2019 Wildfire Conference. Adressing the Challenges of Bushfire Management
Presentations Notes 2019: DAY 1 (Day 2 - Wednesday Nov. 13th, 2019).
Validation of three historical fires in Corsica.
N. Hamamousse (AMU, UO1, LEPM, CEREN), A. Kaiss, J-P. Clerc (IUSTI, Marseille), F. Giroud, N. Bozabalian (CEREN, Valabre), A. Sahila, N. Zekri (LEPM, Oran)
The high number and the frequency of large scale forest fires make them one of the major natural disasters that disturb the ecosystem and endanger gravely the environment of various geographical regions of our planet. In particular, the wildfire activity in Corsica has greatly increased during the last decades. This phenomenon has attracted the interest of an increasing number of researchers because of its serious impact on natural, economical and human resources.
The topography of the terrain, the meteorological conditions and the type and distribution of the fuels in the forest, are one of the key parameters that influence the propagation of large fires. These parameters can differ greatly from a region to another and from a period to another. Due to these local heterogeneities, wildfires can exhibit a fractal shape (Caldarelli et al., 2001). Forest fires are very complex to model because of the number of intricate factors that can influence their propagation, and their multi-scale nature. Indeed, the characterization of their behaviour depends on the spatial resolution considered. In fact, there are generally three relevant scales, each one corresponds to a specific behaviour of fire, and is characterized by a particular set of physical and chemical phenomena. The microscopic scale (0.1m-10m) corresponds roughly to what sees an observer on the ground; it is the scale at which the heat transfer, thermal degradation and combustion of the vegetation occur. The macroscopic scale (10m-1Km) is typical of observations from an aerial positions; at this altitude, the fire front separates distinctly the burning zone and the unburned fuel.
Finally, the gigascopic scale (1Km) corresponds to the scale of satellite images where the fractal behaviour emerges (Caldarelli et al., 2001).
The small world model was proposed for the first time in 1998 by Watts and Strogatz in order to describe social behaviour (Watts D. J. and Strogatz S. H., 1998). It is characterized by a high clustering coefficient and its ability to include randomly generated long-range connections. The aim of this paper is to validate three different historical fires occurred in Corsica: Suartone (July 28, 2003), Saint Cyprien (September 08, 2002) and Sainte Lucie of Porto-Vecchio (September 28, 2006), by using a hybrid stochastic/deterministic model designed to simulate fires in a heterogeneous landscape.