New tools for forest fire fighting and forest ecosystems recovery

Trabajo presentado en las Jornadas de la Red Temática Efectos de los incendios forestales sobre los suelos (FUEGORED), celebradas en Sevilla-Cortegana (España), del 4 al 6 de noviembre de 2009The N.W. of Spain together with the North of Portugal are the European zones most affected by forest wildfires and they are within the worldwide areas with the most number of fires per ha and inhabitant (Carballas, 2007). In these and other countries wildfires are the main factor of forest ecosystems degradation mainly due to post-fire erosion and consequently they produce enormous ecological, social and financial damages. Therefore measures for fire prevention, fire fighting, post-fire soil protection and ecosystem recovery have to be taken. Modelling and knowledge-based systems are important tools for forest fire fighting which were successfully explored by our group (THOR GROUP) with this aim.We have first developed an Expert System, which comprises three computer tools: a) a Forest Fire Risk Index for fire prevention, that predict four days in advance (or more if reliable forecasts are available) the areas with risk of fire (four levels: low, medium, high and extreme risk) with 89% of fire estimation and less than 15% of overestimation; b) a System for the Management of the Resources for Fire Extinction, which gives the difficulty of fire control based on an algorithm that predicts fire propagation (rate and direction), establishes the fire level and produces a specific Plan for fire extinction; and c) a System for the Management of the Burnt Soils, which after analysing the fire damages on soil and plants, gives a plan for the immediate protection of the burnt soil and for the ecosystem recovery (Betanzos et al., 2003). Then, we has built: d) a Mathematical Model for the Prediction of the Forest Fires Peaks, which predicts three months in advance, with an error of no more than 3 days, the temporal period and the initiation date of the winter and the summer peaks, that usually are registered in Galicia (N.W. of Spain) and other regions, which is very important for the temporal contract of very expensive aerial resources; e) a Forest Fire Risk Index for Asturias (N of Spain); f) a Forest Fire Risk Index for Cantabria (N of Spain); and g) a Forest fire Risk Index based on satellite images (Sánchez et al., 2008).The algorithms for the different systems were elaborated by using the historic forest fire and meteorological databases for more than 30 years, taking also into account structural characteristics. The models, as well as the corresponding GIS, initially made for the Galicia Community, were or can be adapted to other regions or countries having forest fire data registered for at least 10 years.Peer reviewe