneously, either within a region or across regions [s1]. The distinction matters operationally: fire suppression depends on moving crews, aircraft and equipment between areas that are not burning and areas that are. When fire weather synchronises, that assumption breaks [s1].

Over the period 1979 to 2024, the analysis found significant increases in synchronous fire weather, with more than a twofold increase in most regions [s1]. The authors estimate that over half of the observed increase is attributable to anthropogenic climate change [s1].

Baseline synchronicity is not evenly distributed. The paper reports climatologically elevated intraregional synchronicity in boreal regions, and interregional synchronicity linking northern temperate and boreal regions — meaning that fire seasons in those parts of the world are more likely than average to line up with each other [s1].

Natural climate variability still exerts a large influence on top of the trend. Equatorial Asia experiences 43 additional intraregional synchronous fire weather days during El Niño years [s1] — a reminder that in any given year, the mode of variability may matter more than the trend does.

The health link in this paper is indirect but explicit: synchronous fire weather is strongly correlated with regional fire-sourced PM2.5 in multiple regions globally [s1]. The paper does not estimate deaths. It establishes that the conditions producing smoke are increasingly arriving everywhere at once.

Pricing the smoke

The second paper, in PNAS, takes up the health accounting the first one leaves open, for the United States [s2]. Climate-induced wildfire smoke, the authors note, is rarely included in estimates of the societal costs of climate change, despite its damage potential [s2].

Their framework estimates PM2.5 from climate-induced wildfire smoke and the associated mortality across emissions trajectories and global mean surface temperature levels, and it accounts for fire–vegetation feedbacks by empirically estimating how past fires affect future burn probability [s2]. That last piece is a real methodological difference from simpler approaches, because a landscape that has recently burned is not equally available to burn again.

Under 3 °C of global warming relative to 1850–1900, the paper estimates smoke exposure will lead to 64,000 deaths annually in the United States (95% CI 33,500 to 112,300), calculated using historical population — a 60% increase above the estimated annual smoke deaths during 2011–2020 [s2].

Limiting warming to 2 °C reduces smoke-related mortality by 14%, or 8,900 deaths per year, relative to the 3 °C estimate [s2].

Converted into the currency policy analysts actually use, the authors calculate a partial social cost of carbon of $11.2 per additional tonne of CO2 emitted in 2025 (95% CI $1.1 to $41.6, in 2020 dollars) attributable to US wildfire smoke mortality alone [s2]. Folding that into existing non-wildfire damage estimates raises the US domestic social cost of carbon by 74% [s2].

What the numbers can and cannot carry

Both papers are modelling exercises, and their uncertainty ranges are wide enough to matter. The PNAS mortality interval at 3 °C spans roughly 33,500 to 112,300 deaths a year — more than a threefold range [s2]. The social cost figure spans $1.1 to $41.6 per tonne [s2]. A reader who takes the central estimate and discards the interval has taken the least informative part of the result.

The mortality estimates are also calculated using historical population, so they isolate the climate signal rather than projecting how many people will actually be living in smoke-exposed areas [s2]. Population growth in the western United States would push the count one way; changes in housing, filtration and behaviour could push it the other.

The attribution claim in the Science Advances paper — over half of the increase in synchronous fire weather attributable to human-caused climate change [s1] — is a statement about fire weather, the meteorological conditions, not about fires themselves. Ignition, fuel management and land use sit between the two.

Why the pairing is the story

Individually, neither result is unexpected. What they say jointly is more specific: the geography of wildfire smoke exposure is broadening at the same time that its health cost is being formally priced into climate policy analysis.

Synchronised fire weather complicates the mutual-aid arrangements that fire agencies rely on and degrades air quality over wide areas simultaneously [s1]. That is a different problem from any single severe fire season, and it is not one that better local preparedness solves on its own.

For readers, the practical implication is narrow: smoke episodes in one region increasingly coincide with smoke episodes elsewhere, which is a reason to treat air-quality forecasting as a routine rather than exceptional part of a fire season. Neither paper offers guidance on individual protective measures, and neither claims to.

Sources

• [s1] Increasing synchronicity of global extreme fire weather. Science Advances, 18 February 2026. https://doi.org/10.1126/sciadv.adx8813
• [s2] Valuing wildfire smoke-related mortality benefits from climate mitigation. Proceedings of the National Academy of Sciences, 19 February 2026. https://doi.org/10.1073/pnas.2533772123