Scientists set up equipment in front of a spreading fire

(Note from Bill: A Fire Behavior Assessment Team (FBAT) has been studying the King Fire east of Placerville, California. Rae Brooks, an Information Officer at the fire, sent us this article describing what a FBAT does.)


These scientists set up equipment in front of a fire to collect fire behavior data

by Rae Brooks

FORESTHILL, Calif. — Fire scientists call them “plots.” Dotted ahead of the leading edge of the King Fire, they were deliberately placed in the anticipated path of the flaming front. Each plot contained a video camera, wind-speed gauge and other monitoring devices.

If the flames came, a data logger buried a foot underground would collect information that would allow the scientists to better understand the science of wildfire, gauge the effectiveness of fuel treatments, and contribute to firefighter safety.

So while thousands of firefighters were building line, clearing brush from roads and bulldozing contingency lines to suppress the King Fire, the scientists patiently waited for flames to sweep over their plots, if control efforts failed.

“We want our plots to burn,” said Carol Ewell, co-lead of the FBAT, or Fire Behavior Assessment Team, at the King Fire. “Firefighters put the fire out. It’s a difficult balance.”

fire behavior plot

A plot BEFORE the fire passed through. Photo by FBAT team at the King Fire.

fire behavior plot

The same plot AFTER the fire passed through. Photo by FBAT team at the King Fire.

Mark Courson, a division-qualified firefighter and an operations section trainee, served as operations lead for the FBAT at the King Fire. His job was to keep the team safe and advise on site selection.

“Usually I’m thinking of putting the fire out,” said Courson. “Now I’m bucking the system, thinking where suppression might not hold it.”

The team, a U.S. Forest Service module, draws members from around the country to work 14-day assignments on wildfires. Since 2006, the FBAT has recorded data at 16 wildfires, including last year’s Rim Fire, the third largest in California history. Generally, emerging fires with potential for growth suit the FBAT better.

At the King Fire, just three of 10 sites selected burned over. The team averages 50 percent, but sometimes gets lucky and finds all its plots burned.

“It’s a big gamble,” said Ewell. “Our success rate is quite variable, and I’m not sure that’s a hurdle that we can fix.”

The King Fire was particularly difficult to read after it made a speedy 15-mile run northward beginning late in the afternoon of Sept. 17, Ewell said. Rain has since quelled the fire, which is now 89 percent contained.

Despite the inherent difficulties, free-burning wildfires provide conditions that cannot be replicated in laboratory, experimental or prescribed fires. For instance, no prescribed fire would ever be set during California’s current historic drought. The team has recorded active crown-fire runs, fire whirls, spot-fire ignitions, and merger of spot fires with the main flame front.

The team’s prime mission at the KIng Fire is to study the effectiveness of fuel treatments. Team members set up plots in treated and nearby untreated areas to provide comparisons.

Because they are working ahead of the flaming front, they follow standard firefighting safety protocols, carefully considering whether they can get safely into and out of selected sites. “Unburned fuel between you and the fire,” is one of the 18 Firefighting Watch Out Situations, and, by necessity, their equipment must be placed in unburned fuel ahead of the fire.

“There is risk involved,” said team member Matt Dickinson, an ecologist at the Delaware, Ohio, location of the Forest Services’s Northern Research Station. “One way we mitigate is to set up plots early in the day before the peak burning period. We pull out quite often when we’re not feeling comfortable.”

It takes about an hour for nine FBAT members to install their gear and inventory the vegetation at each site. If firefighters are trying to build line and the team is trying to set up a plot, the firefighters, of course, have priority, said Ewell. “In comparison, we’re just a geek squad,” she said, although most FBAT members are experienced firefighters.

Some of the team's equipment. Photo by Mike McMillan.

Some of the team’s equipment. Photo by Mike McMillan.

At each plot, the team sets up temperature sensors, heat flux sensors, anemometers to measure wind speed, and video cameras encased in heat-resistant steel boxes. The cameras start when trigger wires are burned over. Each camera captures about 80 minutes of footage.

The flames often melt the anemometer’s plastic cups, so wind speeds might only be collected before flames arrive. At each site, team members also bury a data logger in a military surplus ammunition box. Other members collect information about the vegetation, down to counting sticks on the ground.

At the King Fire, the team also recorded fuel moisture data to help fire behavior analysts working on the fire.

In the future, Ewell would like to equip the team with more heat-flux sensors, Go-Pro video cameras with new triggers, and anemometers that can better withstand heat. The team relies on grant money and project funding from the Forest Service to continue its work.

When sites burn over, team members return, when it is safe, to collect their equipment and the data. They also record how the vegetation has fared. Plots that don’t burn are permanently marked with rebar, so they can provide fuels information for other uses.

The team spends evenings entering data and crunching numbers, and tries to complete a summary report before demobilizing from a fire.

Seeing a wildfire burn during severe drought has been an eye-opener, said Dickinson. Most of his experience has been with prescribed fires. He found the tremendous consumption of fuels and the severe damage to trees hard to witness.

William Borovicka, who normally works at the Vinton Furnace State Experimental Forest in McArthur, Ohio, was a first-time FBAT member at the King Fire. Back home, he studies how oak and hickory forests, if left undisturbed, convert to beech and maple woods. His work in Ohio, he feels, plays into management techniques to stabilize the oaks and hickories.

Using FBAT findings to gain a deeper understanding of fire behavior might similarly help forest managers someday with decision-making, Borovicka said.

“Maybe more prescribed fires, or different harvesting techniques,” he said. “Whatever it takes to prevent this kind of blow-up.”


Revised guidance for safety zones is released

Safety Zone Calculation

Safety Zone Calculation, released July, 2014. Bret Butler.

In his continuing efforts to improve the recommended standards for wildland firefighters’ safety zones, researcher Bret Butler has released a revised version based on additional research. Dr. Butler developed the guidelines that had been used for years which were based on the height of the flames, but in May, 2014 released a new recommendation that was based on height of the vegetation, wind speed, slope, fire intensity, and a constant number. This new July, 2014 version replaces the one that was released in May.

A safety zone is an area where wildland firefighters may be forced to take refuge from an approaching wildfire. There, a firefighter should be able to survive without being injured from exposure to the radiant and convective heat from the fire, and would not have to deploy and enter a fire shelter.

The latest version of the guidelines released a few days ago is based on height of the vegetation, wind speed, slope, and the same constant number (8). It removes a factor that could be a little subjective or difficult to quantify accurately in the field, fire intensity.

The new system, like the one unveiled in May, calculates the Safe Separation Distance (SSD) between the fire and the firefighters. To determine the SSD, using the table above multiply the constant number (8) times the number from the table (Slope-Wind Factor) times the height of the vegetation.

Example for 15 mph wind, 24% slope, 6-foot vegetation:

The Safe Separation Distance is   8 x 3 x 6 = 144 feet

Dr. Butlers’ Additional Considerations:

  1. For a 20-person crew, add 10 feet of radius and for a vehicle add another 5 feet of radius.
  2. The area in red requires large natural openings or construction by mechanized equipment.
  3. The proposed rule is to be used for flat ground rather than the existing flame height rule.
  4. Also consider additional lookouts on the ground and in the air to monitor fire activity with early egress to escape routes and safety zones.
  5. At 30% or greater slopes, hot gases tend to stay close to the ground.

Dr. Butler’s disclaimer: This proposed safety zone rule should be considered preliminary because it is based on limited data and analysis and subject to increase or decrease based on additional data. It is presented for release this fire season with the intent of increasing firefighter safety and reducing risk of injury. It is likely that an updated rule will be released in the next year.

For more information see the article in the International Journal of Wildland Fire titled: Wildland Firefighter Safety Zones: A Review of Past Science and Summary of Future Needs

We will let you know if another revised version of the guidelines is released in two months.

(NOTE: if you want a copy of the table above, click on it to open it in a window of its own, then click on Print in your internet browser.

Thanks and a hat tip go out to Ryan.


Unique display of results from fire effects study

Lake Clark fire effects

Screen shot from the fire effects study portal about a fire in Lake Clark National Park and Preserve

The Alaska National Park Service Fire Ecology crew returned to the site of the Currant Creek Fire in Lake Clark National Park and Preserve in Alaska one year after it burned in order to collect data about the effects of the fire. They created an interactive map tour of their findings, which turned out to be a pretty interesting educational tool.

Thanks and a hat tip go out to Traci.


The Atlantic, on wildfire research and the Yarnell Hill Fire


The Atlantic has produced two very good pieces on wildland fire.

One is the video below, about research into the science of combustion and how fires spread. It was filmed at the Missoula Fire Sciences Laboratory and has excellent production values and photography. Some of the researchers featured will be familiar to those who follow the topic; they include Mark Finney, Jack Cohen, and Sarah McAllister.

The other is a long-form article about the Yarnell Hill Fire that killed 19 members of the Granite Mountain Hotshots June 30, 2013. There have been several similar articles, but this one, written by Brian Mockenhaupt, is better researched and written than some. In addition to describing the fire, and the fire fight, It includes information about what went on behind the scenes at various dispatch and coordination levels, as well as the personal lives of the firefighters.

Here is a brief excerpt:

“To our families and friends, we’re crazy,” [Crew Superintendent] Eric Marsh wrote in the spring of 2013, in a sort of Granite Mountain manifesto addressed to the town of Prescott. “Why do we want to be away from home so much, work such long hours, risk our lives, and sleep on the ground 100 nights a year? Simply, it’s the most fulfilling thing any of us have ever done.”


Marsh took the lead in hiring new recruits, and focused as much on character as on stamina. “When was the last time you lied?” he asked in every interview. “Tell me about that.” Truth telling was a guiding principle for Marsh. He had quit drinking more than a decade earlier, and being honest with himself and others had become a big part of his sobriety.


Like many others who fought the Yarnell Hill Fire and who knew the hotshots who died, [Prescott Fire Department Wildland Division Chief] Darrell Willis has spent a lot of time asking himself why they did what they did. Part of the answer he’s come up with involves the very natural urge to fight and protect our own. “They wanted to reengage,” he said, standing by the posters. “Sure, they could sit up there in the black. But if they could try to get back in the game, they were going to. What they had been doing was lost. And that happens a lot. You put a day’s worth of work into something, and all of the sudden it’s gone, and you have to have a new starting point somewhere. There’s a lot of sweat and expended energy. So what do we do, just sit up here and watch it go by? They knew there was an evacuation going on, they knew there were people staying in their houses. So what would the public think? ‘You’re not going to help us? Why did you even show up?’ ”



Fighting fire in a beetle-killed forest

Mountain Pine Beetle, matchWhen a forest that has been attacked by pine beetles is on fire, there is a lot that we do not know about the flammability, crown fire potential, and resistance to control of these burning stands of conifers. Testing the torching potential of individual beetle-killed crowns was conducted in the winter over a ground covered with snow using a propane burner as the heat source. Flammability of vegetation has been evaluated in a lab. But it has not been proven that existing fire spread models can accurately predict the rate of spread of a stand of trees that has been attacked by pine beetles. As the authors of the paper below stated:

It is a shocking admission that the only empirical investigation of fire behaviour in live, lodgepole pine stands is limited to a single study, involving surface fires, carried out in British Columbia, Canada, 45 years ago (Lawson, 1972;1973).

In an effort to summarize what we do and do not know, three scientists, Wesley G. Page, Michael J. Jenkins, and Martin E. Alexander, collaborated on a paper titled Crown fire potential in lodgepole pine forests during the red stage of mountain pine beetle attack. The entire paper can be read here — their conclusions are below:



True insight into understanding and predicting the possible effects of recent [Mountain Pine Beetle] MPB-caused tree mortality on surface and crown fire potential in lodgepole pine forests has so far proven to be largely an intractable problem. While significant progress has been made in recent years documenting the effects of MPB-related tree mortality on fuel complex structure as well as seasonal and diurnal fuel moistures, trying to accurately assess potential fire behaviour using either operational or physics-based fire behaviour models has proven problematic. Except for the recent development in British Columbia, Canada, with respect to astatisticalmodel(Perrakis et al., 2012), existing models tend to be either inappropriate and/or un-validated for use in MPB-attacked forests. Current operational fire behaviour models used in the US are not capable of addressing the complex spatial arrangements of crown fuels that occur in recently attacked stands. Physics-based models such as WFDS may in time serve to be useful research tools and aid in understanding the dynamic nature of fire behaviour, but until the limitations and sources of error are better understood, interpretations of the resulting simulations must be viewed with scepticism (Alexander and Cruz, 2013a).

Observations from experimental fires and wildfires indicate that a real and considerable increase in crown fire potential exists in recently attacked stands with an increase in rate of spread on the order of 2 –3 times the no-tree mortality predictions. However, the amount of red foliage within the canopy has important implications on the duration of the increased crown fire hazard. Site-specific factors such as the total and yearly amount of tree mortality, the length of the outbreak, and the preexisting stand conditions could all be important factors that could affect the severityand duration of the crown firehazard. Additional factors such as the juxtaposition of red and green crowns and the relative importance of needle drop and subsequent decreases in CBD vs the increased flammability of red foliage may be important to evaluating crown fire hazard but as yet are not fully understood.

Limitations in the ability to accurately assess crown fire potential in MPB-affected stands are likely to persist until accurate wildfire observations and/or experimental fires can be used to either validate current fire behaviour models or derive the needed empirical proportionality constants in VanWagner’s (1977) crownfire initiation and propagation models applicable to MPB-attacked stands. A program of experimental fires (Alexander and Quintilio, 1990; Stocks et al., 2004a) coupled with more systematic monitoring and documentation of wildfires (Alexander and Taylor, 2010) is needed in order to address these current shortcomings and gain insight into the underlying processes controlling fire behaviour in MPB fuel complexes. It is a shocking admission that the only empirical investigation of fire behaviour in live, lodgepole pine stands is limited to a single study, involving surface fires, carried out in British Columbia, Canada, 45 years ago (Lawson, 1972;1973). Additional information on the physical processes of foliage ignition and the relative effect of moisture content under varying heat fluxes will also aid in the development and modification of physics-based models that would greatly enhance our understanding of fire behaviour in these forest ecosystems (Ma¨kela¨ et al., 2000).

As the number and size of MPB outbreaks in western North America declines, opportunities to conduct experimental fires and observe fire behaviour in recently attacked stands will decrease. Simulating MPB-attack, similar to Schroeder and Mooney (2009; 2012), by girdling trees provides a potential way to extend the window of opportunity for experimental fires and to control for confounding factors. Investments in gathering and compiling fire behaviour data by fire management and fire research organizations will help provide a means to objectively assess fire behaviour potential in this unique fuel complex, which will increase the margin of safety for future wildland firefighters and aid in operational planning for fire managers. Meanwhile, wildland firefighters should continue to be vigilant in recently attacked MPB-affected lodgepole pine forests and follow the guidelines outlined in the fire environment factors listed in the ‘Look Up, Down and Around’ table for insect-killed forests found in the Incident Response Pocket Guide (National Wildfire Coordinating Group, 2010).”


Photos from the Fire Lab

Test fire in lab

A test fire in the Fire Lab

These photos were taken as part of the Large Fire Conference during a tour of the Fire Sciences Laboratory in Missoula, Montana, is operated by the U.S. Forest Service. Most of the pictures were taken in sections of the lab where researchers work with actual live fire.

Soil heating test fire

Live fire to test the amount of soil heating

Soil test fire

Live fire to test the amount of soil heating

Fire test in wind tunnel

Fire test in wind tunnel

Below is an 11-second video showing a fire whirl the scientists created in the lab.

As bonus, since you made it this far into the article, are a couple of photos that were taken at the University of Montana during the Large Fire Conference in Missoula.

tethered balloon

A tethered balloon which could be flown hundreds of feet above the ground to host radio repeaters, cell phone repeaters, a camera, or a wireless router for internet service.

Tanker 41

A BAe-146, Tanker 41, flyover May 21 at the University of Montana during the Large Fire Conference at Missoula. Photo by Bill Gabbert.