Burning Cuts the Worst Wildfires by 92% — and their Smoke

✅ Key Takeaway: In Californian conifer forest, low-severity fire cut the most severe wildfire risk by 92% and saved more smoke later than its application produced.

Burn a patch of conifer forest at low intensity and the chance it later burns at the highest severity falls by 92% on the same ground, with the protection lasting up to a decade.

That is the headline result of a new study in Science, drawing on two decades of Californian satellite data, which reaches a more contested conclusion as well: the smoke avoided by such burning outweighs the smoke the burning itself produces.

One limit frames what follows. The study examines California conifer forest, not UK heather moorland or blanket bog. The vegetation, fuel structure and fire behaviour differ enough that the figures cannot be lifted across to British uplands. What does carry across is the trade-off at the heart of it, the same calculation now argued over in UK upland fire policy.

The researchers found that ground burned at low severity was far less likely to burn severely again. The chance of a very severe wildfire in the same location fell by 92% straight after a low-severity burn, and the reduced risk persisted for up to a decade. The benefit also spread outward, with lower fire risk detectable in unburned land as far as five kilometres from a low-severity burn.

How the study was done

Deliberate prescribed burning covers very little land in the western United States, so the team used patches of real wildfire that burned at low severity as a stand-in, having first checked the two were comparable in severity.

Working from satellite records of almost every Californian wildfire between 2000 and 2021, they matched each treated patch to similar untreated land on weather, vegetation and terrain, then tracked how often each burned over the years that followed.

In conifer forest the effect was strong and durable. The protection against the most severe fires lasted beyond a decade, while the reduction in overall fire risk faded within several years. The authors attribute the lasting effect to the removal of ground and ladder fuels, the low vegetation that lets flames climb into the canopy, a mechanism supported by earlier fieldwork in western US forests.

Other vegetation behaved differently. In shrubland, an early drop in fire risk faded quickly and produced no reliable reduction in later severe fires; the authors suggest fire exclusion may protect such systems better than burning them. No effect was detectable in hardwood or mixed woodland, where too little low-severity fire had occurred to measure one. The benefits were specific to forest with a particular fuel structure.

Counting the smoke

To weigh the trade-off, the team linked each fire's severity to the smoke it produced, measured as fine particulate pollution (PM2.5, the airborne particles most closely linked to harm to health). More severe fires produced proportionally more smoke. They then simulated treating Californian conifer forest with low-severity fire and set the smoke this would create against the wildfire smoke it would prevent.

The balance came out in favour of treatment. Burning would raise smoke at first, but the smoke avoided in later wildfires substantially outweighed it, with the net balance turning positive within a few years and the benefit growing over the following decade.

A sustained programme of treating conifer forest at scale, counting the benefit to neighbouring land, would deliver a meaningful cut in California's cumulative wildfire smoke.

What it means for upland management

The figures belong to Californian forest, but the principle is one upland managers will recognise. This is quantified evidence that reducing fuel load lowers the risk and severity of future fire, and that the smoke cost of burning now can be repaid, with interest, in wildfire smoke avoided later.

That trade-off is the core of the argument over managed burning in the British uplands, where the same question of whether to emit now or emit more later is rarely answered with numbers this concrete.

One caution matters most for moorland. The clear benefits here were in forest; the results in shrub-dominated systems were weak. UK heather moorland is a dwarf-shrub system rather than forest, and different again from Californian chaparral, so the headline figures should not be read across to it. The transferable lesson is that the trade-off exists, not how large it is.

The cost side is real and specific. The authors assume burns stay contained; escapes are rare, at around 2%, but not trivial, as the 2022 Calf Canyon-Hermits Peak fire in New Mexico showed when an escaped treatment burned more than 260,000 acres.

They also point to the south-eastern United States, where roughly 11 million acres are burned each year and prescribed fire has been linked to higher air pollution and a threefold rise in lung-disease burden in the most heavily treated areas.

The case for managed burning is not that it costs nothing. It is that, used in the right place and at the right scale, the cost of acting is smaller than the cost of waiting for the wildfire instead.