Estimación del peligro de incendios a partir de teledetección y variables meteorológicas: variación temporal del contenido de humedad del combustible

Abstract

This paper presents the correlations found between field foliage moisture content measurements, meteorological fire danger indices and indices derived from Terra-MODIS satellite images. Both meteorological indices and those derived from MODIS data provide a good estimation of the moisture content of the Mediterranean species sampled. The meteorological subindex of the US system, h1000, as well as the subindex DC (Drought Code) and the final index FWI (Fire Weather Index) of the Canadian system, better estimate fuel moisture content of live fuels. Computed r values between those meteorological indices and gassland moisture content was 0.92, -0.82 y –0.85, respectively. Regarding the dead fuels the best estimations were found for the h1 moisture code (one of the subindices of the US NFDRS system). Concerning those indices derived from MODIS data, the best estimations of moisture content were observed for the Site Moisture (SM), the normalized vegetation index (NDVI) and vegetation index decrements (ARND). For gasslands, the Pearson correlations computed were between 0.91 and 0.98, while for shrub species the correlations were lower (0.79 to 0.89). For oaks (Quercus faginea) the correlations found were not significant with any of the calculated indices.