Home Science When bushfire experts disagree

When bushfire experts disagree

Science relies on dispute – especially in the face of the climate crisis.

The 2019–20 megafires are seared into the minds of most Australians in a series of red-tinged images: thousands seeking safety on beaches, exhausted firefighters catching a moment’s shuteye, massive towers of smoke rising from the landscape, corpses of wildlife among scorched forests.

During that long summer, scientists found themselves in the media daily to explain the causes, behaviour and impact of the fires – much of which we don’t actually fully understand. The Black Summer fires were unprecedented in their scale and power, exhibiting extreme behaviour such as an extraordinary number of pyrocumulonimbus events: thunderstorms generated over intense fires.

“Not all fires are created equal.”

Professor David Lindenmayer

Those same scientists are now working hard to retrospectively understand the fires and inform how we can reduce their risk and impact in the future.

But it’s not always easy to research a topic of such intense public interest.

Earlier this week, a paper published in the journal Nature Ecology and Evolution sparked a backlash among Australian ecologists – suggesting that forest management, including native forest logging, had little-to-no impact on the severity of the Black Summer fires.

At the centre of the debate are two prominent Australian ecologists: David Bowman from the University of Tasmania, who was lead author on the paper, and David Lindenmayer from the Australian National University, who has called the paper “rather unfortunate” – “poorly framed, badly analysed, with the narrative actually not matching the data or the analysis”.


Hard numbers: the Black Summer bushfires, 2019–20

  • Largest fire from a single ignition point ever recorded in Australia: 500,000 hectares burned at Gospers Mountain, NSW, started by a lightning strike
  • Between 650 million and 1.2 billion tonnes of CO2 expelled into the atmosphere
  • More than 23,000 bushfire insurance claims lodged
  • 5.4 million hectares burned in NSW

Source: The Climate Council


Bowman, a professor of pyrogeography and fire science, says he’s “a bit puzzled” by the strong response to his findings.

Some people, he says, are suggesting “the paper is wrong because the science is settled… You’ve got to be kidding. It’s just starting.”

The paper delves into the main drivers of the scale and severity of the 2019–20 megafires. Bowman and his team used satellite imagery to examine 2.35 million hectares of land across NSW and Victoria, spanning a third of the total area burnt in the Black Summer fires, and compare it to historical records on past fires, logging, weather, vegetation and land use.

Statistical tests then revealed which factors were associated with higher damage to the forests. Perhaps unsurprisingly, the dominant contributor was weather (including prolonged drought and the persistent hot, windy conditions), as well as topography (the shape of the landscape, with ridges burning more severely than valleys).

“Blaming land-management practices for the Black Summer bushfires is wrong.”

Associate Professor Brett Murphy

The results align with previous research, which has found that it is now virtually certain that human-caused climate change is making bushfires longer, more frequent and more intense.

So why do other ecologists have a problem with it?

Lindenmayer’s issue has nothing to do with the climate – it’s about the paper’s conclusion that forest management made no difference.

The study team found that more than 44% of native forests suffered severe canopy damage, despite a reduction in logging in recent years.

“The extent of past logging in the last 25 [years] was much lower (4.5% in eastern Victoria, 5.3% in southern NSW and 7.8% in northern NSW) suggesting no connection between fire severity and logging,” they write in the paper.

But Lindenmayer, a professor of ecology and conservation biology, disagrees with this.

“There is an extensive body of science that shows there are strong links between logging and fire severity and the data in this paper actually reinforces exactly that, despite the misleading title of the paper,” he says. (‘The severity and extent of the Australia 2019–20 Eucalyptus forest fires are not the legacy of forest management’.)

It might seem like logging would reduce the available fuel and make a landscape less likely to burn, but previous research suggests the opposite. Not only does logging generate a lot of waste biomass of coarse woody debris, but it also makes the forest drier, and the smaller trees that sprout up in place of an older forest are more flammable.

“Nobody has ever said that that logging was responsible for the fires. We’ve always known that the effects of climate and weather are the key drivers of fire… but forest management is also a contributor and you can see it in the data.”

Professor David Lindenmayer

The forest then burns more often, creating a vicious cycle that strips out moisture and encourages the new growth – which Lindenmayer terms a “landscape trap”. Environmentalists have welcomed this idea as providing scientific grounding to oppose native forest logging – it’s a big deal if stopping logging could reduce the risk of megafires.

But Bowman’s new paper places logging way down on the list of contributing factors to the fires’ intensity, overshadowed by weather and terrain.

“Forest management – including whether there had been logging or prescribed burning in recent years – was of little importance,” summarises Associate Professor Brett Murphy from Charles Darwin University, who was not involved in this study but previously worked in Bowman’s lab.

“It seems likely that under very extreme fire weather conditions, such as those experienced in the summer of 2019/2020, the effect of forest management becomes irrelevant. Blaming land-management practices for the Black Summer bushfires is wrong.”

But Lindenmayer counters: “Nobody has ever said that that logging was responsible for the fires. We’ve always known that the effects of climate and weather are the key drivers of fire… but forest management is also a contributor and you can see it in the data.”

He expresses concern that Bowman and team have conflated two different kinds of fire in their analysis.

“When a fire burns through a landscape,” he explains, “you can have relatively benign kinds of fire that lead to the damage of vegetation, or you can have very, very high severity fire that actually burns the crowns of trees.”

These “crown fires” are extremely difficult to control, leading to fast spread and even generating intense pyrocumulonimbus storms.

“The concern is that where it happens, logging makes more crown fires that are impossible to control and have huge effects on the final size of the fire.”

Dr Philip Zylstra

“A lot of work has already shown that crown fire is more likely in logged forests,” adds Dr Philip Zylstra, a bushfire researcher at Curtin University, who was also not involved in the study.

But the study did not address this issue, instead focusing on crown scorch, which Zylstra points out can be caused even by quite small flames.

“They did actually show that logging increases the likelihood of scorch so that logged forests were about as likely to scorch on mild days as undisturbed forests were on very high fire danger days,” Zylstra says. “Despite that, they argued that logging still wasn’t an issue because it wasn’t responsible for the amount of scorch over the entire burnt area.

“Of course, no one has ever said it was; the concern is that where it happens, logging makes more crown fires that are impossible to control and have huge effects on the final size of the fire. This study didn’t address that at all.”

“You need to be careful about what kind of fire you’re analysing, because not all fires are created equal,” notes Lindenmayer.

He says that the paper’s analysis averaged the two, minimising the effect of crown fires and glossing over the nuance of fire severity in specific areas.

But Bowman says he’s “not downplaying anything – I’m foregrounding what made those fires so extreme, and the answer is weather and climate.

“The effect of prior burning and prior logging…is not even a consistent effect. It depends – it’s contextual.”

He says that it would have been great to show that every time fires sweep through forestry areas, they generate pyrocumulonimbus clouds: “It would mean that there’s very strong advice that forest management is affecting…those events, which are impacting the Earth system by injecting particles and CO2 into the stratosphere.”

But he says there is not yet a definitive link between these clouds and a land management signal.

“I’m not downplaying anything – I’m foregrounding what made those fires so extreme, and the answer is weather and climate.”

Professor David Bowman

“What I would like is to zoom in at the smallest scale, and go ‘Right, let’s really tease out at this fire – what effect did this fire have under these fire weather conditions, what effect did forestry have?’” Bowman says – on both the flammability of the forest and its resilience.

This, it turns out, is also what Lindenmayer is doing; he is releasing a paper in the coming weeks, analysing the effects of forest management on a burnt area of north-eastern Victoria on a smaller and more detailed scale.

Lindenmayer says that although large-scale analyses are important, understanding the nuances of how fire moves through different types and ages of forests is also crucial to both preventing and fighting fires.

“The relationship between the age of a forest, the flammability of the forest, and the severity of fire is what we call nonlinear,” he explains.

Young forests (up to seven years) are less likely to burn, while from seven to 10 years forests see a steep increase in the risk of fire severity, which declines again once a forest is over 40 years old.

“The lowest severity fires are actually in the oldest forest,” Lindenmayer says. “So if I have a fire in a heathland and a fire in a 20-year-old forest, I’m going to go to the 20-year-old forest first because that’s the one I really need to suppress quickly.”

This nonlinear response was lost in Bowman’s analysis, he notes.

But why is this debate about forest management getting so heated when both Bowman and Lindenmayer agree on the most important issue – that the climate is the dominant influence?

“Just having this sort of extreme polarised confusing argument – I don’t think it benefits anybody.”

Professor David Bowman

“We know at a certain point, once the weather conditions escalate, nothing is going to stop the fire from burning straight through,” Bowman acknowledges.

But as Lindenmayer says, “climate change is a colossal, wicked problem” – a long-term problem with long-term solutions that can paralyse us.

“A shorter-term problem that also adds to the flammability of the system is forest management,” he adds, and this is easier to act on.

Continuing to research and better understand the impact of land management on fire is therefore necessary as we look to the future.

Disputes and contestations, like those in the ecology community this week, are part of the zigzagging research process that relies on rigorous peer review.

Some people, Bowman says, have “a Hollywood view of science – you have that eureka moment, you publish the paper and the world changes… That’s not real.”

“I see this as an educational process. I don’t mind if somebody can find a way of doing what we did better or identify a problem.”

“I don’t welcome the sort of abuse that’s come from this…on both sides. I don’t think it helps.”

Professor David Lindenmayer

But he would like “people to take our analysis and pull it apart and rebuild it and publish a more critical or a different analysis – something that’s evidence-based, rather than just opinion.

“Just having this sort of extreme polarised confusing argument – I don’t think it benefits anybody.”

“I don’t think that helps,” agrees Lindenmayer. “I welcome the Bowman et al paper,” he says. “I don’t welcome the sort of abuse that’s come from this…on both sides.

“I think there are flaws in their analysis – but they might think that about some of the things that we’ve done. I think it’s important to have these debates.”

According to Bowman, the research is really only just beginning. Through many publications from many different groups, he is “100% certain there will be a convergence, where we’ll go ‘Yeah, okay, now we get it.’”

Exit mobile version