UPDATE, October 22, 2014: The Wildland Fire Lessons Learned Center uploaded a better, higher-resolution version of the video. We replaced their original version we had earlier embedded, with the improved one you see above.
****
This video is a recording of an October 8, 2014 webinar on the effectiveness of firefighting resources in suppressing large fires. I hesitated to embed it here because about a third of the dozens of the graphics are illegible. They only use a portion of the available screen and the resolution is very low. Expanding the video to full-screen does not help. However, the content is interesting.
Here is how the webinar topic was described:
Dave Calkin presents on webinar on October 8, 2014. Wildfire management currently represents over 50 percent of the US Forest Service’s total budget. Suppression of large fires represents the single largest category of fire management and typically exceeds $1 billion annually. In both 2012 and 2013 large fire suppression exceeded the Agency’s budget allocations by over $400 million. Despite the scale of this investment relatively little is understood about how suppression actions influence large wildfire spread and those conditions that ultimately lead to containment. There is considerable uncertainty in managing large wildfires including the quality of weather forecasts, complex environmental conditions, variation in the type and quality of suppression resources, and whether or not requested suppression resources will be assigned.
In this presentation we review several recent studies that attempt to understand how suppression actions influence fire progression as well as review variation among Incident Management Teams in the amount of resources that they use to manage large wildland fires in the US. Despite these recent efforts, there remains limited understanding of suppression effectiveness. These results suggest that modeling large fire containment as a production process of fireline construction similar to traditional initial attack models is inappropriate. Improved understanding of large fire management effectiveness and efficiency will require spatially tracking individual resource assignments, activities, and tactics within the broader suite of fire management objectives and strategies.
One of the key facts the researchers needed in their study was how resources assigned affected the containment of the fire.
The National Wildfire Coordinating Group defines “contaiment”:
The status of a wildfire suppression action signifying that a control line has been completed around the fire, and any associated spot fires, which can reasonably be expected to stop the fire’s spread.
It is well known that many incident management teams do not accurately report the daily containment percentage, usually pulling a number out of their rear end that is much lower than the actual amount of fireline that is constructed. They don’t have the courage to report the facts so they lie, fearful that if there is competition for resources a lower containment percentage will enable them to obtain and sometimes hoard firefighters, crews, engines, and aircraft — regardless, in some cases, of greater needs elsewhere. On a fire we visited in 2013 managed by a Type 1 incident management team we found that even though it had been contained for a couple of days, and there was very little mopup that still needed to be done, the Incident Commander reported a very low containment percentage in order to make it easier to justify an evacuation order to the public.
The researchers realized this, so they ignored the official percentages reported on the daily Incident Status Summary report, the ICS-209. They analyzed fires for which perimeter maps were available for each day. When a section of the fire perimeter stopped moving permanently, for the purposes of their study they considered that area “contained”.
They found that on 50 fires they looked at, when the entire perimeter stopped moving the average containment reported was 64 percent. Of course, there may be good reasons for not declaring a section of line held or contained. It may not move in that area, but it could still require fireline to be constructed. Reasons for a fire to stop moving other than proactive suppression, include changes in weather, fuels, and topography.
So it is not possible, using ICS-209s or mapping data after the fact, to accurately determine the actual containment of a fire. However, the method used by these researchers may provide a figure closer to reality than the data reported by many incident management teams.
Geographic Area Coordinating Centers and Multi-Agency Coordinating Groups that have to allocate scarce resources may be tempted to use the method described in this webinar to truth-check the information reported by incident management teams.
“…usually pulling a number out of their rear end that is much lower than the actual amount of fireline that is constructed.” Sounds like a classic case of rewarding (with resources) bad behavior (under-reporting containment).
The usual fix in other domains is to change the reward structure, often by changing what is considered a “reward”. For example, measure performance based on efficiency of use of resources versus total resources used.
I don’t know enough about wildfire to know how this could, or even if it would, work. But it’s a thought.