FIRE EFFECTS GUIDE - National Wildfire Coordinating Group

much lower for light, airy fuels such as fine grass, about 12 to 15 percent
(Sneeuwjagt 1974 in Albini 1976), than it is for dense fuel beds such as pine
needles, in which it has been measured at 25 to 30 percent (Rothermel and
Anderson 1966 in Albini 1976). Under favorable burning conditions, the moisture
of extinction has little effect on fire behavior, but when "conditions for burning are
poor, it can cause significant changes in predicted fire behavior" (Rothermel
1983).
(3) Slope. The steepness of slope is measured as the rise of the ground in feet
for every horizontal foot traversed, commonly referred to as "rise over run."
Percent Slope = Rise / Run x 100
Percent slope can be measured directly with instruments or calculated from
topographic maps.
(4) Wind. Both windspeed and direction are used as inputs to the Fire Behavior
Prediction System.
(a) Midflame windspeed. The speed of the wind is measured at the midpoint of
the height of the flames because this best represents the wind that blows directly
on the fire. Most weather forecasts, and most weather measurement stations,
give the windspeed at 20 feet (6 meters) above the ground or above local
obstructions. For fire behavior calculations, the 20-foot windspeed is reduced to
the speed occurring at the midflame height. This compensates for the friction
effect of vegetation and land surface that slows the speed of the wind. The
adjustment factor varies with vegetation type, amount of canopy closure, and
position on slope.
20 Foot Windspeed x Wind Adjustment Factor = Midflame
Windspeed
(b) Effective windspeed. As an intermediate step in obtaining solutions to the
fire spread model, effective windspeed is determined. This value integrates the
additive effects of slope steepness with a wind that is moving across or up a
slope.
c. Outputs of the Fire Spread Model. The accuracy of predictions depends on
how representative the fuel model chosen is of the fuels on the site, how
accurately inputs are measured or estimated, and to what degree the situation
meets the spread model assumptions. For predictions to be within a factor of two
of actual fire behavior (from one-half to two times) is considered to be an
acceptably accurate estimate (Norum 1993). The model is flexible enough that an
experienced practitioner can make fairly good projections of fire behavior by
carefully estimating or measuring the input values and tempering the results with
judgment. Personal experience in a particular fuel type is necessary for refining