There is one institute of which I know that they have a longstanding research activity in wildfires, and that's the JRC (European Joint Research Centre) in Ispra (Italy)

An interesting URL to have a look at is:

http://bioval.jrc.ec.europa.eu/products/fire_probability_82-99/global-prob_82-99.php

where you can find global fire probablility maps, which are produced since quite some time now (1982 - ). Some of the results of the JRC show a low probability of fire occurring in the first trimester of the year (winter) in the Mediterranean area and also at the high latitudes in boreal regions. The probability strongly increases during the second and third periods of the year.

Time series from 1982 till 1999 clearly depict a temporal shift (of about two trimesters) between fire activity in the northern and the southern hemispheres. This temporal shift between the tropical and medium latitudes in the northern hemisphere is around one trimester.

The fire season in some regions is very stable for a 17 year record (1982 - 1999) compiled by the JRC. These areas have either a very high or a very low probability during any particular trimester. This is the case for the high and medium latitudes in the Northern Hemisphere, the Far East of China-Russia, Central Africa and the African Savannahs. The hypothesis can be put forward that in these areas, the inter-annual variations in climate has little or no effect on fire seasonality – though of course it may still influence the size, intensity and efficiency of a (wild)fire.

In other regions, even when if there is always a maximal probability in a given season, the probability peak is less evident, i.e. the timing of the fire activity across the 17 years studied is very variable. A temporal more homogeneous probability can be interpreted such that particularly fire prone areas/ecosystems are more sensitive to inter-annual climate variability and therfore temporal shifts in fire activity are prominently occuring. This is typically the case in Indonesia, the Mediterranean Basin, Southern and East Africa, California, Australia, Southern East Asia (Indonesia) and Central and Latin America.

Hence, it seems that climate cycles are linked to fire regimes in geographically diverse ecosystems, hence climate–fire relationships may be widely applicable to the management and restoration of fire-dependent ecosystems worldwide.

Therefore the key to answer of the question posed by F-N Robinne, seems to boil down to establishing the properties of climate - fire relationships for the most important global ecosystems, which are still in a more ore less natural state, and not too vulnerable to (speculative) anthropogenic fire ignition.

I must agree that these ecosystems are not so frequently encountered any more.

Cheers,

Frank

Hi Frank,

Thanks for this answer. So if I'm correct, the way you would explore this question goes through the research of the relationship between vegetation-climate-fire ?

Actually, I don't really focus on natural/human fire cycle, as many vegetated ecosystems are fire-dependent as a result of human activity (typically in the Mediterranean basin). Instead, I'm looking for their dependency on fire, whatever their natural state.

From your answer, a potential way to approximate the answer could be the creation of a ratio of current vegetated ecosystems actually covered by fire probability maps. It would be a first step, but it does not fully answer the question of dependency on fire.

Dear colleague I think that in ecology, a disturbance is a temporary change in average environmental conditions that causes a pronounced change in an ecosystem. Outside disturbance forces often act quickly and with great effect, sometimes resulting in the removal of large amounts of biomass.

On the other hands, Ecological disturbances include fires, flooding, windstorm, insect outbreaks, as well as anthropogenic disturbances such as trampling, forest clearing, and the introduction of exotic species. Disturbances can have profound immediate effects on ecosystems and can, accordingly, greatly alter the natural community. Because of these and the impacts on populations, these effects can continue for an extended period of time.

Hi, South Africa holds one of the world's centres of endemicity namely The Cape Floral region. In this region fire is an important NATURAL ecological factor. This region has hot and dry summers and is often on high fire alert and with the natural heavy dry biomass load, it can burn for prolonged periods. Fire is important in this region as some Protea plant seeds do not germinate without the chemical stimuli from ash and other plants in this region flower after fires. There is a large seed stock on the ground that waits for fire to germinate. Because of the measures put in place by man to prevent fire, some plants seem to go "extinct" till after a fire where ground seed stock is germinated. I would say in areas where natural fires a prevalent, plants are dependent on the fire. If you have time look at the South African weather forecast reports around the cape region and see fire warnings, also look into plant seeds that do not germinate unless placed in ash.

Forest Fires are influenced by good numer of factors. If one has to allocate or get prepared to lose certain percentage of area through fires, the best model is obtained through statistical analysis of global incidence of fires say for the last few decades. Most of the natural disasters are amenable for statistical analysis with fair amount of confidence levels. The question is what would be the most likely proportion of vegetated ecosystems in the world dependent on fire. We can just modify this question by considering the population of species instead of the "percentage of the area".

Okey permit me to view this from two factors viz., density independent factor and density dependent factor.

First consider the density-independent factors such as a forest fires which destroy the natural ecosystem. The issue here is how far the species survive . It can be stated that it would be the same proportion of individuals in a population regardless of population density.

In contrast, when we consider the effects of density-dependent factors intensify as the population increases in size. These types of factors, or for that matter the most likely proportion of forest ecosystem ,are a direct result of a species' population size.

It may be noted here, though not directly related to the issue under discussion, that these factors arise because the populations of the species have to compete for the same resources in a limited geographical extent.

Let me come back to the first answer, I mentioned above. As such a forecasting model based on statistical analysis of past events is least complicated approach according to me

I want to add some more information namely about forest fire DSS at NRSC /ISRO/India-

The Decision Support Center (DSC) is established at National Remote Sensing Centre (NRSC) as part of Disaster Management Support Programme of Department of Space (DOS), for working towards effective management of disasters in India. Considering the importance of forest fire management in India, a comprehensive Indian Forest Fire Response and Assessment System (INFFRAS) is invoked under DSC activities of NRSC, which integrates multi-sensor satellite data and ground data through spatially and temporally explicit GIS analysis frame work.

Synergy of Indian Remote sensing Satellite (IRS) systems covering IRS-P6, IRS-1D, and data given by TERRA/AQUA Moderate resolution Imaging Spectroradiometer (MODIS), National Oceanic and Aeronautic Administration - Advanced Very High Resolution Radiometer (NOAA-AVHRR), Defense Meteorological Satellite Program-Operational Line scan System (DMSP-OLS), Environment Satellite (ENVISAT) are useful in forest fire detection, active fire progression monitoring, near real time damage assessment, and mitigation planning. For more details contact .