Instantaneous and Automatic Identification of Weather Types for a Forest Fire Decision Support System

Abstract

The objective of this study, carried out as part of a European Union Contract, was to elaborate an aid package that enables automatic and 'real time' identification of the instantaneous weather situation during seasons with high forest fire hazards. The weather types are defined on the basis of discrimination between the breeze and synoptic wind regimes established from a network of 21 automatic sensors installed in the southeastern extremity of France. The establishment of rules such as: if at station X the wind is from a direction of 100-140° and blows at 2 to 5 m/s; and if, etc..., then the situation is such — enables, in a logic of artificial intelligence, the proposal of a 'symbolical', 'analog' model of the weather behaviours to be expected over a given area.