Design and Rationale of the Global Vegetation Fire Inventory (GFVI)

Design and Rationale of the
Global Vegetation Fire Inventory (GFVI)


Assessments of past, present, and future atmospheric chemistry and the consequences of changes of atmospheric chemistry on the global climate rely in part on inventories of emissions of appropriate chemical species and aerosols constructed on appropriate spatial and temporal scales.

Global biogeochemical cycles and the chemical properties of the atmosphere are highly influenced by emissions from combustion of fossil fuels and of plant biomass.

Whereas the investigation of fossil-fuel burning has been receiving most attention and meanwhile providing reliable information, the inventory of global emissions from free-burning vegetation fires (wildfires, fires in land-use) and other plant biomass burning (burning of fuelwood and charcoal) is still inadequate.

Several global estimates of emissions from vegetation fires (Hao et al. 1990, Weiss 1990, Crutzen and Andreae 1990, Andreae 1993, Goldammer 1993) are still considered as uncertain due to the very weak databases.

The main reason for the weakness of basic statistical data is that in the most important regions and countries no statistical data are collected systematically on the spatial and temporal extent of free burning fires in

  • forests (wildfires, prescribed fires, burning for planned forest conversion)
  • tree, brush, and grass savannas (wildfires, prescribed fires)
  • use of fire in agriculture (burning of harvest residuals)

The main reasons for the lack of reliable data is that the origins of vegetation fires are highly diverse (natural fires, a large variety of reasons of human-caused fires). Their spatial and temporal extent is not known in areas with lacking infrastructures (remote sites with low human population densities, e.g. savanna and boreal forest lands). The use of spaceborne sensors for fire detection, classification and mapping has been used mainly for research purposes and case studies. On a permanent operational base the use of satellites is restricted to few locations (e.g. Brazil).

The systematic collection of fire data through administrations, e.g. forest services, is restricted. While forest fire statistics are systematically collected and published only within the ECE region (Economy Commission for Europe: Member countries are all European countries including the CIS, plus the U.S.A. and Canada; c.f. ECE/FAO 1993), most other statistical data are in the “grey literature” and practically not available for international interpretation. The FAO Global Wildland Fire Statistics (FAO 1992) are a first attempt to collect country reports and to summarize information on wildfires in forests. The findings for the decade 1981-1996 show that a large gap of information exists on reliable statistical coverage.

The existing estimates mentioned before have been reexamined recently and improved through the integration of the vegetation fire component into a High Resolution Biome Model (HRBM) (Mack 1994, Mack et al. 1996).

This model approach provides the temporal and spacial extent of vegetation fires and allows the modelling of all kinds of chemical species emitted by the fires. For instance, the model reveal a global gross flux of carbon emitted from vegetation fires to the atmosphere of 4.14 Pg (Mack 1994) and 4.8 Pg (Mack et al. 1996) respectively (1 Pg = 1 Gigaton = 1 billion tons) per year. This compares with the previous estimates as follows (biofuels and agricultural burnings are not included):


Hao et al. (1990): 2-2.5 Pg (tropics only)

Crutzen and Andreae (1990): 2.6 – 3.3 Pg

Andreae (1993): 2.74 Pg

Weiss (1990) and Goldammer (1993): 4.5 Pg (“upper limit of fire potential”, tropics only)

HRBM Model I (Mack 1994): 4.14 Pg

HRBM Model II (Mack et al. 1996): 4.8 Pg


The obvious discrepancies between the various estimates and the model, and the inherent weaknesses of a model itself, clearly justify the necessity of a new comparative approach.

This first step for a global fire inventory aims to collect a comprehensive set of data on vegetation fires. These are all free-burning fires occurring in all types of vegetation, including fire used in the agricultural sector.

Objectives of the GVFI

With a global fire dataset it will be possible to feed regional and global models, e.g. in atmospheric chemistry, carbon cycle, etc. Such dataset will be an important step to realize the Global Vegetation Fire Information System as proposed by the 1992 Dahlem Conference on “Fire in the Environment” (Crutzen and Goldammer 1993).

In the context of the establishment of the Global Fire Monitoring Center (GFMC) the GVFI will also serve

  • decision makers in forest fire management, fire and land-use policy development
  • development of international science programmes
  • decision-support for international fire emergency responses

Current state of the GVFI

The aim of establishing a Global Vegetation Fire Inventory (GVFI) is supported by various international research activities.

Among various international activities the most actively involved research structures and programmes, which are supporting GVFI, are:

  • The International Global Atmospheric Chemistry Project (IGAC), a core project of the International Geosphere-Biosphere Programme (IGBP) with its operational projects BIBEX (Biomass Burning Experiment) and GEIA (Global Emissions Inventory Activity).
  • The International Boreal Forest Research Association (IBFRA), Fire Working Group
  • The ECE/FAO Team of Specialists on Forest Fire
  • The UN International Decade for Natural Hazard Reduction (IDNDR)

Methodology of Phase I

In the Phase I of the Global Vegetation Fire Inventory it is attempted to collect experience in developing, applying and improving methodologies of data collection and a first set of “raw data”.

Through extensive literature research and a network of regional expertise/correspondents the worldwide occurrence of forest fires (wildfires: uncontrolled/unwanted fires) is investigated. To support the research done by the Fire Research Group of the Max Planck Institute for Chemistry and the Global Fire Monitoring Center, experts from all over the world had been encouraged to collect relevant data within the region of his or her expertise.

In order to utilize the country- and biome-specific expertise of the regional correspondents it is suggested to proceed on a country-by-country base and vegetation type (fire regime type) respectively.

In Phase I a simplified approach is chosen for the GVFI in order to obtain a reasonably quick result for first discussion and further improvement of accuracy. The statistics to compile are forest area burned, fraction per month and fire cause. The emphasis is done only on the land use forest because of problems of availability and consistency for the data set.

The data is obtained from various sources like the ECE/FAO forest statistics, the International Forest Fire News as well as other available sources. The data “fraction of area burned by month” is taken from available literature and is otherwise estimated by experts for the specific country. The third figure collected is the distribution of the fire cause by country (only natural vs. human-caused).

The development of outline and structure of the following Phase II will require a more thorough discussion, based on the results obtained by this first phase of the GVFI.

Phase II: The Vegetation Types to be Investigated

The vegetation types of which the fire data for Phase II will be collected are given in a blank online form.

The preliminary classification of vegetation types included is based on the following consideration:

Current fire statistics usually distinguish between fires which have affected “forest” and sometimes “other land”. Evaluating these two types of fire has a restricted use because only a fraction of global vegetation burning is usually in these statistical data and only restricted information can be derived on the ecological, atmospheric chemical, cultural and economic significance of fires.

The use of a complex vegetation classification with dozens or more vegetation types distinguished, on the other hand, would clearly be too difficult to be handled by the data-collecting authorities.

In Phase II of the GVFI the following categories of vegetation types affected by fire will be collected:


I: Wildfires (uncontrolled/unwanted human-caused and lightning-caused fires)

Forest 1 Intensively managed / utilized and protected (incl. plantations):

All forests which are managed following the principles of sustainable forest management. This includes all forest plantations, even if these are not considered to be “sustainable”. In most cases the fire cause an economic damage (except low-intensity surface fires which may lead to minor damages or are considered beneficial).

Forest 2 Not intensively managed and protected:

All forests which do not belong to category Forest 1. Usually these are forested wildlands in which no forest management or forest protection takes place (e.g. boreal forest outside the fire protection zone, or tropical seasonal (dry, deciduous, semi-deciduous) forests which are regularly burned and adapted to fire. Fires are often not considered to cause economic damages.

Savanna/Shrubland 1 Intensively managed, utilized and protected:

Those ecosystems in which fire creates conflicts with sustainable management of wildlife, animal husbandry, watershed protection, and other natural resources management objectives, usually associated with economic losses and/or environmental degradation. Example: tropical and subtropical savannas, steppe ecosystems of the temperate/subboreal zone.

Savanna/Shrubland 2 Not intensively managed, utilized and protected:

Same or similar vegetation types, but no conflicts or immediate economic or environmental damages caused by wildfires, often due to adaptation to fire (fire climax ecosystems).

Grassland/Pastures Intensively managed and utilized (not in categories savanna 1 or 2):

Grasslands and pastures that are intensively managed grazing resources, usually in property of farms or commercial livestock rangelands.


II. Prescribed fires

This category includes all fire which are set intentionally and following a prescription, regardless whether the prescription follows rules of sustainable vegetation (fire) management or resulting in destruction or degradation on the ecosystem.

Prescribed pastoral & agricultural burning maintenance burning of pastures, agricultural waste disposal (straw of wheat, rice, sugar cane, etc.):

Traditional cyclic (often annual) burning of residues and plant biomass which is not utilized for any production purposes.

Traditional Shifting Cultivation with fallow and forest regeneration phases:

Non-permanent land-clearing following traditional shifting cultivation cycles (slash-and-burn) after which forest cover is allowed to return.

Permanent Land Clearing 1 Conversion of forest for plantation establishment:

Conversion of native primary or secondary vegetation (forest and other wildlands) to prepare land for tree plantations (forest plantations, other tree plantations like oil palm plantations).

Permanent Land Clearing 2 Conversion of forest to agriculture and pasture systems:

Native vegetation is completely removed and replaced by agricultural and pastoral systems.

Other (explain):

Information Sources

Potential sources for the compilation of data for the second phase of the GVFI are different for each vegetation types:

Forest conversion fires: Deforestation rates, e.g. from national or FAO statistics. Burned biomass data from research

Forest wildfires: Statistics, satellite data

Prescribed forest fires: Statistics, research data

Other wildlands (savannas): Vegetation maps. Fire intervals and biomass loads from research

Range management fires: Agricultural statistics (yearbooks). Biomass estimates

Agricultural residues burning: Agricultural statistics, biomass estimates

Methodology and Future Development

The data to be collected shall refer to an annual average. In case of availability of statistical data or other regularly collected records (e.g. remotely sensed fire data) the average given should refer to a recent decade, preferably to the last ten years.

Since the goal of Phase I is to obtain a “quick look” at the global fire scene, it is desirable and helpful to obtain additional and still lacking country reports as soon as possible. Active participation of the scientific community is needed. It was therefore decided to publish the questionnaire on this web page. The questionnaire consist of the layout for GVFI Phase II.

Please note that suggestions are highly welcome. Any scientist or land resource manager who has data useable for the inventory is invited to contribute it to this research.

The inventory sheet is available at the GFMC, please indicate the file format you prefer. After filling the form it can be sent via e-mail to the GFMC address ( or by mail (or fax) to the GFMC postal address. In a later stage the form can be filled on-line.

Please do not forget to state your position and the source of the fire data.

Mailing address:

Global Fire Monitoring Center – GVFI
Fire Ecology Research Group
c/o Freiburg University
PO Box
D-79085 Freiburg

Fax: ++49-761-808012



Andreae, M.O. 1993 The influence of tropical biomass burning on climate and on the atmospheric environment. In: Biogeochemistry of global change: Radiatively active trace gases (R.S.Oremland, ed.), 113-150. Chapman and Hall, News York.

Crutzen, P.J. and M.O. Andreae 1990. Biomass burning in the tropics: Impact on atmospheric chemistry and biogeochemical cycles. Science 250, 1669-1678.

Crutzen, P.J. and J.G.Goldammer (eds.) 1993. Fire in the environment: The ecological, atmospheric, and climatic importance of vegetation fires. Dahlem Workshop Reports. Environmental Sciences Research Report 13. John Wiley & Sons, Chichester, 400 p.

ECE/FAO 1993. Forest fire statistics 1990-1992. ECE/TIM/70, United Nations, New York, 24 p.

FAO 1992. Global wildland fire statistics. FO: MISC/92/4. FAO, Rome, 48 p.

Goldammer, J.G. 1993. Feuer in Waldökosystemen der Tropen und Subtropen. Birkhäuser-Verlag, Basel-Boston, 251 p.

Hao, W.M., M.-H. Liu and P.J. Crutzen 1990. Estimates of annual and regional releases of CO2 and other trace gases to the atmosphere from fires in the tropics, based on the FAO statistics for the period 1975-80. In: Fire in the tropical biota. Ecosystem processes and global challenges (J.G. Goldammer, ed.), 440-462. Ecological Studies 84, Springer-Verlag, Berlin-Heidelberg-New York, 497 p.

Mack, F. 1994. Zur Bedeutung von Vegetationsbränden für den globalen Kohlenstoffkreislauf. Doctoral Dissertation, University of Giessen, May 1994.

Mack, F., J.Hoffstadt, G.Esser, and J.G.Goldammer. 1996. Modelling the influence of vegetation fires on the global carbon cycle. In: Biomass burning and global change. Vol. I (J.S.Levine, ed.), 149-159. The MIT Press, Cambridge, MA.

Weiss, K.F. 1990. Abschätzung der jährlichen Biomasseverbrennung und Kohlenstoffemissionen aus Wald- und Savannenbränden in den Tropen und Subtropen. Unpubl. Diploma Thesis, Fire Ecology Research Group, c/o Inst.Forest Zoology, Univ. Freiburg, 88p.

Back to GVFI


Print Friendly, PDF & Email
WP-Backgrounds Lite by InoPlugs Web Design and Juwelier Schönmann 1010 Wien