Previous Page Table of Contents Next Page

Forest fire danger measurement in the United States

By A.W. LINDENMUTH, JR, and RALPH M. NELSON, Division of Fire Research, U.S. Forest Service, Southeastern Forest Experiment Station. Asheville. N.C.

In the United States an average of more than 182,000 fires burn every year, almost 300 fires per million acres (404,690 ha.). Almost two-thirds of these fires, a concentration of about 700 per million acres (1,730 per million ha.), occur in the South, the region expected some day to produce half of the nation's total timber supply, while the Pacific Northwest states of Oregon and Washington, where almost 40 percent of the country's present saw timber is standing, only 3,815 fires burn annually.

Fires range in intensity from innocuous and easily suppressed ground fires to holocausts that defy control. As examples of the latter type, the Miramichi fire of 1825 in Maine and New Brunswick is believed to have swept through 3 million acres (1,214,100 ha.) and taken the lives of 160 people. The Peshtigo fire of 1871 in Wisconsin burned more than a million acres and caused the death of 1,500 people. In 1933, the Tillamook fire raged through one of Oregon's finest timber stands killing 10,000 million board feet (45,300,000 m³) of timber, only 4,000 million feet (18,120,000 m³) less than the total U. S. lumber cut for that year. More recently, in 1947, fires burned 240,000 acres (97,128 ha.) in Maine, 800 homes were destroyed and countless people suffered untold hardship. Fortunately, such fires occur infrequently, but the fact that they can occur emphasizes the need for well-trained and organized protection forces.

Because of the danger, foresters, land managers landowners, and others charged with the administration or protection of forest lands have for many years been seriously concerned with the control of fires either by prevention or by prompt suppression.

This concern has added great impetus to the study and development of techniques and equipment for controlling fires. The fact that the average size of fire has been reduced from 100 acres (40.47 ha.) (5-year average, 1926-1930) to 37 acres (14.97 ha.) (5-year average, 1945-1949), and that five times as many fires per million acres (404,700 ha.) burn on lands not under organized protection as on lands protected by public and private agencies, is proof of the worth of well-developed techniques. No small part of these gains, in all sections of the country, have been accomplished through the use of fire danger measurement systems. As a basis for these systems, stations are located in the forests at strategic points, where fuel moisture, wind velocity, rainfall, and other variables affecting burning conditions are measured with instruments. The knowledge obtained from daily measurement of variable factors directly affecting the flammability of forest fuels, collectively called fire danger, at hundreds of fire danger stations - 300 in the South alone - has been of great value in holding fires to a small size, and to some degree in preventing man-caused fires.

Early Research in Fire Danger Measurement

Early fire fighters undoubtedly were aware that under certain conditions of fuel and weather fires burn with greater or lesser speeds and severity. Rules of thumb, such as estimating wind velocity from the movement of trees or smoke drift, and fuel moisture by the extent to which pine needles would bend without breaking, or the ease with which hardwood leaves would crush in the hand, were useful to an experienced man in an area familiar to him. However, he could not use them with any precision to judge the wide range of conditions that existed between the times when forests were safe from fire or highly flammable.

Recognizing the need for better methods of measuring fire danger, H.T. Gisborne of the Northern Rocky Mountain Forest and Range Experiment Station in the early 1920's began a study of the variables that influence the severity of burning conditions. He found four highly significant factors: (1) season of the year - whether the flammable vegetation was green, in a transition stage, or cured; (2) fuel moisture of heavier fuels as measured by one-half inch sticks; (3) relative humidity, which was a measure of the flammability of finer fluels; and (4) wind velocity. He integrated these variables in a slide rule type of danger meter which for the first time enabled a fire-control officer to obtain a numerical index of fire danger. Burning conditions could at last be precisely defined. The device proved so useful that other Experiment Stations developed systems of fire danger measurement to fit conditions peculiar to their regions.

Regional Differences

Differences in forest types, fuels, climate, and topography, in the principal forest regions in the United States have given rise to variations in methods of measuring danger. The character, arrangement, and volume of fuels; the periodicity and amount of rainfall, wind velocities, and temperature; the aspect and elevation, are only some of the factors that had to be taken into account.

From the Rocky Mountains westward, precipitation averages 10 to 20 inches (25.4 to 50.8 cm.), most of which falls in the winter. The forests are coniferous and the fuels are chiefly deep duff, or heavy slash following cutting. Because of arid summers and heavy fuels, the cumulative effects of drought are very significant in rating fire danger. The moisture content of the fuels decreases progressively during extended dry periods and for that reason one-half inch sticks, which do not dry rapidly, are used as a measure of moisture content.

In the other principal forest area, the East, which extends from a few hundred miles west of the Mississippi River - excluding the mid-continental plains area that supports no important forests - to the Atlantic Ocean and the Gulf of Mexico, rainfall from 40 to 60 inches (1,016 to 1,524 mm.) is fairly evenly distributed during all seasons. Prolonged drought periods are relatively infrequent. The forests are a mixture of deciduous and coniferous types. The fuels are generally grasses in the long-leaf pine type of the South, pine needles in the loblolly type of the Southeast, hardwood leaves in the deciduous forests in the mountain types, and spruce and fir duff in the far North. Because of the plentiful rainfall, fires burn chiefly in surface fuels. Therefore, daily variation in the moisture content of the dried grasses or the topmost layer of coniferous or hardwood leaves is the important element of fire danger. One-eighth inch (0.3175 cm.) basswood slats are used to measure the moisture content of these fuels in the Northeast and South; relative humidity is used as an indicator in some parts of the East.

As continuing research permitted a clearer identification of the regional variables that contribute to fire danger, and as fire control officials became more familiar with the many uses of danger measurement, danger meters for integrating the variables were modified from time to time. Instead of being graduated in a few broad classes, as at first, most types now express danger on a scale of zero, or one, to a maximum of 100 units. The highest and lowest points are not standardized, however, and intervening gradations vary among different types. As examples, in the Northern Rocky Mountain sections, the minimum rating of one indicates conditions under which fires will probably not increase in size; a rating of 100 means that small fires are likely to become unmanageable. In parts of the East the lower 50 gradations of the 100-point scale are divided into 6 classes proportional to the ease with which fires will ignite. The upper half of the scale constitutes a single class which represents explosive conditions.

Uses of Fire Danger Measurement

It is possible to correlate danger ratings with a number of fire phenomena such as rate of spread, behavior, occurrence, or damage, provided the local climate is well sampled with weather measuring stations, equipped with accurately calibrated instruments, and operated by competent observers. The number of stations and their location, as well as the elements measured, depend mainly upon the diversity of the climate and the use to be made of the data.

In the West., where terrain and fuels are ordinarily more difficult than in the East, several elements of fire danger are incorporated in an index to judge the number of men who should be dispatched to a going fire. For example, in California a meter has been devised which integrates fuel type, percent of slope, wind velocity, and fuel moisture, and which correlates these variables with a rate-of-fire-spread factor. This in turn can be converted into figures which indicate the number of men required to construct fire lines under given conditions.

In the northern Rocky Mountains, where danger ratings are correlated with fire behavior and where cumulative drying of fuels during the summer is a prime factor of fire danger, ratings are a valuable guide in placing men on active duty in the many fire-control positions. This western country is a land of rugged mountains, and vast areas are very remote. Hence it is impossible to expand or reduce the number of men employed each day as in the East. During the winter, spring, and fall, when the fuels are covered with snow or are wetted as it melts, the average danger rating is about zero and very few of the fire organisation positions are filled. But as the average fire danger increases in the late spring, when the rating ranges from 20 to 30 for several days, then 30 to 40, and so on, more and more men are hired until all men needed to handle the normal amount of fire suppression work are employed. While employed, these men are not necessarily working at all times on fire-control activities. At intervals, showers or cloudy weather may lower the danger rating temporarily, and during such periods men are shifted from fire-control positions to such jobs as repairing equipment, telephone lines, and similar work. They are still however, to assume their fire-control duties whenever needed.

Danger ratings have been correlated with the relative probability of fires igniting in the pop lows Northeast, where 99 percent of the fires are caused by people; thus the scale on the danger meter is graduated accordingly. It has been found that the probability of fires starting depends primarily upon the number of people in the forest and the care with which they handle or discard igneous materials, extinguish campfires, burn debris, or otherwise employ materials capable of igniting forest fuels. Ordinarily the number of residents or visitors in the forest remains fairly constant over long periods. Their habits also remain about the same unless rather drastic action is taken to change them

That close correlation can exist between accurate danger ratings and fire occurrence is indicated by an analysis of records from a section where the number of people and their habits are fairly stable. The area sampled, comprising some 2,300,000 acres (930,810 ha.), is part of an eastern state. More than 3,000 fires that burned during the three year period, 1947-1949, were compared with average danger ratings - as obtained from seven danger stations distributed throughout the area - for all days during the period. The correlation coefficient was 0.94, which means that 88.4 percent of the average daily variation in the number or fires was accounted for by variations in fire danger.

Thus, in the Northeast the meter can be used with considerable confidence in predicting the number of fires that will ignite from day to day. From forecasts made 24 hours in advance by the United States Weather Bureau, estimates of fuel moisture, wind velocity, precipitation, and temperature, are obtained and the expected fire danger calculated. The fire control administrator, knowing the fire danger and expected number of fires, determines whether he should post observers to detect the ignition of fires, how many men he should hire to be ready to suppress fires, and the number of pieces and type of equipment probably needed during the period. This provides a basis for many other important decisions essential to adequate fire control planning. If the estimate of the number of fires is considerable, the public can be warned by radio and through the newspapers.

If daily ratings of relative flammability are available, it follows that seasons or years for a particular field unit, or between units, can be compared. In the Northeast this is done by summing the daily ratings and determining the relation between the ratings and numbers of fires. Because daily ratings are proportional to the expected number of fires, the same relationship will hold for longer periods. The ratio of one variable to the other can be used as a measure of variations in flammability. A comparison of actual versus expected fires can also be used to measure the effectiveness of fire-prevention measures. This is possible because over large units the number of forest users remains relatively constant. Thus, the number of fires is affected chiefly by the care with which users handle or use fire (one of the elements of risk) and by flammability (as measured by danger ratings). A graph of the ratio of expected to actual fires by time established whether fires are on the increase or decrease and thus provides a gauge for the effectiveness of prevention activities.

From this type of comparison it has been determined that the number of fires per unit of fire danger has been reduced about 21 percent in the Northeastern region during the past five years. In several states the reduction has been highly significant. Inasmuch as the population of the region has not decreased, the smaller number of fires can only be accounted for by good prevention programs. Strict enforcement of debris-burning permit laws, and the issuance of warnings during critical periods, both based on the estimated danger, are the measures which are believed to have brought the best results.

The summing of units of burning index for successive days or weeks can be a useful indicator of impending critical situations. This is illustrated by an analysis of the burning conditions which prevailed prior to the period October 21-25 in 1947, when tires caused tremendous damage in Maine. A comparison of records for July, August, and September of 1947, with averages for the years 1943-1946, reveals nothing that might have given fire officials concern. However, by cumulating units of burning indexes by 5-day periods beginning September 1, as is done in the following table, it is readily apparent that early in October there were signs of increasing danger.


Units 1947

Units 1943-1946

September 20...



September 25...



September 30...



October 5...



October 10...



October 15...



October 20...



October 25...



October 31...



The sharp increase in burning index units during the period October 5-15, had the information been available, should have alerted all fire-control agencies and the public. By October 20 it would have been clear that conditions were extremely critical. Methods now in use will detect the buildup of any similar critical period.

In the South, where it has been found that variations in danger account for more than half of the variations in average rate of spread, the chief use of danger ratings is to guide fire-control action. When fast rates of spread are indicated, more observers are used to speed the discovery of fires, larger crews are dispatched rapidly to all fires, mechanical fire-line builders are utilized fully, and such other action taken as will aid in holding fires to a small size.

Besides their direct uses as guides and measures danger ratings are sometimes treated as a single factor among several that are integrated in order to arrive at other indexes. For example, danger ratings are one of the factors recognized in the East when appraising fire damage. It has been determined that the number of trees killed or injured by fire by specific timber type, stand density, and tree sizes, is closely associated with the danger ratings.

Perhaps it should have been emphasized earlier that no mechanical device such as a danger meter can be a substitute for sound judgment in fire control. There are so many variables that influence fire behavior there can be no fixed rules as to how a fire may be controlled. A knowledge of what is likely to occur in certain degrees of danger can, however, avert catastrophes. Also, during critical periods when it becomes necessary to place responsibility on less-experienced men, a fire plan, based on the degree of danger estimated, can be a very valuable aid. Fighting forest fires requires good men armed with good tools. Fire danger measurement is one of these tools.

Henry Solon Graves, 1871-1951

Henry S. Graves

Henry S. Graves, forester and educator, died in Brattleboro, Vermont, U. S. A. on 7 March 1951 following a heart attack. As Chairman of the Technical Committee on Forestry and Primary Products of the United Nations Interim Commission on Food and Agriculture, Mr. Graves, by his vigorous personality objective judgment, and genial nature, gave inspiring leadership to the world's foresters who had the task of formulating the program and Organization of FAO's Forestry Division.

He was born in Marietta, Ohio, on 3 May 1871, and was educated at Phillips Academy, where his father was professor of Physics and Chemistry, he took his A. B. degree at Yale College in 1892. At the urging of Gifford Pinchot, the first American professional forester, he studied at Harvard University and at the University of Munich.

In 1898 Mr. Graves was appointed as Assistant Chief and Superintendent of Working Plans in Mr. Pinchot's Division of Forestry in the U. S. Department of Agriculture. His work on research, public education, and co-operation with private forest owners, was handicapped by the lack of professionally trained foresters. As a result,: Mr. Pinchot's father, James W. Pinchot, and his family, endowed the Yale Forest School, and in 1900 M. Graves was chosen to organize the School and to serve as its Director.

In 1910 Mr. Graves succeeded Mr. Pinchot as Chief Forester of the U. S. Forest Service and served in this position for 10 years.

Commissioned a Major in the Corps of Army Engineers in June 1917 and advanced to Lt. Colonel in two months, he organized the regiment of forest engineers to cut timber in France for the American armies.

In 1922 Mr. Graves returned to Yale School of Forestry as Dean and was appointed as Sterling Professor of Forestry. For a period of 4 years he also held the post of Provost of the University, and served as Chairman of the University Educational Planning Committee.

In 1939, after he had retired from active duty at Yale, he became Chairman of the Joint Committee on Forestry of the National Research Council and the Society of American Foresters, to study postwar problems in forestry.

He played an important part in the formulation of a number of the late President Franklin Roosevelt's conservation projects. Not the least of these was the Civilian Conservation Corps.

He wrote some 170 articles and bulletins as well as two of the earliest American textbooks in forestry - Forest Mensuration and Principles of Handling Woodlands and was the co-author with C. H. Guise, of Forest Education.

Mr. Graves was awarded a number of honorary degrees, including M. A. from Harvard University in 1911, L. L. D. from Lincoln Memorial University in 1922, from Syracuse University in 1923- and from Yale University in 1939. He was awarded the Sir William Schlich Memorial Medal in 1944, for distinguished service to American Forestry. The French Government in 1947 awarded him the rank of Officier du Mérite Agricole in recognition of his services to forestry. The first award of the Gifford Pinchot Medal was made to Mr. Graves in December 1950.

His broad vision and high objectives are well expressed in the 1945 Report of the Technical Committee on Forestry and Primary Forest Products to the United Nations Interim Commission on Food and Agriculture:

The basic objectives of a world forest policy should be:

(1) The adequate consumption of forest products to improve houssing, clothing and living standards in all parts of the world.

(2) The managed use of the world's forests and forest soils for the continuous production of raw materials.

(3) The conservation of all forests performing social or protective functions.

"In due course, governments may wish to consider a formal declaration that recognizes the achievement of these basic objectives as a duty to their peoples, to each other, and to the world."

To be sure there would be no misunderstanding, his report lists six "specific goals":

(1) Application of efficient silviculture and management to all forests in use.

(2) Afforestation on denuded lands with particular regard to Asia.

(3) Maintenance of protection forests and their extension to meet the requirement of different areas.

(4) Development of unexploited forests.

(5) Promoting of integrated, modem forest industries.

(6) Balanced expansion of processing facilities for various primary forest products, of forests.

"The achieving of these goals should be accompanied by progressively better incomes and living conditions of forest workers, operators, and owners. Simultaneously, a relative reduction in the cost of forest products should become possible and facilitate their efficient and economical distribution."

Previous Page Top of Page Next Page