0738-B3

Comparative analysis of meteorological factors in Russian and foreign fire-danger rating systems

T.M. Sofronova, M.A. Sofronov, P.M. Matveev and A.V. Volokitina ¹

Abstract

Fire danger of vegetation cover depends upon many factors, natural and anthropogenic, therefore almost every country has its own fire-danger rating systems. Mutual improvement of systems is effective when natural conditions in countries are similar (e.g. in Russia and Canada).

Level of forest fire drought is an integral meteorological factor, which produces the greatest impact on daily fire danger value. This level is expressed by different indices, which are usually connected with moisture content of standard vegetation fuels. In Russia and in Canada, standard fuel is similar - it is a green moss cover.

In Russia, the following is taken into account to calculate the forest-fire drought index PV-1 (which is a modified Nesterov's index): PV-1 value for the previous 24 hours, air temperature, and dew point temperature at 3 p.m. as well as day and night gross rain precipitation (from 7 a.m. till 7 a.m.) over 0.5 mm. PV-1 index value is connected with moisture content of a moss layer. There is PV-2 index connected with moisture content of a duff layer. But there is no generalized index for both layers.

In the Canadian Forest Fire Weather Index System (FWI), first of all moisture content codes of each of three standard vegetation fuel layers are calculated for the current day (code FFMC for the upper layer 1.2 cm thick; code DMC - for the middle layer 7 cm thick, and code DC - for the low layer 18 cm thick). The following is also taken into account: value of codes for the previous day, and data on weather in the midday (air temperature and relative air humidity, wind speed, gross rain precipitation for the previous 24 hours). Impact of the three codes and wind speed is synthesized in the FWI index, which reflects burning intensity of standard fuel. Impact of other fire danger factors (vegetation character and state, ignition sources, etc.) is taken into account by means of empirical regional scales in Canada as well as in Russia.

Further modification of the Nesterov's index goes on in Russia: PVG index is proposed, which takes into account hygroscopicity of fuels (like in Canada), rain precipitation duration (like in the United States of America), calculation using weather forecast has been elaborated, technique of regional fire danger scales creation has been unified. Index similar to FWI should be added to the system. Western Europe has become interested in the Canadian index FWI and is testing it.

Introduction

Wildfires are distributed very unevenly in time and space under influence of meteorological and other factors. In economical terms, it is impossible to provide every region with all the necessary means enough to successfully control emergency fire situations and daily carry out air patrol of the whole territory. Therefore, fire danger rating and forecasting are a most important element of every country's forest fire protection.

Fire danger connected with wildfires is a very complicated notion. In fact, it is a whole complex of notions having different scales in time and space and connected with different tasks in wildfire management. Fire danger depends not only on climatic and weather factors but also on the structure of a vegetation floor, its seasonal dynamics, relief of the territory, distribution of ignition sources both in time and space, economic and cultural development of the country, etc. Therefore almost every country develops its own systems of fire danger rating.

Every system usually consists of three sets of factors. The first set is connected with pyrological characteristic of a vegetation floor as an object of burning. The second set unites factors responsible for moistening and drying of vegetation fuels and factors influencing the process of burning. The third set of factors includes ignition sources.

Since there are no ideal systems, it would be useful to compare them in order to borrow some ideas and methods. It should be noted that it is impossible to compare systems on a quantitative level for they have qualitative differences.

Most developed fire danger rating systems are in the US - National Fire Danger Rating System (NFDRS, 1972-1978) and in Canada - Canadian Forest Fire Danger Rating System (CFFDRS, 1971-1984). Apparently, some ideas and elements of these systems can be used to improve a fire danger rating system in Russia. For this purpose, it is necessary to make a comparative analysis of these three systems. This paper gives analysis of the second set of factors, i.e. fire weather danger rating - fire danger rating according to weather conditions in the forest.

Standard vegetation fuels (VF)

Any fire weather danger rating in forests includes assessment of forest-fire drought level in the mode of daily balance of weather factors, which dry and wet vegetation fuels (VF). This balance is directly or indirectly connected with moisture content of standard VF or standard VF complex and with possibility of their burning.

In Russia, the following forest-fire drought indices were elaborated: V.G.Nesterov's "complex index of fire occurrence" (1949) and "moisture indices" PV-1 (connected with moisture content of the whole moss layer) and PV-2 (revealing moisture content of a duff layer) (Vonsky, Zhdanko 1976). Green moss cover 5-8 cm thick in pine stands on drained soils was used as a standard VF complex. Duff was 2.5 - 4.5 cm thick under a moss layer.

In the American system NFDRS, all VF are divided into two large categories: dead and live, since it is considered that live plants are always able to maintain high moisture content, whereas dead VF moisture content depends upon processes of fuel moistening and drying under influence of weather conditions (Deeming, Burgan, Cohen 1977).

Such division is appropriate for the major part of the US where moss and lichen cover is practically absent in forests. Live lichen and moss get dry and moist in a similar way as dead VF since they are unable to keep up their moisture content. In boreal forests of Russia, moss-lichen covers are widely spread, consequently, American VF division is not appropriate for Russian nature conditions.

In the US, dead VF are divided into four classes depending on the value of its time lag (Table 1). Time lag is a period of time, during which there is loss of 2/3 (63%) of the moisture quantity, which can evaporate under standard air conditions from a given VF sample (layer), i.e. 2/3 (M - M_e), where M is VF moisture content at a given moment; M_e is value of equilibrium moisture content of a given VF for standard conditions.

Table 1 - American fuel classification

In NFDRS, standard air conditions are: air temperature - 26.5_C, relative air moisture - 20%. It should be noted that time lag is a constant value for a given sample (VF layer) and does not depend upon its moisture content. This accounts for by the fact that drying rate directly depends upon the difference (M - M_e). In the process of drying, this difference decreases and correspondingly, drying rate decreases.

It should be noted that litter and duff (peat) layers drying rate depends upon depth of their disposition but also upon regime of their soil moistening. On weakly drained sites, duff drying is more retarded than drying on well-drained sites. The structure (loose or compact) and disposition of VF layers produces a great impact on drying rate. In the American VF classification, moss and lichen layers are not mentioned at all.

In the US, VF of the second class in the mode of wood bars having definite size and quality have been long since used as a standard VF.

In Canada, like in Russia, boreal forests with moss cover predominate, therefore a standard VF in Canada is similar to Russian VF and differs only by an enlarged duff thickness. Standard VF complex in the system CFFDRS (subsystem FWI) is a surface cover of green pinnate moss together with pine needle litter and duff under it. VF are divided into three layers (Table 2).

Table 2 - Standard VF complex in the Canadian system FWI

The indicated moss cover is quite typical of high density forests of Pinus banksiana Lamb. and Pinus contorta Dougl. widely spread in Canada. Thick low layer of duff is met in pine stands on the Vancouver Island and in dark coniferous forests of Petawawa (Van Wagner 1987).

Meteorological factors used and scheme of fire weather danger rating

Earlier V.G.Nesterov's "complex index of fire occurrence" was applied in Russia for fire weather danger rating, at present - "moisture indices" PV-1 and PV-2. The latter two indices are characterized by a more differentiated taking into account of rain precipitation (Table 3).

In the American system NFDRS, first of all moisture content of each VF class is estimated using meteorological factors (Table 3).

Since first class VF (dried out grass, thin litter, etc.) can be mixed with green grass, load percent of the latter is taken into account and moisture content of this mixture - "fine fuel" moisture (code FFM) is calculated (taking into consideration code 1-Hr TL FM). This very code FFM is used instead of code 1-Hr TL FM in calculations.

Then three components are calculated.

• Ignition component (IC) is determined by tables taking into account code FFM, air temperature and cloud conditions.
• Spread component (SC) reveals fire front spread rate. It is estimated for each "fuel model" by tables taking into account wind speed, surface slope, code FFM, and tree canopy state.
• Energy release component (ERC) is estimated by tables depending on codes FFM, 10-Hr TL FM, 100-Hr TL FM, and on presence of these fuel classes in "fuel models".

Using components SC and ERC, index BI (Burning Index) revealing frontal edge burning intensity is determined for each "fuel model" (Deeming, Burgan, Cohen 1977).

By present time in the US, two sets of "styled", i.e. standard models of VF complexes have been elaborated. NFDRS now includes 20 models. Northern laboratory of forest fires in the state of Montana elaborated a bit different set of 13 models for the system of fire behaviour forecast "BEHAVE".

In terms of fuel, models differ by distribution of VF load into classes, height of VF layer, and critical moisture content; active fuel load is taken into account in these models. In biotic terms, both sets of models are divided into four groups: grass, bush, tree plots, and cuttings (Anderson 1982; Rothermel 1983).

Canadian general system CFFDRS includes a number of subsystems. One of them is Canadian forest fire weather index system (FWI).

Table 3 - Meteorological factors taken into account in calculation of moisture content indices (codes) in fire danger rating systems of Russia, USA and Canada

Fire weather danger rating in the Canadian system FWI is carried out in the following way. At first, moisture content codes for each of three standard VF layers are calculated for the current day (FFMC - fine fuel moisture code, DMC - duff moisture code, and DC - drought code) using data on weather and yesterday's codes (Table 3). Then on the basis of FFMC and wind speed, initial spread index (ISI) is determined, and on the basis of DMC and DC - buildup index (BUI) revealing VF load consumed per unit of area. Using these two indices, the final FWI is found. This index reveals burning intensity of frontal edge of a fire, which spreads over a plot covered with standard VF. This very index FWI is used in fire danger rating in forests (Canadian Forestry Service 1987; Stocks et al. 1989).

FFMC is in reverse dependence from midday moisture content (M) of the upper layer (if FFMC is equal to 0, M constitutes 250%, if 25 - 132, 50 - 68, 75 - 27, 80 - 18, 85 - 12, 90 - 9, 95 - 5, 99 - 2%).

To calculate FFMC, i.e. upper VF layer moisture, midday data are taken at 12 a.m. or 1 p.m., though creators of FWI are aware of the fact that extreme values can be at 3 - 4 p.m. All codes and indices in the system FWI are determined with the help of tables or a computer program. In spring, the code is calculated on the third day after snow has melted or after midday temperature of 12°C (Van Wagner 1977, 1987).

Some doubts arise as to the usage of wind speed in calculation of FFMC: according to calculation tables, strong wind can increase the FFMC value 1.2 times. But it is known that wind speed near the surface cover in the forest is extremely low. However, in the open places in sunny weather, there is an active mixing of the near-surface air layer due to convection. Neither in the US, nor in Russia, wind is taken into account in evaluation of VF moisture content. But then in the US, cloud conditions are taken into consideration since solar radiation is an important factor of VF drying.

In calculation of FFMC the following regularity is used: the more rain precipitation falls, the less part of it is held by VF layers. It is also taken into account that moisture content of a dried-up layer can increase again in case of relative air moisture increase, i.e. without rain (Canadian Forestry Service 1987).

Moisture content (M) has an expressed diurnal cyclic recurrence. This causes diurnal cyclic recurrence of fires spread rate. Therefore, when forecasting fire behaviour, it is necessary to forecast M (to be more exact, FFMC) for any time of day and night. In 1972, Van Wagner made a table of standard FFMC dynamics in the interval from 12 a.m. till 8 p.m. of a given day depending on the FFMC value at a standard time (Van Wagner 1972). In 1977, he elaborated a computer program to calculate hourly FFMC, using initial FFMC and hourly (including forecasted) data on air temperature, relative air moisture, wind speed, and rain precipitation (Van Wagner 1975). This program is of great interest for Russian fire ecologists as well.

BUI (Buildup Index) is connected with VF load in the middle and lower layers being usually consumed in a flame regime (this load is added to the load of an upper layer). BUI is determined by DMC and DC. Its value depends mainly on DMC, since a lower layer practically does not participate in flame burning. Even complete drying of duff from a lower layer increases BUI less than twice.

ISI (Initial Spread Index) depends upon FFMC value and wind speed. According to tables, a weak possibility of burning occurs at FFMC equal to 40-50 units that corresponds to 60-75% moisture content. Stable spread of combustion is observed beginning with FFMC equal to 78-79 units (25% moisture content), and increase of FFMC by 4 units doubles ISI value. If wind speed increases by 4 m/s, ISI value doubles as well. It should be noted that wind speed is meant not under forest canopy near the fire edge but at the altitude of 10 m in the open place.

Final FWI reveals intensity of a frontal fire edge (standard VF complex being consumed at day time under current weather and drought conditions). FWI is determined using BUI and ISI. It is approximately in direct dependence from ISI and doubles if BUI increases 5 times (Canadian Forestry Service 1987). It should be noted that FWI is connected with only potential current fire danger rating since probability of fire sources occurrence is not taken into account in this index.

Conclusion

Forest fire danger rating systems usually correspond to country's nature conditions. Boreal forests of Russia and Canada bear great resemblance, therefore their fire danger rating systems, especially fire weather danger rating, have much in common. Canadian system is more developed: FWI successfully integrates layer-after-layer moisture differences, takes into account wind as the most important factor of combustion spread and intensity rate; it facilitates assessment of combustion daily dynamics. In our opinion, ideas and elements of the Canadian Forest Fire Weather Index System can be used to improve Russian forest fire danger rating system.

References

Anderson H.E., 1982. Aids to determining fuel models for estimating fire behaviour. - Ogden. - General Technical Report INT-122. - 22 p.

Canadian Forestry Service, 1987. Tables for Canadian Forest Fire Weather Index System. For. Techn. Rep. 25., Ottawa. - 49 p.

Deeming J.E., Burgan R.E., Cohen J.D., 1977. The national fire-danger rating system - 1978. USDA Forest Service. General Technical Report. Int-39. Ogden, Utah. - 66 p.

Nesterov V.G., 1949. Forest fire occurrence and methods of its determination. Moscow, Goslesbumizdat. - 76 p. (In Russian)

Rothermel R.C., 1983. How to predict the spread and intensity of forest and range fires. - Ogden: USDA, Forest Service Inter-Maintain forest and range experiment Station. - UT 84401.- General technical report. - INT-143. - 161 p.

Stocks B.J., Lawson B.D., Alexander M.E., Van Wagner C.E., McAlpine R.S., Lynhem T.J., Dube D.E., 1989. The Canadian Forest Fire Danger Rating System: an overview.// The Forestry Chronicle, vol. 65, NO 6.- p. 450-457.

Van Wagner C.E., 1972. A table of diurnal variation in the Fine Fuel Moisture Code. Environ. Can., Can. For. Serv., Petawawa For. Exp. Stn., Inf. Rep. PS-X-38. Chalk River, Ontario.- 8 p.

Van Wagner C.E., 1975. A Comparison of the Canadian and American Forest Fire Danger Rating Systems. Petawawa Forest Experiment Station. Inf.Rep. PS-X-59. Chalk River, Ontario.- 22 p.

Van Wagner C.E., 1977. A method of computing fine fuel moisture content throughout the diurnal cycle. Fish. Environ. Can., Can. For. Serv., Petawawa For. Exp. Stn., Inf. Rep. PS-X-69.Chalk River, Ontario.- 15 p.

Van Wagner C.E., 1987. Development and structure of the Canadian Forest Fire Weather Index System. Can. For. Serv., Petawawa Nat. For. Inst., For. Techn. Rep. 35., Chalk River, Ontario.-37 p.

Vonsky S.M., Zhdanko V.A., 1976. Principles for elaboration of forest fire danger meteorological indices. - Leningrad, LenNIILH. - 48 p. (In Russian)

¹ Siberian State University of Technology, Prospekt Mira, 82, Krasnoyarsk, 660049, Russia. [email protected]