UNDERSTANDING A CRITICAL FIREGROUND SAFETY THREAT
OBSERVATIONS OF VORTICITY-DRIVEN LATERAL SPREAD IN A WILDFIRE

BY RICK MCRAE

In September 2024, the Airport Fire burned on a wooded range of hills that sits within the outer suburban areas of greater Los Angeles (Figures 1a and 1b). The fire did something both spectacular and remarkably dangerous. Field fire crews operating in hilly or mountainous terrain with forest, woodland, or perhaps even heathland fuels need to understand what happened, as do their supporting technical specialists and incident management team leaders.

Imagine you were a crew leader in the headwaters of the valley below and west of Santiago Peak. Sometime in the morning you would have seen smoke rise up as the fire burned onto the far side of the ridge to your south-west. It is a long way off to the side, and the wind was nowhere near blowing the fire toward you. You could see several fire aviation units swooping in on that area. There is so much happening that you call in a situation update.

Shortly after 10 a.m. the fire reaches that ridgetop. You discuss it with your crew. Suddenly the smoke goes dark and the flames become visible even though they are two miles away. The fire starts spreading sideways just below the ridgeline at right angles to the wind – straight toward your crew. In the next few minutes, you see spot fires starting downwind of this and running backwards uphill as well as upwind. The fires quickly merge.

You discuss your safety and egress options. The decision is made to leave. It is a half-mile walk up a steep slope to the trucks.

By the time you reach the trucks and stow your gear, the fire is half a mile away at right angles to the wind, but it spans the valley downwind to your south. The fire is moving toward you with a massive, rotating black plume and spotfires everywhere. Embers hit the trucks as you finally drive off.

This is a scary scenario; it is know that this type of event has occurred a number of times, leading to several burnovers and catastrophic outcomes. (It is possible that the Yarnell Hill Fire in Arizona on June 30, 2013, was such an event.)

Noteworthy known and suspected catastrophic VLS events:
In the United States
• Mann Gulch Fire (Aug. 5 1949, at −111.9°, 46.9°)
• Yarnell Hill Fire (June 30, 2013, at −112.6°, 34.7°) In Australia
• Linton Fire (Dec. 2 1998, at 143.6°, −37.7°)
• Canberra Firestorm (Jan. 18 2003, at 149.0°, -35.5°)
There are many more.

Welcome to the strange world of vorticity-driven lateral spread, or VLS. I discovered VLS just after the 2003 wildfires devastated Canberra, Australia’s capital city. In airborne infrared imagery, the fires were clearly spreading sideways at up to three miles per hour (five km/hr) as well as downwind with lots of dense spotting on a landscape scale. It took a few years for the science to peg the actual mechanism. It is now confirmed in models and wind tunnel tests.

As an aside, VLS had been seen before 2003, but without high-quality imagery it had neither been identified nor understood. In fact, many events had been explained afterwards using “normal” concepts – banging the round peg into the square hole. This is true for both Australia and the United States, and probably also Canada.

Wind blows straight up the windward side of a steep ridge arranged nearly at right angles to the prevailing wind. If the wind exceeds about 15 mph (25 km/hr) it goes over the ridge top but cannot drop down quickly; it rises off the ground, and a lee-slope vortex or eddy wind forms in the gap underneath. Here the lee-side surface wind is also blowing upslope!

If the fine fuel moisture content is under, say, five per cent of oven dried weight, then spotfires can start easily. If fire burns into the eddy, lots of embers swirl around inside it; the fire heats up and gets unstable. As the eddy is a “closed system”, low oxygen burning occurs, but embers are nevertheless ejected downwind to start spotfires.

 

 

 

 

 

 

 

 

 

 

Figures 1a and 1b. Location map. Black arrows show movement of vorticity-driven later spread (VLS) edges. Grey areas are urban. The High Performance Wireless Research and Education Network (HPWREN) cameras are on Santiago Peak at 117.53° 33.71°.

The Airport Fire did all of this within view of the pre-existing high-resolution High Performance Wireless Research and Education Network (HPWREN)
fire camera of the Santiago Peak Communications Site. We got the best ever record of what goes on. This happened downwind on Sept. 9 and, separately, upwind on Sept. 10. I have been able to do photogrammetry on the second VLS event and measure it. The numbers are scary.

HPWREN:

The High Performance Wireless Research and Education Network (HPWREN), a University of California San Diego partnership project led by the San Diego Supercomputer Center and the Scripps Institution of Oceanography’s Institute of Geophysics and Planetary Physics, supports Internet-data applications in the research, education, and public safety realms. https://www.hpwren.ucsd.edu/

WHAT DID WE SEE?

The lateral rate-of-spread (ROS) along the ridgeline started at 0.4 mph (0.7 km/hr) then rose to about 2 mph (3 km/hr) when the plume got active – an overall average of 1.2 mph (1.9 km/hr). Once high ground on the side of the valley nearer the cameras got involved things sped up – to 12 mph (20 km/hr) – sideways!

When the spotfires got started near the camera, it took one minute for things to get nasty. The downwind rate of spread of the fire envelope (a line around all of the merging spotfires) was 28 mph (45 km/hr). When fire got very close to the camera, we could track smoke and embers.

Imagine that you were in our hypothetical ground crew and couldn’t reach your trucks, and that you were forced to activate burn-over protection protocols in a fire like this. If the Santiago Peak event is truly typical, then this is what you would experience:

Thirty-one minutes of ground-hugging, often dense smoke.An ember storm with the first smoke; all the embers were following the ground, not ballistic.

Three pulses of heavy smoke, every 10 minutes, each lasting eight minutes.Seven ember storms, three minutes apart on average, each lasting on average two minutes.

There was a spotfire recorded below Modjeska Peak. It ran uphill for 500 yards (450 metres) and burned 12 acres (five hectares). Extrapolated to an hourly value, that gives a slope-affected headfire rate of spread of nine mph (15 km/hr) and a burn rate of 150 acres/hr (60 ha/hr) just after ignition.

Another key observation was of an ember storm. In Figure 2, two video frames, 10 minutes apart, are overlaid after strong filtering. Spotfires have started just below the camera and are rapidly merging. The black dots are embers. They are radiating outward from a single point in the images (from a wide-angle camera). They are flowing in a layer over the terrain. This is called an ember storm. It is very different from ember (or firebrand) attack with a ballistic trajectory – falling out of the sky. Ember storms can blow under your fire truck or get sucked into the engine’s air intake.

So, if you are on a fireground where VLS might occur, there will need to be back-room technical specialists who are tasked with keeping you safe; they will need to be able to do the terrain and wind modelling to tell if there is a chance of VLS occurring and, if so, where and when. They need to model fine fuel moisture content to see if dense spotting can occur. They need to be embedded in an incident management team that immediately understands and acts when the fire behaviour analyst says “all crews must leave Sector Charlie immediately due to a threat of VLS.” They also need to understand the futility of retardant drops on VLS fire – they don’t work as the fire just spots past the purple drop lines! Also, critically, a VLS event can suddenly put new community elements at risk and compromise any evacuation options. Effective prior warnings must use VLS potential.

These insights come about because of the hill-top fire cameras. The cameras are run by

HPWREN out of the University of California San Diego and by Alert California. We have started using such cameras here in Australia. We desperately need to record more events like these. We do not know if this event was typical; future numbers may also be much larger than those in any training material. In Australia, we need to get a set of these numbers and observations under our belts to better learn what to expect. More observations, more scientific validation of ideas, more fireground safety.

The VLS wind-terrain model:

In general, the following characteristics are necessary for VLS:

• A ridgeline arranged within 40 degrees of being at right-angles to the prevailing wind.
• Slopes generally above 25
• A rapid change in slope. Typically, this is a ridgetop (with a switch to opposite aspects), but it can happen without a change in aspect such as at a clifftop.
• Winds generally above 15 mph (25 km/hr).
• Fine fuel moisture content below about 5%.

VLS rarely occurs in flashy fuels, such as most grasslands; they do not produce enough heat to interact with the lee-eddy.

 

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 focussing 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 now an adjunct professor with the Bushfire Research Group at UNSW Canberra.