Denise Cullen
Cosmos correspondent Denise Cullen is reporting on the annual Australasian Fire Authorities Council conference in Brisbane this week.
Researchers have called for a radical rethink on how to predict, detect and manage bushfires as their scale and incidence escalates hand in hand with global warming.
More than 2,000 delegates attending the first day of the Australasian Fire and Emergency Services Authorities Council (AFAC) conference in Brisbane this week heard how, according to a UN report, extreme fires are expected to increase globally by up to 14% by 2030, 30% by the end of 2050 and 50% by 2100.
Two decades ago, when Canada’s Kelowna fires destroyed 239 homes, Dr Jen Beverly, then a new research scientist tracking wildfire evacuations, described them as an “outlier”.
“If you’re a data person, and you’re looking at 28 years of data, and you’ve got this one big spike, [then] obviously that’s an outlier,” Beverly said in her keynote address to Australasia’s largest emergency management conference and exhibition.
“But it wasn’t really an outlier – in the years that followed, we had event after event after event.
“Kelowna really was the start of something new.”
Beverly, a former firefighter who is now Associate Professor within the Department of Renewable Resources at the University of Alberta, points to a subsequent series of catastrophic Canadian wildfires – Slave Lake in 2011, Fort McMurray in 2016 and Lytton in 2021.
But the most compelling evidence of a fundamental shift came with this season’s fires which Beverly describes as “the worst on record by a long shot”.
Still ongoing at the time of writing, these wildfires have burned more than 13 million hectares so far, forced the evacuation of 200,000 people, and killed 4 firefighters.
Historical data and antiquated methods can’t be relied on to plan for the future.
Jen Beverly
Beverly reveals photo after photo of communities burning – Old Crow, Gun Lake, Hay River, Buffalo Narrows and more.
“It’s remarkable how many there are and how similar they all look,” she says.
“And after a time, you’re just completely saturated by these orange skies, and people fleeing and stories of devastation and heartbreak.”
At the same time, the smoke still hangs over the Hawaiian island of Maui after the United States’ most deadly wildfires in more than a century killed at least 114 people there earlier this month.
While the precise cause of the blazes remains unknown, it’s believed that the hot, dry conditions arising from climate change contributed.
Fires have also raged in Croatia, Portugal, Greece, Spain, Italy, Algeria and more.
The horror unfolding in the northern hemisphere may provide a glimpse into the upcoming Australian fire season, with the Black Summer fires of 2019-20 still fresh in people’s minds.
According to the Bureau of Meteorology, an El Niño event, involving hotter and drier conditions, is now more likely than not.
Beverly argues that the enormity of the fire risk means that new approaches are necessary.
“Historical data and antiquated methods can’t be relied on to plan for the future,” she says.
She scorched the dominant approach – computationally complex burn probability simulations – given that fire event probabilities couldn’t be assigned with any degree of certainty.
Through comparing these predictive maps with actual events, she found that fires were burning predominantly in low burn probability areas.
“We need to move away from the fire event as a unit of analysis and decouple variable and fixed risk factors – so we can focus on what we know,” she says.
Factors such as fuel moisture availability, wind speed and direction, and ignition agents, were highly variable within a fire season, with lots of predictive uncertainty, Beverly explains.
On the other hand, factors which remained mostly fixed within a fire season included what burns and where it is (fuel), operational capabilities, and the locale (receptors).
In a paper published by Natural Hazards this year, Beverly and her colleague at the University of Alberta, Air Forbes, explored the role of wind speed and direction in wildfire disasters.
They also developed a simple, standardised method for measuring and displaying the directional vulnerability of a locale, based on how exposed the landscape was to any encroaching fire.
“We don’t have to grow [by mathematically modelling] 100,000 fires,” she told the conference.
“Why don’t we just look around at every point in the landscape and see if fire can get to it?”
Her paper notes that when combined with current and forecast weather conditions, directional vulnerability assessments could provide an instantaneous red flag for informing suppression priorities and related response actions, such as evacuation decisions.
Directional assessments could also be used to help plan fuel reduction treatments around a community, proactively identifying candidate locations for supporting operational tactics such as fireline construction in the event of a fire or identifying transport network-related evacuation vulnerabilities.
Closer to home, the Minderoo Foundation, a not-for-profit founded by Andrew and Nicola Forrest in 2001, has set its sights on developing the capabilities to put out dangerous fires within an hour.
“If the Black Summer bushfires taught us anything, it’s that we need to change the way we fight fires,” says Rania Poullos, the Minderoo Foundation’s Fire Shield Lead.
Poullos said that while a combination of techniques, including lightning storm monitoring, observation towers, ground-based cameras and aerial monitoring, are used to detect fires in Australia, the majority are still reported through triple-zero calls.
If the Black Summer bushfires taught us anything, it’s that we need to change the way we fight fires.
Raina Poullos
In practice, this means fires can go unnoticed for quite some time in unpopulated areas.
According to the CSIRO, a 10-minute delay in detection can see a fire’s area grow by up to 1,500 square metres.
“For early detection and monitoring to be fit for purpose for emergency services, a global constellation of satellites fitted with sensors to detect fires at a few metres in diameter every 15 minutes is what is needed,” Poullos says.
She described the current baseline satellite capabilities that are used for fire detection in Australia as “not fit-for-purpose”, as they provided limited coverage or did not see at a resolution that was useful for detecting fires early.
For instance, Himawari-8 can provide the fastest revisit time of 10 minutes but has low resolution imaging (500-2000m) which is not useful for early fire detection, while Sentinel 2A, 2B provides 10-60m resolution imagery but only has a revisit over Australia every five days.
“Low earth orbit satellites with fit-for-purpose fire monitoring sensors and near-real time image processing and analytics can provide unprecedented situational awareness of bushfires at state, national or even global scales,” she says.
“Technologies exist to develop these systems and a new global competition has been developed to encourage global innovation in fire detection.
“A coordinated global approach can see the outcomes of this competition deliver impact at the operational level on a large scale.”