Bigger, stronger, disastrous: How climate change fuels wildfires

February 15, 2018

Mark Cochrane has been studying the characteristics and behaviors of wildfires since graduate school. Now a professor at the University of Maryland Center for Environmental Science's Appalachian Laboratory, Cochrane discusses how wildfires differ across regions, if places like California are more prone to fires than others, and his research that shows wildfires will occur more because of climate change.


This is part of the transit out to where Cochrane does most of his fieldwork in Indonesia where peatlands catch fire.
Mark Cochrane and a master's student of his, Chris Moran, canoe through Minnesota. They had to paddle across three lakes with two portages to get to the burned forests.
Mark Cochrane with his research team during field work.

Story transcript

The fires were started with the best of intentions.

Where forests were cut back to make way for farmland in the Brazilian Amazon, landowners would plan controlled burns to give their cattle a place to graze. Their fires would race to the rainforest’s edge and stop, denied the fuel they needed to keep running.

It was a tradition passed down through families, but then one year, a shift in the weather would change that. Fire no longer obediently stopped at the forest’s edge, and what had been the end would prove to be only the beginning.

MARK COCHRANE: Once we got to the point where fires got into these forests, they don’t recover. You would need years, if not decades, maybe even a century before they would completely recover. Once you pass that threshold, now fires tend to get into those forests every year.

UMCES: This is Mark Cochrane, a professor at the University of Maryland Center for Environmental Science’s Appalachian Laboratory. He has been studying wildfires since graduate school, and recently was part of a research team that scoured 11 years of satellite data and 23,000 fires worldwide. In the end, they determined that large, high-intensity forest fires will increase. The reason was similar to what was happening in Brazil – higher temperatures, drier conditions and stronger winds.

COCHRANE: What we find is throughout much of the planet we have a much greater chance of having the appropriate weather conditions to spawn very large fires.

UMCES: In the news on a regular basis, it’s one of the more tangible examples of climate change. To understand how we got to this point, we have to go back to Brazil and Cochrane’s graduate school years.

COCHRANE: I basically got into it as part of my graduate studies, working in the Amazon. It was one of the things I was working on together with other elements of remote sensing to look at landscapes, but I really started focusing on looking at fire in rainforests. Rainforests, by definition, aren’t really supposed to burn, but I was looking at the fires that escaped into the rainforest. So that’s how I began, by looking at these dynamics and systems that weren’t really considered to be fire prone and looking at how they actually are impacted by these fires and what dynamics are at play.

UMCES: What do you mean when you say escaped into?

COCHRANE: It used to be that you would burn a fire and it’s a rainforest so they would burn it and have it go up to the forest and it would go out because it’s so wet. We started seeing more of these fires that wouldn’t go out – they would get to the forest and just keep going. It was burning through ranches and fruit tree plantations, but it was also burning through their sustainable forest fire projects, it’s at the location where their firefighters are based.

Nobody could stop these fires. The way they burn is kind of interesting. You have these small fires and they go along and then at about 5 o’clock in the evening, they go out. You think everything’s fine, the problem is the next day at 10 a.m. they all start again. What’s happening is at night, the humidity level rises, the leaves get too moist for it to burn, so the fires go out but they stay smoldering in some of the downed logs. The next day as things dry out again, once the leaves get dry enough the flames start again and start marching off again and this can go on for weeks.

Once they did burn, everything changed. What you had effectively was a very small fire moving really slowly, which doesn’t sound bad except for the fact that these are forests that are not at all adapted to fire, and because the fire moves very slowly it imparts a lot of energy to the tree around the bark and kills it.”

UMCES: These little fires would kill as much as 30 percent of the trees, carving out a path of scorched branches and dead leaves for the next fire that would inevitably be more devastating. Forest fires are the concussions of the natural world – each new one is more devastating than the last.

COCHRANE: When they burn the next time they were drier and there was a lot more fuel lying around, so they burn more severely. So the next time around it killed another 40 percent of the trees, but this was like half the biomass at this point.

We were able to show what happened was these fires got in during a drought during that year, got into the areas where the landowners couldn’t see, and that previously had some logging, burned so much that it looked like deforestation even though it wasn’t intentionally deforested. So the fires created this deforestation event in that region in 1995, but it wasn’t really widespread. After 97, 98, we had this El Nino and it became much more widespread and since then, it has just expanded again and again.” 

UMCES: Working toward his degree, Cochrane arrived in Brazil in time to see devastating El Nino’s effect. It forced him to rethink his next steps. So, he started to collect data from the fires that were burning, but Cochrane knew he couldn’t just look at a forest from the ground up. He also uses satellite data to survey forests from above. 

How do you use satellites to study forest fires?

COCHRANE: The way I tended use it most is what you call LandSat – LandSat is a system that looks at landscapes in a few bands, so you can see blue and red and green and some of the infrared, but it uses pixel sizes that are 30 meters on a side so about 100 foot on a side and you can see the difference between something that’s green in a forest or something that’s been cleared, that’s a field.

I was trying to look at these small fires that were burning underneath the canopies and not the active fires but what happened afterward. So the problem was with a jungle you can’t easily map what burned. It would kill some of the trees, but what that would do is just give you a forest canopy with a few holes in it. It’s only when you’re able to look at this other way of basically pulling out what’s there. It’s kind of like, if you have a fever, you still look like a perfectly ok person, but when we take your temperature, we go, ‘Uh oh, we have a problem.’

So how do you quantify how much area burned? We had to develop a method to do this and this method needed to see something smaller than the smallest pixel, so to speak, so we developed a method for looking at the fraction of dead material within that given pixel so that we could actually map on a landscape how big some of these fires were.

Being able to see this impact on these forests, that allowed us to say, Oh, we have a bigger problem. We could then go on the ground and verify these areas had burned. With that, we were able to start mapping this over larger areas and over time to say how frequent these fires were and how wide ranging.

UMCES: Can you talk about some of the places your research has taken you and why maybe some places might be more prone to them than others?

COCHRANE: There are different issues in different places. That’s one of the problems, we think of fire as one thing, it’s not. Fire is a combination of the heat and the oxygen and how it moves on that landscape. When it moves on a landscape, you can think of it almost as a living thing that’s just trying to eat, it’s trying to move to the next combustible fuel. What gets complicated is we have different amounts of fuel, we have different types of fuel, different sizes of fuel, different land that it’s on.

When we look at how a fire spreads, the first thing you need is something that burns. If it’s dry, that’s somewhat easy because if it’s wet, it takes a lot of energy to get the moisture out before you can burn. When it’s on a slope, heat tends to rise, so more of it gets into the next fuel, so it’ll tend to move up that slope faster. When we get wind, that brings in more oxygen to that flame, but it also tends to push that flame over such that the flames are now closer to the fuels to the ground and therefore, it tends to move faster in that direction. Grass it will consume very fast. One thing I don’t think people realize is more people die in grass fires than forest fires because they move so fast.

Worked in Australia where we look at a lot of systems there. There are drier systems there where we can talk about how the dynamics change in some of the more open grasslands of Northern Australia and how they’re trying to manage there by doing more burning early in the season. So they create lots of little fires instead of getting lots of big fires later in the season. Because if they burn early in the season they will go out at night so they don’t keep growing, whereas later in the season when it’s very dry, those fires just grow day and night and become immense.

Right now I’m doing a lot of work in Indonesia where we’re working in these peat swamp forests, which are basically tropical forests on top of very deep peat lands. There, they’ve gone in, cleared forests, and drained those peat lands so now when fires get in there, they actually burn into the ground surface and keep burning to the point where they release huge amounts of carbon, way more than any other system, to the point they can be somewhere between 5 and 20 percent of all global emissions in some years. We’re trying to work there, trying to understand how that dynamic plays out, so they can manage those lands better.

I’ve also done a lot of work in the United States where we tried to actually look at all the fuels management we’re doing here. You’ve probably heard of prescribed burning. We also do mechanical thinning where we’re trying to cut out some of the trees. We’re trying either to prevent fires or keep the fires from becoming too bad when they do occur.

Right now, we’re looking at fires here especially in the east to try to look at whether or not we are likely to have a shift in fire behavior here, because there used to be a lot of fire so at some point we’re going to have to have fire here to get back to any kind of a natural system. At present in the east, we’re not very well conditioned to deal with wildfire. It’s considered very unnatural here when it’s anything but.

UMCES: Better management will become increasingly important with climate change – and recent wildfires might be proof we aren’t as prepared as we need to be. 

What did you think about last year in the United States – was that coming, was that a fluke, was that something we might see more?

COCHRANE: It’s basically inevitable. We have a combination of factors it’s always hard to pin it down to just one thing. One of the main things we have is all kinds of people living out in these flammable areas. Some places it’s a factor of us having kept fire out too long so things keep growing and we have more fuel. Some of it has to do with the weather patterns that come along so for example California had these Santa Ana winds, so it acts like a hair dryer burning down through there and creating huge fires, but all around the country we have more and more of this and in the east we don’t think about it anywhere near enough. It used to burn a lot more here in the eastern US as well.

We had an example of that in 2016 when we had the Gatlinburg fires where we had immense fires and very severe burning that people had not seen in their lifetimes.”

UMCES: In the late fall of 2016, wind fueled two devastating fires in Gatlinburg, Tennessee. In all, the fires burned 17,904 acres, mostly in the Great Smoky Mountains National Park, killed a dozen people, and injured 191.

COCHRANE: Those fires were just beyond anybody’s experience, it’s not that they were any worse than the fires that happened in the west in California. It’s just that they were unprecedented for that region. It’s not that the fires are becoming beyond what has ever been seen before, but they are changing beyond what has been seen in the locations that are occurring. 

This is why we’re concerned about climate change. Climate change is not about the temperature going up a degree or two; it’s about everything that goes with that, all the weather patterns and all the changing disturbances, whether it’s snow or rain or wildfires in this case.

So this is another thing we’ve been studying. We’ve been trying to look around the world at how things are changing. It doesn’t change in one way everywhere. For example, one study we’ve done was looking at what we call fire danger – when are the conditions right for having severe fires? When we look across the planet, it’s been increasing over the decades, about 18.7 percent total increase in the chance of having these really severe fires, but it’s concentrated in only on about a corner of the planet where it’s changing a lot. Unfortunately, in the US, the western US is one of the places where it is changing quite dramatically. Other areas are changing such that we’re changing fire behavior. When we look at that we have a lot more chance of having very severe fires and the likelihood increasing over time because of changing climate.

All around the planet right now we’re seeing these changes – some places less fire, but many places more fire and when we get those fires they tend to be larger, they burn longer, they burn at the wrong time of year, and so they’re becoming more and more problematic for people.

UMCES: In other words, wildfires will likely make even more headlines in the future, unless we can cut global carbon emissions to curb the strong winds and dry climate conditions that trigger fires. Cochrane estimates by 2041, there will be 35 percent increase of catastrophic fires per decade. That’s about four extreme fire events for every three that occur now. And some of those fires might break out in places where we’re not wholly prepared to manage them.

But Cochrane is hoping to change that, before the next wildfire ignites.