Crown Fire‎ > ‎

Spotting Fire Behavior

Evaluating spotting fire behavior requires the integration of three factors:
  • the source, size, and number of firebrands
  • the distance the firebrand is carried downwind
  • the probability of igniting a new fire at the downwind location.
Spotting Factors

Short-Range Spotting is not generally considered as significant in the growth of wildfires, because the advancing fire usually overruns the developing spot fire. Based on that, the assumption is that short range spotting is typically accounted for in the spread model outputs

Long-Range Spotting is differentiated from short-range spotting, primarily because firebrands are being lofted by a convection column and carried beyond the immediate fire area.

Estimating Maximum Spotting Distance

Both the included Spotting Distance Nomograms (with the associated worksheet in Section 5.5) and BehavePlus provide methods for estimating the Maximum Spotting Distance from a Torching Tree, or trees.

The maximum spotting distance model requires identification of tree species, height, and DBH of the torching tree to estimate the flame height and duration from the torching tree that will initiate the lofting of the ember into the windfield.

Further, the open windspeed is used to suggest how far the fire brand will be transported as it falls back to the ground.  The nomogram, because it assumes level ground uses the surface (20ft) windspeed and direction. 

The downwind Canopy, or Tree Cover, Height (reduced by half for open canopies) is used to factor out embers intercepted by the canopy before reaching surface fuels.

The graphic here depicts additional inputs to the BehavePlus spotting module.  In mountainous terrain, ridge top winds are used if wind is blowing across valleys as shown.  The shape of the valley is considered with inputs for Ridge-to-Valley distance and elevation difference.

Terrain Factors

Quick Reference Spotting Distance Table (under construction)

Maximum Spotting Distance Nomogram Instructions

Inputs Required:
Torching Tree: Species, Height, DBH
Open 20 ft Windspeed
Downwind Average Tree Height (Divide by 2 for open stands)

Nomogram 1 (Flame Height) & Nomogram 2 (Flame Duration)
Start with input DBH, draw vertical line to interest curve for input torching tree species, turn and draw horizontal line to determine flame height in Nomogram 1 and flame duration in Nomogram 2

Nomogram 3 (Firebrand Lofting):
Divide Flame Height (Nom 1) by the input torching tree height and use that value to select the curve in Nom 3. Using the flame duration (Nom 2), draw a vertical line from the bottom axis to intersect the selected curve. From that intersection, draw a horizontal line to determine ratio for calculating firebrand height. Multiply ratio from Nom 3 by flame height to determine firebrand height.

Nomogram 4 (Maximum Spotting Distance):
Using the estimated firebrand ht., draw vertical line from bottom axis on right to intersect curve for selected downwind tree ht. From intersection draw horizontal line to line for input windspeed, then down to spot distance.

Maximum Spotting Nomogram Worksheet 

(follow 1 to 5, left to right on each line)

Spotting Nomogram Worksheet

NOM 1. Maximum Spotting Distance (Flame Height)

Spotting Nomogram 1

NOM 2. Maximum Spotting Distance (Flame Duration)

Spotting Nomogram 2

NOM 3. Maximum Spotting Distance (Firebrand Lofting)

Spotting Nomogram 3

NOM 4. Maximum Spotting Distance (Maximum Spotting Distance) 

Spotting Nomogram 4

Spotting Likelihood/Probability

Though tables for Probability of Ignition are provided in the Fuel Moisture section, they describe only the likelihood that an ember will ignite a fire in receptive fuels. WFDSS spatial analyses integrate the potential frequency and distance for spotting fire behavior, but frequency information is hard to isolate. 

Combining Maximum Spotting Distance with Probability of Ignition

Spotting Threat

Isolating Spotting Spread Potential with FSPro

A suggested method for applying the WFDSS spotting models is to isolate the potential probability of spotting across significant barriers using FSPro analysis. FSPro is normally used to apply probabilities associated with windspeed and direction combined with day to day variability in fuel moisture scenarios.  The MTT spotting spread model includes monte carlo probability assessments associated with embers lofted from crown fires.  

Consider these adaptations to the FSPro Inputs:

  • Assume a 1 day analysis with forecasted ERC, Windspeed and direction favorable to spotting spread.  
  • Assume a number of fires (several hundred to several thousand) using the one day scenario from above
This method assumes that the scenario is favorable for torching and spotting behavior.  Insure that the analysis will produce at least passive crown fire.  The number of fires assumed for the analysis will drive the variability in:
  • ember source location from user applied probabilities
  • ember size and, therefore, distance 
Probability contours produced by the analysis reflect the locations for possible spotting and the relative probabilities for their occurrence.

Integrating Spotting Spread into Fire Growth Projections

FARSITE, FLAMMAP, and FSPro attempt to integrate the estimate of the number of embers and the distribution of distances they travel into the fire growth projection. Estimating maximum spotting distance from nomograms or BehavePlus only suggests an outer perimeter for spotting potential. 

Comparing Spotting Estimation Tools

Tonja Opperman, Fire Applications Specialist, Wildland Fire Management RD&A, assembled these tables based on contributions and discussions among many fire behavior researchers, programmers, and practitioners, including: Pat Andrews, Mark Finney, and Chuck McHugh at the Missoula Fire Lab; Brian Sorbel from the Alaska Region of the NPS; Mitch Burgard and Erin Noonan-Wright from the Wildland Fire Management RD&A; Stu Brittain with Systems for Environmental Management in Missoula; Joe Scott with Pyrologix in Missoula; and Rick Stratton from the Pacific Northwest Region of the USFS. Corrections can be forwarded to Tonja Opperman at tonja_opperman@firenet.gov

Maximum Spotting Distance (non-spatial) 

System

Maximum Spotting Distance 

from Torching Trees Nomograms

BehavePlus v.5.0.5

Inputs

  • Torching tree height 
  • Torching tree species 
  • Torching tree DBH 
  • Average tree height 
  • 20-foot wind speed
  • Torching tree height 
  • Torching tree species 
  • Torching tree DBH 
  • Downwind canopy height 
  • 20-ft wind speed 
  • Number of torching trees 
  • Ridge/Valley elevation difference 
  • Ridge/Valley horizontal distance 
  • Spotting source location

Spotting Process


Nomograms published in Rothermel (1983) are based on the model published by Albini (1979).

Predictive mathematical model to calculate the maximum distance an ember will travel from a single firebrand is lofted from a torching tree to calculate maximum distance.

Spotting from torching trees is based on nomograms with options for multiple torching trees and terrain adjustment (Albini 1979, Chase 1981, Rothermel 1983).

BehavePlus also calculates spotting from wind-driven surface fire (Albini 1983, Chase 1984, Morris 1987) and spotting from burning piles (Albini 1981).

Spotting from active crown fire will be added in BehavePlus v6.0 (Albini et al. 2012).

Outputs


Maximum spotting distance from a single torching tree on flat ground is read from a nomogram.

Maximum spotting distance from torching trees (single or multiple) is displayed in a table and graph.

Probability of ignition is calculated separately.

Assumptions & Limitations


Gives maximum distance only.

Assumes level terrain; single torching tree.

Does not account for likelihood of trees torching, firebrand material availability, the number of spot fires, or the probability of ignition for that firebrand.

Gives maximum distance only.

Accounts for terrain and number of torching trees.

Number of torching trees is used to calculate firebrand lofting height; higher firebrands travel further, all else equal.

Does not account for likelihood of trees torching, firebrand material availability, or the number of spot fires, or the probability of ignition for that firebrand


WFDSS Spotting Spread (spatial)

In all geospatial systems, embers are only generated from passive and active crown fires, not from surface fires, fire whirls, burning piles, or structures. Spotting can be turned off or set to zero in all tools.

System

Short Term Fire Behavior (STFB)

FLAMMAP MTT similar

Near Term Fire Behavior (NTFB)

FARSITE similar

Fire Spread Probability (FSPro)

Inputs


Canopy characteristics from spatial layers (LCP).  

Analyst specifies foliar moisture content.  

Spotting tree species is always grand fir with a DBH of 20 cm (7.9 inches).

Analyst sets spotting probability. 

Wind speed/direction is constant for entire burn period, but can use gridded winds that are modified based on topography.  

Weather is static, though dead fuel moistures can be conditioned. 

Canopy characteristics come from spatial layers (LCP). 

Analyst specifies foliar moisture content. 

Spotting tree species is always Grand fir with DBH of 20 cm.

Distance and perimeter resolutions are determined from the LCP resolution; timestep is 60 minutes. 

Wind speed/direction input can be changed hourly.

Weather is changing through analysis with DFM conditioning throughout. Gridded winds are not yet available. 

Canopy characteristics are from spatial layers (LCP).  

Foliar moisture content always 100%.  

Spotting tree species is always grand fir with a DBH of 20 cm.  

User sets spotting probability for each fire danger (ERC) bin. 

Winds can be probabilistic, forecast or a combination.  

Weather can be probabilistic or forecast or combination. No fuel moisture conditioning. 

Ignition delay is optional.

Spotting Process


Fire behavior is calculated for each cell. Nodes are on fixed grid equal to LCP spatial resolution.

User-set spotting probability determines which predicted crown fire cells (and associated nodes) can produce spots. Those nodes generate a single ember with random distance from zero to the max for that node. For cells predicted to have active or passive crown, 16 incrementally-sized embers are lofted.

Max ember distance & azimuth are calculated using canopy cover, crown fraction burned, elevation, and all available wind information.

Embers landing on unburnable or already burned substrate do not ignite. Embers landing on burnable substrate always ignite (Finney 2002). Similar to spotting in FlamMap 5.0 and FSPro.


Vertices loft 16 incrementally sized embers. The number of vertices depends on perimeter and distance resolutions & timestep. Each ember goes through a random draw process based on user-set spotting probability

Ember distances and azimuth are based on canopy cover, crown fraction burned, elevation, winds, and species/DBH. Embers are tracked until they burn out or land.

Ignores all embers that land within one cell from where the ember originated, as the main fire would eventually burn over these spot fires.

A spotting grid overlaid onto the LCP allows the first spot that lands in a burnable fuel model to “seize” that cell so no other spots can land in that particular cell.

Same process as STFB

Outputs


Models fire behavior for every cell simultaneously for a single scenario, and uses MTT to calculate fastest fire travel paths. Embers produced only with passive and active crown fire.

Randomly lofts a single ember from a node if the predicted fire type is passive or active crown fire.


Simulates lofting and downwind travel of individual embers of different sizes from each vertex that exhibits passive or active crown fire.


Fire probability surface output that may or may not distinguish spot fire activity.

Limitations & Assumptions


Spotting only occurs when passive or active crown fire is modeled. Finney and Scott & Reinhardt methods are available for crown fire; each calculates crown fraction burned (CFB) differently. CFB and canopy cover are used to determine “number of torching trees” (1-10) used in firebrand lofting height.

More embers will be lofted at finer landscape resolutions. Faster ROS will encounter more nodes, but the absolute number of nodes is static. One ember per node; less chance than in NTFB/FARSITE that an ember will travel the maximum distance.

NOTE: Users will probably want to set spotting probability higher in STFB than for NTFB tools.


Grand fir is used as the spotting tree species for the entire landscape. Distance resolution, perimeter resolution, and timestep are automated.

The minimum spotting distance (set to the landscape resolution) essentially skips the first cell'. For example, on a 60-meter landscape, no spots occur in the first 60 meters from the perimeter, but any viable embers that land beyond 60 meters can produce spot fires.

NOTE: Users will probably want to set spotting probability lower in NTFB than in systems using MTT (STFB, FSPro).


Same as STFB. Finney and Scott & Reinhardt crown fire methods are available; each calculates crown fraction burned (CFB) differently. CFB and canopy cover are used to determine the number of torching trees used in firebrand lofting height.

NOTE: Users will probably want to set spotting probability higher in FSPro than for NTFB tools. Calibrate FSPro with STFB utilizes consistent spotting methods.

Ċ
Robert Ziel,
May 6, 2017, 12:20 AM
Comments