The New Fire: Why Wildfire Behavior Has Outrun Our Models

Fire is not an intruder in most of the world's ecosystems. Chaparral, savanna, boreal forest, ponderosa pine, and eucalypt woodland all evolved with it; many species in these systems require fire to reproduce, and their structure reflects centuries of regular, mostly low-intensity burning. Some pine cones open only under the heat of a passing fire. Some seeds germinate in response to smoke chemistry. The problem in much of North America and Australia is not that fire has returned but that a century of aggressive suppression allowed fuel to accumulate at densities those systems never previously carried, so that when fire arrives it arrives as something the ecosystem has no evolutionary experience of. This is the fire deficit paradox: suppression did not prevent burning, it deferred and concentrated it.

Climate change then loaded that fuel with drier air. The key variable is vapor pressure deficit — essentially the atmosphere's thirst — which rises non-linearly with temperature and pulls moisture out of vegetation and soil. Warmer air also extends the fire season at both ends, shrinks the window for safe prescribed burning, and increases the frequency of the dry lightning that starts fires in remote terrain. The result is behavior that fire managers describe as outside historical experience: fires that do not lie down at night as humidity recovers, that spot embers kilometers ahead of the front, that generate their own weather, and that produce pyrocumulonimbus columns capable of injecting smoke into the stratosphere.

The ecological consequences run deeper than the burned area statistics suggest. When fires burn hot enough and often enough, forests do not regrow as forests — seed sources are destroyed, soils are sterilized, and the site converts to shrubland or grassland in a transition that may be effectively permanent under current climate conditions. In boreal and peatland systems the carbon accounting is worse still, because these fires burn into organic soils holding carbon accumulated over thousands of years, releasing stocks that no regrowing forest will recapture on a policy-relevant timescale. Canada's extraordinary 2023 season, which pushed smoke across the eastern seaboard of North America and into Europe, illustrated both the scale and the reach.

It matters here that fire is not uniformly destructive, and that conflating all fire with catastrophe has produced bad management for a century. Mixed-severity burns create the patchwork of ages and openings that many species depend on; black-backed woodpeckers specialize in freshly burned stands, and a long list of plants flowers only in the years following fire. What ecologists worry about is not fire but homogeneity — burns so large, so hot, and so complete that they eliminate the unburned refugia from which recolonization happens. A landscape that burns in a mosaic recovers. A landscape that burns uniformly at high severity does something else, and the difference is largely a function of what fuel structure we allowed to build up beforehand.

Where houses sit has become inseparable from all of this. Decades of expansion into the wildland–urban interface put homes in the most combustible terrain available, which converts fire management into asset protection and consumes the suppression budget defending structures rather than managing landscapes. Insurers have begun to respond in the most direct way possible — withdrawing from entire regions or repricing to levels that function as withdrawal — and that retreat is turning out to be a faster and blunter policy instrument than any zoning code. It is worth watching, because it is one of the first cases in which climate risk has been priced into a market at a scale ordinary households can feel.

Smoke is the part of this that touches the most people. Fine particulate matter from wildfire travels thousands of kilometers, and the health burden it imposes on populations far from any flame is now large enough to register in national mortality statistics. That gives fire policy an unusually broad constituency, and it points toward the same set of underused tools: prescribed burning, cultural burning practices maintained by Indigenous fire managers who were displaced by suppression regimes, thinning where it is ecologically appropriate, and land-use rules that stop putting new houses in the worst places. None of these are technically mysterious. All of them are politically slow.