The Use of Simple Fire Danger Rating Systems as a Tool for Early Warning in Forestry
(IFFN No. 23 – December 2000, p. 32-37)
More and more foresters realize the importance of developing methodologies for monitoring and predicting fuel conditions in the forest to determine the fire danger in a given area. The intense forest fires raging in the province of East Kalimantan in 1997-98, as well as in Brazil drastically showed that the element fire has to be taken into account for management and conservation of forest resources in the tropics. Many of these forest fires could have been prevented if an effective fire management system had been in place at that time. One important aspect of a fire management system is the integration of early warning to ensure that the organization in charge is prepared for possible upcoming fire calamities. Several countries like Canada, Australia and U.S.A. have developed highly sophisticated Forest Fire Danger Rating Systems. In the setting of developing countries these systems are often very difficult to implement, since they are based on a lot of meteorological data and need complicated calculations. The example of East Kalimantan shows that weather stations and equipment like that which is standard in the countries mentioned above simply does not exist and for the near future will not be operationally in use. This paper intends to highlight very simple and inexpensive methods to determine fire danger and to give assistance in setting up such a system.
The Keetch Byram Drought Index
The Keetch-Byram Drought Index (KBDI) (Keetch and Byram 1968) expresses drought as an index on a scale from 0 to 2000, based on the moisture content of the soil. Zero is the point of no moisture deficiency and 2000 is the maximum drought level possible. For almost 5 years the Integrated Forest Fire Management (IFFM) Project has used the KBDI for Fire Danger Rating in East Kalimantan on an operational basis. In 1995 the index was modified and adapted to the conditions in East Kalimantan (Deeming 1995). The computation for deriving the Index is done in a simple spreadsheet by the staff of IFFM on a daily basis.
The major advantage of the KBDI is that only three variables are required to compute the Drought Index:
Mean annual rainfall of a station,
Today’s maximum temperature, and
The Drought Factor equation, which has been slightly modified, is now used for the calculation in East Kalimantan:
where Tmax is the daily maximum temperature and AnnRain is the mean annual rainfall for the area.
The KBDI itself of a given day is the sum of yesterdays rating reduced by 10 times rainfall added to today’s Drought Factor (DF). The Fire Danger, which is expressed through the KBDI, can range from 0 to 2000. To start calculating the KBDI for a given region, one has to go back to a period when the KBDI dropped to “0”, meaning the soil was saturated by water. Keetch and Byram (1968) indicate that point as the day after a rainy period with 150 to 200 mm rainfall within one week.
The index was originally divided into three fire danger classes, for practical reasons and with the focus on the potential end user concessionaires the fire danger rating class “extreme” will be added to the classes:
Tab.1. Fire Danger Rating Classes
IFFM is currently integrating this information into a GIS to evaluate the various fire danger rating conditions for parts of the province.
The Nesterov Index
Another simple Fire Danger Rating Index was developed by Nesterov (1949). This index is used with slight modifications in Russia as well as in other European countries. The Nesterov Index is based on the following parameters:
Days without rain
Dry bulb temperature
Dew Point temperature (calculated from relative humidity and temperature)
N = Nesterov Index
W = number of days since the last rainfall > 3mm
t = temperature ° C
D = dew point temperature ° C
The index requires daily observations of temperature, dew point temperature and precipitation. The difference between daily temperature and dew point temperature is multiplied by temperature and cumulatively added over the days since the last rainfall. Thus the index increases each day until a rainfall of more than 3 mm occurs, at which the index drops back to zero and the process begins again. The system is divided into the following fire danger levels:
Nil 0-300 Moderate 301-1000 High 1001-4000 Extreme 4001 +
Comparison of the two Indexes during the extreme fire season 1997
The Nesterov Index and the KBDI were tested in the 1997 fire season in East Kalimantan. This year serves as a useful example, since it was characterized by wet conditions at the beginning of the year. In February-March the drought started followed by a small amount of precipitation. Between June and October almost no rain was recorded. In this period of high fire risk most of the forest fires were recorded.
Fig.1. A comparison of the performance of the fire danger ratingsystems KBDI and Nesterov (note different scaling)
The comparison between the two indexes shows several interesting aspects of these differing fire danger ratings methods. Since the KBDI is a drought index based on the possibility of the soil to hold water, it is limited by the field capacity. The KBDI system in East Kalimantan is based on the assumption of a 200 mm field capacity. The field capacity multiplied by 10 is hence the upper limit (2000) of the rating. The Nesterov Index does not have this limitation and so has no upper limit, this causes the extraordinary value of 18000 by the end of September, showing extreme fire danger. Both systems follow the development in the same manner, increasing steadily with no rain, and falling down when rain occurs. The Nesterov Index, by definition, falls down to zero if rain occurs. This is a clear limitation of the Index since it assumes no fire danger on a day with more than 3 mm precipitation. For the tropics this 3 mm is much to small, since 3 mm are by far not sufficient to saturate vegetation and duff with moisture, this figure has to be increased to a higher amount. Another limitation of the Nesterov Index is the decrease of the index to zero, since this only describes the situation where rain occurs. In the tropical region there is a high variability in rainfall, the variation is that many rainstorms are very local with tracks only 5 to 10 km wide, so the assumption that fire danger drops to zero for a larger area is not appropriate for fire management purposes.
When comparing the number of days within each Fire Danger Rating Class, the similarities of the two systems are obvious; both systems measure approximately 50% of the days in the high fire danger rating class. The difference in the lower classes results from the sharp drop to zero of the Nesterov Index, when rain with more than 3 mm occurs, while the KBDI drops only slightly, often staying in the same class.
Keetch Byram Index Nesterov Index
Fig. 2. Amount of days in the three Fire Danger Rating Classes according to the KBDI and the Nesterov Index in 1997
The comparison of the two indexes shows that they are both useful tools for early warning. The simplicity of the calculations and the few requirements to measure the input weather data makes both these formulas effective and practicable measures, especially in circumstances where the budget and trained staff is missing. These simple indexes can be calculated by the forest industry as well as the official forestry agencies, requiring only a simple weather station and a spreadsheet programme. The KBDI generally gives a more realistic overview of the fire danger situation due to the only slight decrease if rain occurs, while the Nesterov Index shows the increased fire risk in periods of extreme drought more dramatically. Since the costs for a simple weather station, which can measure relative humidity, rain and temperature in the case of the Nesterov Index, and a weather station that can measure rainfall and maximum temperature for the KBDI are rather small, this methodology is most suitable to the specific circumstances of forest protection in developing countries. An extensive network of simple weather stations that provide data for either index has to be preferred over complex high sophisticated fire danger rating systems like those that exist in western countries like America or Canada.
Crucial for effective fire management are the measures taken with this information, data dissemination methods and Standard Operating Systems have to be in place based on the early warning information, to ensure protection of the valuable forest resources.
The spread sheet used to calculate both the formulas and further information on Fire Danger Rating can be obtained from the Integrated Forest Fire Management Project, Samarinda, East Kalimantan, Indonesia. Contact the author at email@example.com
Chandler, C. 1983. Fire in forestry. Volume 1. John Wiley and Sons, Inc. 449p.
Deeming, J.E. 1995. Development of a Fire Danger Rating System for East Kalimantan. IFFM Report. Contract. No. 1-60134345
Keetch, J.J., and G.A. Byram. 1968. Drought index for forest fire control, USDA Research Paper SE-38. Southeastern Forest Experiment Station
Nesterov, V.G. 1949. Combustibility of the forest and methods for its determination, USSR State Industry Press, 1949. <in Russian>
Georg Buchholz and Doris Weidemann
Integrated Forest Fire Management Project IFFM/gtz
Komp. Perkantoran Dinas Kehutanan Tk. I Kaltim
Jln. Kesuma Bangsa / Harmonika
Kotak Pos 1202