Prescribed Fire Best Practices and Lessons Learned from the US Forest Service and National Park Service

I’ve been playing around with a new tool in the toolbox: SkyNet aka AI. I know, I know, “don’t give in to the Dark Side, use the Force!”

https://notebooklm.google.com if you want to check it out yourself. Feed it source material via uploaded PDF or Doc files, links to websites, even YouTube channels. It will assimilate all the data you feed it, and you can ask pointed questions, or have it create a FAQ, briefing sheet, or study guide for you.

One of the coolest features of this AI app is the ability to generate a “Deep Dive” audio podcast episode with AI hosts that’s customizable - you can have it touch on specific themes or topics, gear the podcast to a specific audience, etc.

I fed it the US Forest Service’s National Prescribed Fire Program Review of 2022 to see what it would do with it.

I read and digested this document thoroughly when it came out 2.5 years ago, but I wanted to experiment and see how the AI would handle such a technically, politically, socially, and ecologically complex subject - not just Prescribed Fire in general, but Prescribed Fire leading to catastrophic wildfire conversions - and how we can be better, safer, and more effective at burning going forward.

The results were … surprisingly good!

If you haven’t already listened, scroll back up and listen for yourself. The only error I caught, which doesn’t really rate junking the whole thing (one unfortunate thing: it is “take it or leave it” - you could use audio editing software to cut out the inaccurate/irrelevant parts), was that the hosts referenced KBDI - the Keetch-Byram Drought Index - as being hosted on the National Drought Monitoring Center’s website, which is incorrect. That site has lots of other drought products, but not KBDI, which is calculated daily as a WIMS output. KBDI can be found on the Wildland Fire Assessment System page in the same section as dead fuel moisture info.

Aside from this one small factual error, everything else discussed in the “deep dive” was actual-factual, if a little generic at times.

The ingredients for that podcast were the uploaded National Prescribed Fire Program Review PDF, and the following prompt to the “audio overview generator”:

Focus on best practices and lessons learned to increase prescribed fire safety and efficiency. Ignore references to “agency administrators” as irrelevant to the target audience. Target audience is private landowners conducting smaller-scale prescribed fires who want to learn from the mistakes and lessons learned of larger organizations such as the USFS.

Don’t want to listen? or want a deeper look at some of the findings and issues? Read on!

Let’s start with how this all came about…

The Hermit’s Peak/Calf Canyon Complex wasn’t the first time a prescribed fire has gone catastrophically wrong. Back in 2000, the Cerro Grande Fire (in the same general region of New Mexico) was sparked by a spot fire and subsequent escape of the Upper Frijoles Prescribed Fire in Bandelier National Monument (National Park Service) on the second day of multi-day burning operations.

Lessons NOT Learned: A Quick Look at the Cerro Grande/Upper Frijoles Investigation Report

Straight from the Executive Summary of that report:

The Fire Investigation Team concludes that federal personnel failed to properly plan and implement the Upper Frijoles Prescribed Fire, which became known as the Cerro Grande Prescribed Fire. Throughout the planning and implementation, critical mistakes were made.
Government officials failed:
• To utilize the correct National Park Service complexity analysis process.
• To provide substantive review of the prescribed fire plan before it was approved.
• To evaluate conditions adjacent to the prescribed fire boundary with regards to fire behavior, fuel conditions, and public safety in the event the fire crossed the planning boundaries.
• To complete and document the onsite review of critical conditions identified in the prescribed fire plan prior to ignition.
• To provide adequate contingency resources to successfully suppress the fire.
• To provide any wind predictions in the 3-5 day forecast for the periods of May 7 to May 9

Keep those things in mind, because fast-forward 22 years, and they show up again.

The US Forest Service National Prescribed Fire Review found many of the same issues, especially concerning escapes and wildfire conversions. These are some of the major bullet point findings from the 2022 Rx Fire Review - you’ll see several of the same themes as the 2000 Cerro Grande Investigation.

Almost all of these factors and findings apply to EVERYONE - yes, you too Joe Burner. This isn’t just about the “big-box-burning” Land Management Agencies!

In addition to the major points, I’m adding additional explanations and interpretation based on 20 years of Fire Management experience and being deeply invested in the study of “Lessons Learned” in hopes of not repeating these same mistakes.

• Inaccurate characterization of fuel types: this is rampant. On the private lands side, many conservation professionals, consulting foresters, etc, not specifically trained in wildland and prescribed fire, are not as attuned to the variations of fuel loadings and arrangements and how seemingly subtle differences can make big impacts in fire behavior. Prime example: a thick, dense Indian Grass or Bluestem field is going to burn with MUCH greater intensity than a somewhat sparser Broom Sedge field, even though they’re both “Native Warm Season Grass fields”.
  Another common mistake in my neck of the woods: writing a plan based on a Timber-Litter fuel model, when a heavy Forest Stand Improvement project (cutting down unwanted trees) is being performed before the burn, changing the unit to a Slash-Blowdown fuel model that will burn INFINITELY more intensely and is far more likely to throw spots and embers than a Timber-Litter model.
  In the public lands side, LANDFIRE is a two-edged sword: on the one side, many times the fuel model estimations are correct; when they’re wrong, though, they’re significantly wrong. Lazy analysts, planners, and modelers may find themselves bitten by the GIGO (garbage in, garbage out) Model Monster - which has downstream effects to planning of prescriptive parameters, fire lines, contingency resources, etc!

• Insufficient consideration of the impact of long-term [and short-term!] drought on prescribed fire behavior: Drought not only affects live fuel moistures, but dead ones as well. Dry soils conduct heat which can cause roots and underground organic material to smolder and burn, contributing to invisible fire spread and surface intensity.

• Not using available forecasting tools or insufficient training in using them: this isn’t just about Behave+ or IFTDSS (fire modeling systems) - numerous models and decision support tools exist for fuel moistures (if it’ll burn, and how hot), weather parameters, etc. Two things I use constantly: the Wildland Fire Assessment System for fuel moistures and KBDI references, and the Fire Weather Dashboard, which serves as a “one stop shop” for all your fire weather planning needs. I like the Dashboard so much I made a “how to use it” YouTube video!

• Inadequate patrolling and mop-up: and moreover, inadequate patrol and mop-up specs that allow folks to leave burning/smoldering material too close to fire lines! Is it smoking, smoldering, or burning? Then either move it WELL interior and away from your lines, or put it out! Keep patrolling your burn at minimum twice a day until it’s OUT COLD. A burn is like a campfire: when there’s no smoke or heat ANYWHERE - THEN it’s “OUT”, not a moment before!

• Underrating the complexity of burns: this one is an interesting one, particularly in private lands - new burners tend to over-estimate complexity, moderately trained/experienced burners tend to under-estimate it, highly trained/experienced burners are usually in the ballpark.
  Agency-side, there are very few Type-1/High-Complexity Burn Bosses nationally, so getting a burn into that Type-2/Moderate-Complexity rating category is necessary for many units to be able to have a realistic shot at getting that burn done. That puts a lot of pressure on folks to potentially underestimate potential risks and outcomes to get down to that Moderate Complexity Rating.

• Test burns not representative of the burn unit: this is double-true for units that have multiple fuel categories - meaning say, a grass field partly surrounded by a timber-litter forest, some of which has been TSI/FSI’d to an open woodland with significant “activity fuel” (logs, tree tops, etc) still on the ground. Folks will test fire in the timber-litter in the morning, see 6-inch flame lengths, and go “Great! This will work perfectly!” and it does…until you wrap around to that grass or open, slashy woodland at the peak heat of the day. Suddenly things are no bueno, and you’re chasing and fighting fire instead of lighting it.

• Burning at the upper end of prescription: This one should have been first, in big flashing neon lights. Corollary to this, when incorrect fuel models are used and/or modeling systems aren’t used correctly, you end up with bad prescription parameters to begin with. This means your prescription may be too hot to begin with, then you’re burning at the hot end of that too-hot prescription - and disaster ensues.

• Fire behavior not what was expected: like the above point, this wraps together the lack of situational awareness of fuels and moisture conditions, incorrect computational and mental models, lack of or misunderstanding of decision support tools, test fires in the most favorable fuelbed as opposed to the most potentially volatile, etc.

• Delaying declaration of a wildfire: in Joe Burner terms, don’t wait to call 9-1-1. If your folks can’t catch a spot, don’t chase it around for half-hour to an hour before making that call. Get the fire department enroute - because again, if you’re at this point, several of the above factors have lined up against you and it’s time to get help now.

• Communications and briefings failing to convey needed information: I’ve never really witnessed this personally on a prescribed burn, but I’m also a notorious question-asker and repeater of important points. For Joe Burner though, here’s a simple mnemonic to help you cover all the bases: I-WORD
  Introductions - experience, qualifications, skills, knowledge
  Weather - brief the latest weather forecast for your unit, and relate the forecast to your prescription
  Organization - who works for whom, “chain of command”, and communication protocols such as radio channels/frequencies, callsigns, etc.
  Risk Management - brief known hazards and issues in and around the unit, and protocols for things such as a medical emergency, spot fire, sudden weather event or fire behavior change that’s not forecasted, etc.
  Directions - ignition plan with quick review of the organization so folks are 100% dialed on who they’re working with and for.

• Gaps in communication with partners and adjacent landowners: this is big in private lands too. Talk to your neighbors. Many private land management consultants plan treatments, including prescribed fire, right out to the fence lines. This leaves no buffer if a spot fire occurs on a burn, and that isn’t the way you want to meet or interact with your neighbors! Early in the planning process, meet them, discuss your goals and objectives with them, and how you intend to ensure that things go according to plan and stay “in the box”.

• Surrounding fuels, values at risk, and potential for escape is another oft-overlooked issue. Folks spend so much time looking “in” at the burn unit they often neglect to look “up down and around” at the surrounding areas as well. Mentally game-out those potentials, or if you have the proper knowledge and training in decision support tool use, use those tools to help you figure out what your worst-case scenarios could be, and adjust your burn plan accordingly!

• Time/Social Pressure: This is huge - the “get ‘er done” mentality is a major “Human Factors” contributor to poor decision-making and accepting higher risks. This can come in the form of pressure from a boss or client, “Hey we’re paying you to do this, we need results,” “we’ve been prepping this unit for 3 years and never burn it” sentiments from crew members, or even your own professional pride and wanting the project off your plate instead of hanging over you.
  The worse yet: peer pressure when you find what seems to be a good burn day, you get all the folks and equipment out there…and the on-site weather isn’t lining up with the forecast, or the test fire is way more spicy than it should be. “Well…everyone’s here, it’s a huge waste and people will be pissed if we call it and send them home….let’s just go for it! It’ll probably be fine!” Famous. Last. Words.

Between listening and reading you’re probably already in for at least a half hour of your day, not to mention time spent digesting all of this and hopefully adding it to your Fire Toolbox, so I’ll not take any more for now.

Read, listen, learn, and don’t just work hard - think hard, Firelighters!