"A scientific bounty from Santiago Peak!"
The Airport Fire, which ignited in Southern California in September 2024, blazed a destructive path across over 23,000 acres, reaching the summit of Santiago Peak. This wildfire not only consumed vegetation but also damaged critical telecommunications equipment, including infrastructure belonging to the High Performance Wireless Research and Education Network (HPWREN).
On September 16th, amidst remaining firefighting efforts, HPWREN received an unexpected email from Rick McRae, an Australian researcher with the Bushfire Research Group at the University of New South Wales. Intrigued by the fire's behavior captured by HPWREN cameras, McRae saw a valuable opportunity to advance his research on fire patterns and firefighter safety. His email, titled "A scientific bounty from Santiago Peak!", expressed his excitement about the data. An initial writeup of his showed HPWREN footage, and also included a pointer to a News story which showed a time-lapse from an AlertCalifornia camera. This serendipitous collaboration highlighted the potential for HPWREN's real-time fire monitoring to contribute to, for example, international wildfire research and safety efforts.
By Rick McRae, Bushfire Research Group,
University of New South Wales, Canberra
On the 18th of January 2003 a massive fire storm swept into the western suburbs of Canberra, Australia's capital city. I was the Planning Officer. Over 500 homes were lost (and over a thousand saved by residents and fire crews). Four lives were lost. In a city where no homes were lost to bushfires in the preceding eighty years, this was a major shock.
In trying to work out what happened, I explored aerial linescans (using the same Thematic Mapper technology as was on early LandSat satellites). These showed odd patterns of fire spread on their upwind edge. The smoke plumes were also anomalous. Using a GIS to explore terrain patterns, it was clear that the fire had spread upslope against the wind and then turned ferocious. This was new.
| Aerial linescan image showing VLS underway west of Canberra on 18th January 2003. The area in the top left had burnt days before. A smouldering tree fell over the containment line, probably only two hours earlier (white arrow), into terrain able to create VLS. Fire spotted downwind for a considerable distance - blue arrow. At the same time the leading edge of the event spread across-wind at an estimated 3 mph, dragging the dense spotting with it. Dense spotting can be seen. Less than an hour later another linescan showed that the spotting had spread nearly 2.5 miles further across-wind (white dotted line). The deep flaming resulted in a fire thunderstorm. Yellow is flame, and green is ember storms. (Imagery supplied by New South Wales Rural Fire Service.) |
|
It's name has changed - in my first conference talk on it, it was "the Breath of the Dragon", while later the fire science paper called it "Fire Channeling" based on its resemblance to forced wind channeling. More detailed studies - using wind tunnels and computer modeling showed it to be Vorticity-driven Lateral Spread, an honest but gruesome name for fire crews - so it is usually called the easier "VLS".
Imagine a fire day in eastern Australia with a westerly wind. As it crosses the crest of a north-south oriented ridgeline, it tries to bend down to follow the slope. If the slope exceeds about 25 degrees and the wind speed exceeds about 12 knots, the wind cannot bend fast enough - it separates from the ground on the lee side.
Underneath, a counter-rotating lee-slope eddy wind forms. Now imagine what happens if fire burns into the eddy. You, like everyone else prior to the 2003 fires, will probably get this wrong.
The simple answer is that the eddy fills with fire, smoke and embers, which are tossed around. Being a semi-closed system, the eddy expands. If it pushes upwards, it injects embers into the main air flow, which start spot fires for miles downwind. Here I have made the other assumption - as well as winds over 12 knots, the fine fuel moisture content has to be low enough to support easy spotting - we use 5%. So if the fuel is dry, you get a sea of merging spot fires downwind. But the eddy also expands sideways, taking the ember source, and new dense spotting, sideways with it. The fire spreads in two directions at once.
We call this "deep flaming", and it is critical because it changes the behavior of the fire plume. In an unstable atmosphere it resists mixing out to the cloud base, where a cloud condenses in it, releasing a lot of latent heat. In bad cases this cloud is a fire thunderstorm or pyroCb, and the latent heat of condensation released can exceed the heat from the fire. The cloud punches up to the top of the Troposphere. It gets to below -40 degrees Celsius and forms pyrogenic lightning. Many say that it changes the weather on the fire ground - but it may be changing over ten thousand cubic kilometers of the atmosphere. Big changes on a grand scale. Astronauts take photos from the ISS, and the ozone layer can be impacted.
If the air above is stable then this can't happen, and the plume stays under a cap. Foehn wind like dynamics arise and the fire can be just as vicious as under a pyroCb. In Australia I often evoke the Santa Ana Wind as an example of a foehn wind, and how it can create serious fire behavior. I was surprised to see that the Airport Fire burnt close to the city of Santa Ana. The videos show the plume spreading sideways overhead, under a strong cap.
In rugged landscapes, globally, VLS is the most serious cause of major fires and serious burn-over incidents. And yet the uptake of the lessons learned about VLS are only slowly being taken up by fire services. It is hard to convey the true import when using terms like vorticity and troposphere. Good photos are few and far between, and post-analyses can often fit established concepts to the evidence without the need to evoke new ones like VLS.
The Airport Fire reached Santiago Peak while the fire cameras were working.
On the 9th of September the cameras recorded the passage of a VLS event traveling sideways past them at close range, clearly showing the plume dynamics as well as distant spotting. The next day the same thing happened again but upwind of the cameras. The lateral spread was recorded in great detail. Spotfire behavior was clear to be seen. The cameras were coated with retardant by a passing large air tanker. Every fire expert that I know who has seen the videos has been stunned!
For the first time we are able to quantitatively measure VLS events, and use this to calibrate computer models and develop fireground safety protocols.
If fire gets into a ridge aligned across the winds, and the fire weather forecast says that the winds and fuel moisture are critical, then you do not want fire crews there. The blow-up fire event that is caused by VLS lasts for up to three hours, and can typically span four miles by seven miles (downwind), but we saw far bigger ones, especially during Australia's Black Summer (2019-2020). You must avoid the threat footprint, and the only effective incident objective for affected sectors is "save lives". All vegetation on all aspects gets burnt at extreme intensities.
Typically afterwards fire intensities stay mild for a while, allowing suppression after things have cooled off. But care is needed to watch for another, separate VLS event nearby.
After finding the new process after the 2003 fires I have studied a number of other VLS events, including in North America, where I have worked as an FBAN.
|
| An HPWREN image of the VLS approaching the Santiago Peak cameras. This form of fire behaviour is very dangerous because the fire is spreading in two directions at once, and it is spread by dense spotting, which can easily cross containment lines and can rapidly degrade situational awareness for fire crews. |
The dynamic involved does not involve fuel loads. The only fuel inputs that you need are firstly a running fire to enter the lee eddy, and secondly you don't get VLS in flashy fuels like most grasses.
We now have good videos of one VLS event, thanks to HPWREN and AlertCalifornia. If we get more in the future we can improve our theories and improve fire ground safety.
On the other hand, I hope to never hear about a VLS event again, because that often means people have had a really bad day.
If you fight forest fires in rugged landscapes and haven't heard about VLS, then you should pass a request up the chain of command to get something into the training curriculum. It is important for keeping you safe.
For more details, videos of this event can be found on YouTube. Examples include: