Forest fire modelling

Abstract

Wildland fires have always been undesired and dangerous events. The danger includes the destruction of a renewable natural resource, damage of the atmospheric environment through the emissions of pollutant gases, which contribute to the greenhouse effect, and threat to the lives of people living in the areas surrounding the place where fire occurs and of the members of firefighting teams. To cope with all this, it is necessary to know the behaviour of the fire in order to be able to make adequate and proper decisions that will assist activities implemented in fire suppression and prevention. In this context, fire behaviour modelling is utilized to determine these characteristics and to simulate fire propagation in a variety of vegetations, under diverse climatic and topographic conditions. In this context, the aim of the thesis work was to study and describe the fire propagation under different fuel features and boundary conditions with particular attention to terrain configurations where fire propagation may be characterized by abrupt variations in intensity and propagation rate. These areas include double-slope domain and canyons where such a phenomena (commonly identified as eruptive behaviour) occurs without any change in the main factors governing the fire propagation (i.e. atmospheric condition, vegetative fuel, domain topography). This work aims to provide an increase of knowledge in the fire spreading, currently still limited, thanks to the adoption of a physically-based code (WFDS) to model flame propagation; in fact, through the numerical resolution of the fundamental balance equation describing the fire phenomenon, it offers a way to analyze the fire behaviour on a scientific basis. To purse this target different aspects, reported in different chapters, were considered in this work. The first part examines shortly the social, economical and ecological impact of wildland fires on a global scale and for the countries of the Euro–Mediterranean region. The role of fire behaviour modelling as a tool for fire fighting activities and prevention management is put in evidence and discussed. The second part endeavours a short review of the main surface fire spread models developed since 1990 with particular attention to physically-based codes. These models are alternative to empirical or quasi-empirical models, which do not have physical basis and are only statistical in nature or make use of some form of physical framework upon which to base the statistical modelling chosen. A comparison between the most frequently used simulation codes is accomplished and the advantages to consider a physical code - for instance WFDS - rather than an empirical or quasi-empirical one is highlighted. [edited by Author]