About five years ago, the infamous Thomas fire swept through California’s Santa Barbara and Ventura counties, burning more than 440 square miles and causing $2.2 billion in damages before finally being contained in January 2018. It was the largest wildfire in modern Californian history at the time — a title it lost just months later to another massive fire that broke out that summer. At least six other California wildfires now surpass it for sheer size.

Megafires are becoming ever more common in California, and the total area burned each year is rising fast. From 2000 to 2020, over five million hectares burned across the state — double that of the previous 20 years, says fire ecologist Jon Keeley of the US Geological Survey at Sequoia National Park, coauthor of a 2022 overview of fire ecology and evolution in the Annual Review of Ecology, Evolution, and Systematics.

The common stories told about California’s wildfire problem center on forest fires, where climate change and a buildup of burnable fuels from a century of intense fire suppression are the culprits. But there’s another story to tell too, says Keeley. Wind-driven fires like the Thomas fire that strike California’s chaparral landscape — shrub, grass and woodland common to the coast and southern region of the state — are also on the rise. For this type of fire, Keeley says, climate change and fuel aren’t primarily to blame — rather, a growing population and faulty power lines.

Big, wind-driven wildfires used to strike once every 30 to 130 years or so, says Keeley; now it’s every 10 to 15 years. And that’s proving lethal to several native plants. Without time to recover between fires, whole landscapes are being transformed and taken over by invasive species.

Keeley spoke with Knowable Magazine about these two distinct kinds of California wildfire, and why he thinks they need to be managed differently. This interview has been edited for length and clarity.

Your work emphasizes that there are two distinct types of extreme wildfire events in California. How are they different?

One type is those fires that strike mostly in the forests of the Sierra Nevada mountains. These are driven by anomalous accumulation of fuels — plant material and forest debris — that has built up because of more than a century of effective fire suppression. These fires are often ignited by lightning.

Then we have fires that are largely not in forests, but in the shrubland vegetation and oak woodlands of coastal central and Southern California. These chaparral landscapes tend to be juxtaposed with large metropolitan areas, and the fires are started by people. These fires are not driven by fuel accumulation; they are driven by extreme wind events that typically happen in the autumn and winter. They are called Santa Ana winds in Southern California, and North or Diablo winds in Northern California.

Once these fires are started, they’re really hard to extinguish. And as a result, even though we’ve had the same policy of suppression for these fires, we have not been effective in excluding fires in the chaparral like we have in forest.

Chart shows years, names, land area burned and cause of eight large California wildfires that occurred between 2012 and 2019.

A chart shows some large California wildfires and their causes. Some fires take hold in forests and are fueled by the buildup of dead plant material. Other fires ignite in chaparral, often sparked by power lines, and are spread by the wind.

Both types of fire are on the rise, but for different reasons. What has changed for the forest fires of the Sierras?

Historically, these fires would burn in the understory, and there was not enough fuel to carry that fire up into the canopy. We would have low-intensity surface fires.

But then, in the early 1900s, state and federal agencies embarked on a program of trying to eliminate fires from our forest. It was so effective that the area burned in the forested landscapes dropped to very low levels, far lower than historically was ever the case. Without fire, more and more dead material accumulated. That has resulted in fires now spreading from the surface up into the canopy and changing the fire regime from one of a low-intensity surface fire to a high-intensity crown fire.

Is climate change also playing a role?

In the forests, yes. We know that those forest fires are sensitive to temperatures in the spring and summer: The higher the temperatures, the more area burns. We have 100 years of data showing this pattern. As temperatures are increasing due to global warming, we should expect more area to burn in these forests.

What about for the wind-driven fires of the chaparral?

We have no evidence that climate change is playing a role for these fires. We did a very extensive study published in 2021 that examined 70 years of wind events and area burned. We looked at the temperatures and the precipitation levels and we found no relationship between climate and area burned during these autumn Santa Ana wind fires. Global warming is not a big factor because there are other things that override it in these coastal areas.

Why, then, are these wind-driven fires becoming more common?

These fires are 100 percent ignited by humans. We’ve had an increase of 6 million more people in California over the past 20 years. That translates into a great increase of the power grid, and more power lines means more opportunities for fire. Maintenance is probably the No. 1 issue. PG&E has been faced with a number of lawsuits related to failure to maintain these lines. The 2018 Camp fire, for example, was ignited by a power line that was wasn’t being properly maintained.

Can wildfire be good for the landscape?

Well, it depends. In the forests of the Sierra Nevada, frequent, low-intensity surface fires can be very valuable. The primary advantage of frequent fires in forest is it keeps the surface fuels down and prevents them from accumulating to the point where fire spreads from the surface to the canopy. Under these conditions, fire also creates openings in the forest, maybe 10 to 20 acres in size, and those are sites where certain species like ponderosa pine regenerate.

The difference in chaparral is, it’s a crown fire system, meaning that when fire burns through, it burns the entire canopy. The site requires several decades to recover to the point where it can withstand another fire. That period of time is something we’re working on now to try and determine: How short is too short in these systems? In general, it appears you need 20 to 30 years without a fire for that system to recover fully.

What happens to chaparral when wildfires are more frequent than that?

Chaparral has a range of different regeneration strategies for different species. Some species like oaks resprout from the base after a fire, and it almost doesn’t matter what the frequency of fires is — they seem to resprout very well. So, in those species, frequency of fire is not a threat.

But then we have another group that represents the majority of the diversity of woody vegetation: obligate seeding shrubs of Ceanothus, which is sometimes called California lilac or buckbrush, and Arctostaphylos, which is known as manzanita. The majority of species in those two genera depend 100 percent on seedlings coming up after the fire, and to do that, they need time to accumulate a seed bank.

Photograph of a hillside covered with shrubs, including some that have lilac-colored blossoms.

Chaparral shrubs, including manzanita and California lilac (shown here in bloom), need time to accumulate a seed bank between fires.


We’ve done a series of studies over the last two years, looking at regeneration in stands of different ages after they burn, and have found that in Ceanothus, for example, if the fires burn at less than 15-year intervals, there’s just not enough seed in the soil. And we get sometimes complete extirpation of those species from the site. For manzanitas, it’s a similar pattern. It often requires even longer periods, maybe 20 years or more, in order to accumulate sufficient seeds.

What are the implications of that?

When you wipe out these plants, you can wipe out a huge part of the landscape. For example, in Orange County, we have data from the 2020 Bond fire that over-burned the 2007 Santiago fire; it was a 13-year interval between those fires. And we showed that there was something like several thousands of skeletons of Ceanothus per hectare. They had no seedlings at all. They were wiped out from the site.

These systems do not fill in with native species; you end up opening up an ecological vacuum and non-native weedy species come in, and then you convert it from chaparral to non-native grasslands. That changes the water-holding capacity of the landscape. It increases the probability of erosion and flooding. And also, shrublands generally only have a fire season of about six months out of the year; grasslands have a 12-month fire season because these grasses are capable of drying out very rapidly. You can get a rainfall event and a week later, it can burn.

You just change a lot of things in the ecosystem that are not desirable. You eliminate a lot of native species: bird species, as well as lizards and rodents.

Learn more about megafires in our video.


What does all this mean for prescription burning: the practice of intentionally lighting controlled fires to burn an area. Is that a good or bad idea?

Prescription burning is a really important management tool in forests. I work in Sequoia National Park, and there we have clear evidence that prescription burning, which has been going on for 50 years, has been highly effective at preventing catastrophic fires. For example, the 2021 KNP Complex fire burned something like 80,000 acres, mainly in the Sequoia and Kings Canyon National Parks. It destroyed vast stretches of forest, but when it hit the Giant Forest, a landscape that has been subjected to prescription burning, it died down. Giant Forest is the poster child of why you want to do prescription burning. If you can burn frequently enough, you can keep the fuels down and prevent massive crown fires.

When you get to Southern California, there is a belief by some fire management agencies that they work the same way: that we need to put fire onto the landscape. But these are landscapes that are already suffering from excessive burning due to human ignitions. There’s the potential for prescription burning to actually be damaging to resources. If agencies are going to burn in chaparral, they need to try and focus on areas that are older than a couple of decades.

Photograph of a live giant sequoia tree, with green canopy. It is surrounded by trees with burned trunks and dead foliage.

A giant sequoia tree survives a fierce fire that killed off many of its fellow trees. Prescribed burns, also known as controlled burns, have been highly effective at protecting trees in Sequoia National Park.


Is that a hard message for policymakers to hear?

The problem is that chaparral is an extremely dangerous fuel. And the older the vegetation is, the harder it is to control the fire. So agencies don’t want to wait that long between fires, for the sake of buildings and other infrastructure that might be damaged.

The US Forest Service and National Park Service are very concerned about natural resource issues. For example, the National Park Service in the Santa Monica Mountains about two decades ago rewrote their management plan to greatly reduce prescription burning based on our research, because they realized they’re already getting way too much fire for the resources on their landscapes.

For most of Southern and Central California, fire protection is the responsibility of Cal Fire (the California Department of Forestry and Fire Protection). They mostly deal with rural lands, which are private property. And so resource issues aren’t always their primary concern when it comes to management decisions. What we need to do is balance fire hazard reduction with resource conservation.

Why is it so difficult to change people’s perceptions about California’s wildfires?

Well, there’s absolutely no question that policymakers, agencies and journalists like a simple story. They like a story where one model fits all. And so generally, mostly, what they hear is the story about fires in forest, which says: We don’t have enough fire, we need to put fire back into the system, we need to do prescription burning to get it back. And very few people hear the story that for some landscapes it’s exactly the opposite.

We have two different landscapes with two very different fire regimes that require two very different management practices. That’s really what we’re trying to focus on.