United States 

 Department of 

 Agriculture 



Forest Service 



Intermountain 

 Forest and Range 

 Experiment Station 

 Ogden, UT 84401 



Research Paper 

 INT-305 



February 1983 



Predicting Wind- 

 Driven Wild Land 

 Fire Size and Shape 



Hal E. Anderson 



INTRODUCTION 



When an unplanned wild land fire occurs, the fire manage- 

 ment staff needs to know the expected size and shape so the 

 impact on land resources can be assessed and suppression 

 fbrces dispatched. Early attempts to satisfy the need included 

 analysis of fire records by Hornby (1936) that provided a table 

 of minimum, probable, and maximum perimeter for various 

 fire sizes. This table is in the Forest Service Handbook, FSH 

 5109.12, Fireman's Handbook, June 1966, amendment No. 4, 

 figure 2, page 33.5 — 3, and has been reemphasized by Cargill 

 (1970) in terms of forward rate of spread and elapsed time. 

 Most Forest Service regions carry some form of this table in 

 their Fireline Handbook supplements. 



Rothermel's (1972) mathematical model provides the means 

 for predicting how far a fire would travel in a given situation, 

 but in the early 1970's there was no way to use that informa- 

 tion to estimate fire size and shape. However, Fons' in 1940 

 provided data that could be applied to the problem. Fons' 

 work will be referred to throughout this paper. It was possible 

 to develop from Fons' data an approach where only the down- 

 wind spread distance and the windspeed at midflame height 

 were needed to estimate the fire's acreage, perimeter, and 

 shape. ^ This approach is used for estimating fire behavior, 

 utilizing computer facilities for calculations, and evaluating fire 

 hazards of slash (Albini 1976a, 1976b; Puckett and others 

 1979). In addition, the procedure is incorporated into the S-590 

 Fire Behavior Officer course conducted at the National Ad- 

 vanced Resource Technical Center at Marana, Ariz., and the 

 proposed S-390 Fire Behavior training package (see appendbc I). 



Fons, Wallace L. Forest fuels progress report No. 6, May 20, 1940. California 

 Forest and Range Experiment Station, copy on file at the Pacific Southwest Forest 

 and Range Experiment Station, Forest Fire Laboratory, Riverside, CA. 



Anderson, Hal E. Memorandum to R. C. Rothermel and W. C. FiScher, on 

 file at Northern Forest Fire Laboratory, Missoula, MT, August 10, 1973. 



This paper (1) documents the development and formulation 

 of the procedure, (2) reviews observations and methods of 

 assessment, and (3) provides examples that will aid testing the 

 procedure. Illustrations show how this model can be used to 

 confirm other fire behavior models. 



BACKGROUND 



Firefighters and researchers generally agree that wild land 

 fires are circular in shape immediately after ignition, but as 

 wind, slope, and other environmental factors influence the fire, 

 its shape becomes elliptical (Hawley and Stickel 1948; Brown 

 and Davis 1973; Peet 1%7; Pirsko 1961; McArthur 1966; Curry 

 and Fons 1938; Mitchell 1937). McArthur (1966) states that the 

 stronger the wind the more narrow and elongated the fire 

 bums. He presents a relationship between fire shape and wind 

 velocity, using a straightforward ellipse where the length to 

 width ratio varied from 1.0 to approximately 6.0. This range in 

 length to width ratio primarily reflects the grassland fuels being 

 considered. The most probable ratio selected by Cheney and 

 Bary (1%9) is 4:1 for grasslands, while Van Wagner (1%9) uses 

 a ratio of 2: 1 as an example for a forest fu^e. This compares 

 with the average fire shape found by Hornby (1936) where the 

 perimeter is about 1.5 times the perimeter of a circle of equal 

 area. An ellipse with a length to width ratio of 5:1 represents 

 that average fire shape. Work by Peet (1%7) in the western 

 Australia Jarrah Forest indicated the ratio of 2:1, but he noted 

 the fires became more ovoid in shape as rate of forward spread 

 increased. 



Similar observations by Curry and Fons (1938) show the 

 change with a steady wind or a variable direction wind. An ad- 

 ditional display of fire shapes presented by Fons (1946) contrib- 

 ute to using two semiellipses to define size and shape. Mitchell 

 (1937) observes that fires become oval or egg-shaped after a 

 few minutes, with the narrow end being in the direction of for- 

 ward spread. These features of wind-influenced wild land fires 

 become more obvious with the use of infrared imagery for fire 



1 



