When it comes to what we know about bushfire behaviour, big questions remain unanswered.

This week, the New South Wales state coroner, Teresa O’Sullivan, handed down a 734-page report after a two-year inquest into the black summer bushfires. She described fire prediction as an “inexact science”, and said there was a degree of uncertainty in all the existing models used to forecast what would happen as the risk worsened with the climate crisis.

One of the major controversies is over the way fuel loads for bushfires are modelled. There are markedly different views among scientists working in the area. The modelling used to estimate forest litter on the ground guides hazard reduction and firefighting plans.

The differences were highlighted when Prof Mark Adams, a veteran ecologist, published an academic paper earlier this month arguing the modelling usually relied on is fundamentally flawed. He says lives are at risk because of the way forest litter is estimated in Australia’s most populous states.

Not everyone agrees. Associate Prof Phil Zylstra, an ecologist at Curtin University, calls Adams’ position a “distraction from the real issue” and argues surface fuel loads do not drive bushfire risk.

He says the question of what drives fire has become “very emotive”, and a major area of debate.

What to make of this?

• Sign up for Guardian Australia’s free morning and afternoon email newsletters for your daily news roundup

Adams, a Swinburne University professor, believes leaf litter is crucial in understanding fire risk.

He says modelling used by the Rural Fire Service means the New South Wales agency is likely to have underestimated how much fuel had accumulated across the state’s vast forests before the catastrophic 2019-20 fire season.

NSW, Victoria and Australia’s major fire research agency use the Olson model to predict fuel accumulation and calculate fire danger ratings. Developed more than 60 years ago, the model assumes that leaf litter, twigs and bark – known as fine fuels – rapidly accumulate at first, and then fall on the ground and decompose at the same rate, reaching a “steady state” that remains unchanged unless they are burned.

View image in fullscreen

Adams says this model leads to “significant underestimates of fuel” – in some cases as much as half of what is really on the ground.

“Why use a 60-year-old model, which is not used anywhere else in the world?” he says.

“Every gardener would know that twigs decompose at a much, much slower rate than leaves. To assume they all decompose at the same rate is simply non-scientific.”

Zylstra strongly disagrees. He says fire intensity is not determined by the weight of litter on the forest floor. “It’s all about the arrangement of it and the flammability of the plants themselves that are burning,” he says.

“Think of a newspaper laid flat on the ground. If you set that on fire, you can get it to burn but it won’t burn very well,” he says.

“If you take a single page from it, and just loosely crumple that page up, you’ve got a much, much lighter fuel load, but set that on fire and you’ll get a much bigger flame.”

The NSW RFS relied on the Olson model to publish a list of “comprehensive vegetation fuel loads” before the black summer bushfires of 2019-20 burnt through more than a quarter of the state’s 20m hectares of forest.

Adams says given most of these forests had not burnt for “many, many decades”, the RFS is likely to have underestimated how much fuel there was to burn.

But the RFS risk management director, Dr Simon Heemstra, defends the research. He says the Olson model “works very well” in NSW, although he acknowledges there are some “nuances” – such as variations in climate and weather – that it misses.

“There are issues with our fuel models, we’d definitely like to improve [but] the reason we use modelling is it’s just not physically possible to sample the whole landscape,” he says.

Adams recently had a paper outlining his arguments against the Olson model published in the journal Forest Ecology and Management after what he acknowledges was a “challenging” but ultimately successful peer review process. Adams and his co-author, Mathias Neumann, accuse governments, agencies, and researchers of relying on misinformation.

Some government agencies say they use this type of modelling and have faith in it as the best method available to calculate bushfire fuel loads.

A spokesperson for the Victorian environment department says there is “promising” research identifying alternatives to the Olson model but they are not yet “fit-for-purpose”.

A spokesperson for Victoria’s Country Fire Authority says it does not use the Olson model but does use “similarly shaped” modelling to determine fuel hazards “in some circumstances” – which it says it validates with field assessments.

View image in fullscreen

The internationally recognised ecologist Dr Luba Volkova, from the University of Melbourne, is particularly concerned what the current approach means for estimates done by the major government-funded research agency working in this space, Natural Hazards Research Australia.

Volkova says the agency, which informs state and federal bushfire policies including hazard reduction burning, should put more work into field research rather than using models that rely on assumptions about the way fuel accumulates.

“[The] US, Europe have a huge chain of permanent sampling plots across the country – even New Zealand has it – which are measured every couple of years or every year,” she says.

The agency’s chief executive officer, Andrew Gissing, says it isn’t surprising that scientists don’t always agree on bushfire management, given its complex nature, and that all science is “built on healthy, robust debate”.

Adams says he is concerned that recent literature on fire risk and fire agency fuel predictions rely on “an unpublished and unreviewed report” and a separate 2014 paper that used the Olson model to calculate fine fuels in eucalypt forests in south-eastern Australia.

“Neither the report nor the paper provides evidence of ecosystem steady state for even a single study forest, let alone all 80 forest formations [in south-east Australia]” he writes.

But one of that paper’s authors, Ross Bradstock, refers to Adams’ recent journal article as “Mark’s story”. Bradstock says it’s wrong to suggest that conventional forest litter estimations “somehow lead to underestimation of bushfire risk”.

“It sounds really dramatic, but whether there’s underestimation or overestimation is actually unimportant,” he says. “The variation in most real world fuel data far outstrips the magnitude of any under or overestimation.”

He’s vowed to produce a “very comprehensive response” to Adams’ paper, which would also need to be accepted by a scientific journal and peer-reviewed.

“The Olson model has served us well,” Bradstock says. “Scientists do modelling to help people.”

Source: theguardian.com