Intercomparison of satellite-derived water stress indices in the dynamic representation of the forest fire risk

Resumen

Many probabilistic indices that predict both the risk and severity of forest fires are determined using various types of properly standardized parameters, such as terrain orography, type of fuel, the presence of infrastructures, vegetation cover, humidity conditions and other weather variables, among others, being the temporal variation of some of these variables that accentuate or soften the aforementioned risk of wildfire. This study will compare the use of several vegetation indices, as indicators of water stress status of forests, derived from NOAA-AVHRR and Terra-MODIS sensors data, over the same probabilistic model. The main objective is to analyze the dynamic evolution of these models of risk in the months previous to the most important fire events that have taken place in the Canary Islands (Spain) from 2004 to 2009 and to determine the consistency and degree of representativeness of each vegetation index in the composite value of risk. The adaptation of these dynamic risk models to the special climatic and orographic characteristics of an insular environment will be discussed and the use of them as an aid tool in both prevention and in the simulation of evolution of a fire will also be analyzed.