This article first appeared in Wildfire magazine.
The author, Rick McRae, argues that the impacts of climate change must be better included in wildfire management strategies. McRae is an adjunct professor with the Bushfire Research Group at UNSW Canberra.
After a career as an ecologist, senior emergency manager, and bushfire scientist I have a particular view of where climate change is taking us; it is fundamentally based on Australian conditions, but I have an international perspective that is both operational and scientific. A lot of people say a lot about this problem, but too many are saying different things. Who does one listen to, especially if you enjoy your comfort zone? You may disagree with my views, but rather than dismiss them, start a conversation with your colleagues and think about how what I am saying might affect both them and your collective goals. My career and my research has all been aimed at reducing wildfire risks. Here I simplify some topics, and omit a lot of necessary technical detail, but I completely support the outline that I present.
The global collective of fire management wisdom is clearly focused on a fuel-oriented path forward in the face of climate change. The Landscape Fire Governance Framework that arose from the 8th International Wildland Fire Conference in Oporto in May 2023 is the latest element of a global framework. The framework states that fires are getting worse due to a combination of too much wildfire suppression, a lack of investment in fire management, and changes to how communities handle fire on the landscape. A common theme in discussions is the need for more fuel management, either through more fuel reduction burning or a switch to Indigenous practices.
To this end, planning typically includes a focus on risk reduction through hazard reduction via fuel management. Training, equipment, and systems are focused on this system, matched by budget allocations. Satellites show that certain countries produce a lot of smoke from this risk-reduction effort. For normal wildfires, fire services and their communities do a very good job mitigating the risk. (Can this ever be enough?) Climate change is increasing fire danger as the world warms up, and fire services and land managers are correctly adapting to the heightened risks.
At the same time, the world is being severely affected by what are called extreme wildfires, which dangerously couple with the atmosphere above.
It is critical to correctly use the terms normal and extreme: normal fires spread by quasi steady state fire behaviour – if you know the fuel and the terrain, then you largely know what the fire will do; extreme wildfires have one or more blow-up fire events (BUFEs), where the fire couples with the atmosphere and exhibits dynamic fire behaviour, which often involves feedback loops and so the details are largely unpredictable. Figure 1 shows their relationship.
For BUFEs, there is no explicit role for fuel load (beyond the need for a prior fire), indicating that fuel management – central to the framework – is unlikely to be an effective preventative action. We do, however, need to explore how fuel management can be targeted to prevent future dynamic fire escalation. Extreme wildfires do not occur in flashy fuels such as most grasslands: they are mainly a problem in forests and woodlands and they have, in recent years, occurred in new ecosystems (discussed below). (See figure 2.)
When an extreme wildfire couples with the atmosphere after being triggered by dynamic fire behaviour, a BUFE occurs, lasting up to three hours, and typically burning 50 to 100 square kilometres (20 to 40 square miles). With little opportunity for fire suppression the only real incident objective is to save lives. Saving structures may put fire crews at risk for little return. This minority of fires cause the majority of damage.
Figure 1. The relationship between the fire drivers for normal wildfires with quasi-steady state behaviour and the fire drivers for extreme wildfires with dynamic behaviour. The left is quasi-deterministic while the right involves unpredictable feedback loops.
The incident action plan for affected sectors and divisions during a BUFE looks very different to that for a normal fire. Locally appropriate strategies and tactics need to be formulated to help save lives.
There is an archive of decades of high-quality satellite data that is informing many aspects of the challenges associated with extreme fires; it will become increasingly important that we get the full leverage off the datasets involved. The complexity of the changes already underway can be overwhelming. It will be important for end users to make clear what their needs are, and for them to accept the answers produced.
While many authors have used forward-looking climate models to anticipate how climate change will impact fire risks, observations are now showing a far more alarming picture overall.
Fire thunderstorms, called pyroCbs, are the most obvious manifestation of extreme wildfires. A recent study found that there has been no recent global trend in the frequency of pyroCbs. Global pyroCb activity has always been dominated by fires in and around Boreal forests. However, areas such as Australia, South Africa, South America, and the Mediterranean have only recently started having problems with extreme wildfire. Canada, in 2023, experienced the most protracted ever season for extreme wildfires, globally. Australia’s Black Summer was just as prominent with record breaking intensities.
Figure 2. The drivers of fire risk. The “depleted” column is where dynamic fires usually occur.
An important step must follow on from recognition of the wildfire-type dichotomy: operational doctrine must be revisited. As an example, in Australia, the national doctrine for operations in the urban interface lacks any dynamic fire behaviour elements. This document is founded on decades or experience during fire fighting and is state of the art – for normal fires only. What is different? When a BUFE arrives at the urban interface, it is characterised by: (1) a lack of a headfire, with a switch to dense spotting, and a high chance of loss of overall situational awareness; (2) an ember storm (a sea of flowing pea-sized embers flowing over the ground), which is very different to typical ember attack (which is more like a mortar attack); (3) strong turbulence; (4) a darkened sky; and (5) much deeper penetration of the urban edge. Air ops are likely to be impeded.
Also, standard doctrine is often founded on past damaging fires, but key lessons from previous events may need revisiting if, as is often the case, those fires were driven by processes subsequently discovered, such as the key elements of dynamic fire behaviour.
Several past landmark fires have featured descriptions of the fire spreading sideways on the lee face of a ridge. We have seen this in news footage, with chief officers waving their hands sideways during media briefings, or even in official post incident reports. After being identified in 2003 in the Canberra fires, a scientifically validated concept called Vorticity-driven Lateral Spread (VLS) is now known to be the cause. VLS is by far the main cause of forest fire damage in rugged landscapes, globally. Fire service operations based on key lessons learned need to adapt to this. A lookout at a fire where VLS might occur has to be trained to look to the rear at certain landform elements, as opposed to the prior practice of focusing on the headfire. To avoid VLS-driven BUFEs, it may sometimes be an option to burn-out VLS prone areas ahead of the main fire when fuels are too damp to support spotting. Another key instance of the need to rethink is that dynamic fire behaviour is often associated with large air tanker accidents. Climate change is leading to large aircraft flying out of aviation weather into fire weather while climate change is turbo-charging weather close to the ground.
It used to be that different countries had different types of fire, and therefore different operational approaches. Climate change is reducing these differences. I identified a fire near Canberra in 2004 as being foehn-wind driven. Some time after that my collaborators and I wrote a paper on this, introducing Australian firefighters to an idea that has long been a mainstay of training in North America and the Mediterranean Basin. Over the following decade we found only a few good cases of local foehn-wind driven fires. Then during Black Summer, with hundreds of BUFEs, perhaps 50 per cent of those were of this type. That is a massive escalation.
These changes clearly suggest that the world needs a multi-pronged adaptation strategy to climate change’s impacts on wildfire risk. The strategy for normal fire is well understood and must be implemented and continually improved upon. The strategy works better than is acknowledged, because the metrics for success were developed using data from both types of fire. The inclusion of dynamic events with bad outcomes biases the outlook.
In passing, a serious issue arising from lumping all fires together is the mis-training of artificial intelligence and machine learning systems being developed to help mitigate bushfire risks. Just because a fire was attributed as something in a database 25 years ago does not mean that that is correct in today’s thinking. Climate change will not be forgiving to field crews using poor intelligence.
A new strategy is required for rapid adaptation to extreme wildfires. The ongoing escalation suggests a need for the multi-pronged approach to be created as quickly as possible. I have developed a framework for predicting dynamic fire events in the forests of south-east Australia, which aims to show the potential for new thinking (Figure 3). The framework seeks to predict BUFE events using hydrology, remote sensing, and fire ground data in a multi-scaled way.
For the adaptation strategy to work it is necessary to define the following: ownership (by a global body); working membership; protocols; data and accounting needs; professional development protocols; and dissemination channels.
The mandate for climate change adaptation for wildfires might include:
- Focussing on extreme wildfires (to complement on-going collaboration on normal wildfires);
- Defining, owning, and disseminating research goals;
- Providing a hub for research outcomes;
- Providing a forum for international exchange of relevant operational lessons;
- Maintaining a global overview of wildfire problems and tracking the overview’s evolution;
- Rapidly disseminating new information or certified lessons from major fire events.
Figure 3. Two decades of predictive analysis on the potential for pyroCbs in the forests of southeast Australia. PyroCbs (red bars) occur when alerts are generated by the system, either due to temperature anomolies (green bars) or landscape hydrology (blue bars). The orange line clearly shows the impacts of climate change on air temperatures in Canberra, while the purple line shows a more worrying trend for offshore sea-surface temperatures. The difference between the two sets of 12-month average anomalies – the Canberra Dipole (black line) – is critical for BUFE potential. At the peak of Black Summer, Canberra had an extraordinary 12-month average temperature anomaly of 3C. Similar frameworks could work elsewhere.
Students of the evolution of wildfire can look at the references cited in many new wildfire papers and see – from the references alone – where the paper was written and what technical specialty it is from (for both the authors and the journal). However, this Fire Tower of Babel situation is not good enough. In a similar vein, if we are to collaborate on these problems, we must standardise the terminology. The use of alternative terms, and the widespread misuse of others does nothing to aid adaptation –foundation terms such as pyroCb or megafire are key examples – and surely reinforces the previously mentioned issue with the training of machine learning systems.
The wildland fire sector needs to stop being overly distracted by fuel loads, otherwise we will all be affected by extreme wildfires and their impacts on ecosystems, communities, soils, hydrology, biodiversity, traditional practices, and the upper atmosphere – including the ozone layer.
Rick McRae served as a headquarters technical specialist in what evolved to become the ACT Emergency Services Agency in Canberra from 1989 until his recent retirement. He worked in business planning, arson investigation, multi-hazard risk assessment, as planning officer for major incidents, weather specialist, and as a research scientist focusing on extreme wildfires, and especially pyroCbs. McRae has conducted case studies, described new phenomena, and developed predictive tools. He maintains a website that aims to present operationally useful material on extreme wildfires: https://www.highfirerisk.com.au/.
McRae is an adjunct professor with the Bushfire Research Group at University of New South Wales Canberra.