The Mann Gulch Fire killed 13 people in Montana’s Gates of the Mountains Wilderness in 1949.
The South Canyon Fire killed 14 people on Colorado’s Storm King Mountain in 1994.
And the Yarnell Hill Fire killed 19 Granite Mountain Hotshots near Yarnell, Arizona in 2013.
In each case, the firefighters were caught in a deadly burnover. When the weather changes abruptly, firefighters can die. And as communities push into the wilderness, they face the same danger, as residents of Paradise, California, can attest after their community was virtually wiped out in November 2018.
Until fairly recently, there was little that could be done to prepare for severe weather events that lead to rapid and unpredictable changes in wildland fire behavior, something which exposes both firefighters and communities to increased risk. But now scientists at U.S. Forest Service’s Fire Lab in Missoula, Montana, are working towards changing that equation.
The U.S. National Fire Danger Rating System (NFDRS) is an important tool used to quantify the timing and location of extreme events that lead to very destructive wildfires.
NFDRS helps maintain direct communications between land management agencies and communities. It also supports decision making at multiple scales, ranging from local Initial Attack decisions to long-term Extended Attack fires managed by Incident Management Teams, as well as regional and national level resource prioritization and allocation decisions.
This can be done at the Geographic Area and National Interagency Coordination Centers by utilizing a wildfire hazard projection system that can map areas with unusually high potential for extreme fire behavior and rapid fire growth.
Work is underway to incorporate high-resolution fuels and terrain information into these risk areas from the US interagency Landfire program. Landfire provides 30 layers of data per 30 meter X 30-meter map pixel, something which fire managers find to be valuable, especially when combined with NFDRS weather data.
A Little History
NFDRS has a surprisingly long history, as Missoula, Montana, Fire Lab Research Ecologist Matt Jolly relates.
“NFDRS started right here in the Fire Lab almost fifty years ago when the lab was tasked to build the first one. It started in the late sixties when the Fire Lab was designing and building a new spread model to use for simpler applications that could be applied over really big areas, and to support these decisions where you don’t really need a high-tech spread model.”
Even though the NFDRS has a long history, there was also a period in which little changed, as Jolly continued.
“In 1972 the Fire Lab released the first official NFDRS version, which was updated in 1978 and modified again in 1988. After that it lay dormant scientifically for about 30 years.”
Then, two years ago, the Fire Lab released an update to the system, picking up where they had left off 30 years before, making some improvements to the system with new science and some better sub-models, while maintaining continuity with what’s being taught to firefighters and fire managers, then rolling it out to the field.
Twin Goals of the Project
Jolly summed up what his aims are for NFDRS.
“I have two goals: to provide the best information that we can to our firefighters so that they can effectively manage risk; and to provide better information to communities so that they can be more proactive.”
The importance of the NFDRS is that it can provide information ahead of time for dangerous wildfire events and where they’re likely to happen so that communities can be informed and fire resources can be prepared to deal with the incidents early when they’re small.
“The idea of pre-season planning and staffing includes maintaining adequate resources during periods of high fire potential and using that information to decide how we would respond to a new ignition during initial attack, in terms of how many resources we would send and the tactics we would use to mitigate risk for First Responders and communities,” Jolly explained.
“Twenty years ago wildfires weren’t considered to be a big financial risk. People would say that one bad hail storm in the Midwest could cost insurance companies more than a wildfire in California. That is not the case anymore.”
The Nuts and Bolts of NFDRS
The current NFDRS is run on point source data measured by 2,200 remote automated weather stations scattered across the U.S. that is uplinked via satellite, making it near real-time. The new technological piece is migrating toward gridded data (a chunk of land versus a single weather station).
“We can use the data we get from the National Weather Service to provide not just current conditions but forecast fuel danger fire conditions for the next 7-16 days, depending on the forecast we use,” said Jolly.
“We’re really trying to expand the spatial part of it so we can map individual locations on a landscape with high potential, and also to provide those forecasts out longer so that we can be more proactive about when, where, and how we engage fire.”
To lengthen the projections, Jolly uses the National Weather Service’s National Digital Forecast database, which provides a 7-8 day forecasts throughout the U.S. in grids that are each 2.5 km X 2.5 km resolution.
“The real challenge is that we have coastal weather stations, but the coasts are a problem because we don’t have any traditional weather stations at sea. That’s why this new, gridded approach moves us to a better place. We take our weather station network and use it for baseline for bias correction and initialization of forecasts using the grids. It’s a good hybrid approach where we leverage our weather stations, some of which are not located in normal places. They’re in backcountry locations or mountainous terrain. Most weather stations are at airports or roadways. Having a network of stations that represents the terrain complexity and how that terrain influences the weather is a good thing.”
And resolution could be improving in the near future.
“The National Weather Service plans to move to a finer resolution, which we’re building into better spatial mapping tools,” said Jolly.
“WFAS - Severe Fire Weather Potential Mapping System – provides the current day’s and the next day’s forecast. These are based off the NFDRS indices, which can be applied anywhere on the grid. Any point is normalized so that we can show areas that are abnormally high for fire danger. We have one index which provides hot/dry anomalies that we call “energy release components”. We have another index which covers wind events that we call “the burning index”. We can combine those indexes into one to get the severe weather fire potential.”
So why go from point sources to grids?
“We’re trying to break out of the paradigm of needing to anchor every decision or assessment on where the closest weather station is located. Kansas, for instance, has four weather stations, whereas California has hundreds.”
And California is adding more all the time as power utilities install scores of weather stations for their own emergency centers. The bankruptcy of industry giant Pacific Gas & Electric, which was drowned in a sea of lawsuits following back-to-back years of deadly wildfires, some of which were directly attributed to PG&E’s equipment, has not gone unnoticed by power utilities in the state, which are quickly implementing measures that will protect their infrastructure as well as their shareholders’ profits.
Crisis Equals Opportunity
The utilities’ panic has also provided an opportunity for the Fire Lab. “We’re doing a proposal for the California Energy Commission, which is a really big group of scientists and technology companies,” Jolly remarked. “Next week California’s hosting a wildfire technology summit, where they’re gathering folks from the utility companies, CAL FIRE, and federal research organizations all in the same place.”
The Fire Lab is collaborating on a proposal with Technosylva to evaluate NFDRS for use by the utility companies. Part of that project would be to help utilities figure out where best to locate weather stations by grids, where the major channelling areas are because there are wind patterns in California that don’t really exist anyplace else in the country.
“The utility companies have developed their own version of fire danger ratings, but they’re all meant to be the same thing,” Jolly observed.
“They basically take weather and any other associated data and produce an index that is related to the likelihood that a fire will start or spread really fast. Ultimately they may end up using a hybrid of some of the things they have and some of the things we have in the NFDRS for their danger rating system.”
And the future looks bright with more organizations getting involved.
“We want to work with the utility companies, then it will be a true partnership. We won’t know for a while if we get this grant, but if we do we will be working directly with the meteorologists from the utility companies and the technology companies and universities. Then we will have all of these smart, capable people that can really add a lot of value to the process. The utility companies are getting on board by hiring really capable meteorologists who can anticipate events so that they can be proactive about managing the risk to their infrastructure and the consequent risk of their infrastructure starting wildfires.”
Jolly wants to see a concerted investment in technology to help improve how these events are anticipated, but he knows there are going to be bumps along the way.
“Moving forward there are still some learning curves. The momentum is starting to build, but it takes a while.”
For More Information
A prototype seven-day forecast system that has been running operationally for 3 years as part of the Wildland Fire Assessment System (WFAS). A desktop dashboard and mobile application are in development in collaboration with Technosylva, LLC.