Remote sensing of large wildland fires in the european mediteranean basin.
The general goal of the MEGAFiReS project was to establish an integrated framework for using satellite remote sensing images in wildland fires management. This general objective was split in three aspects, following the main phases of fire planning: pre-fire (risk estimation), during-fire (early detection) and post-fire (effects assessment). Following the three objectives proposed at the beginning of MEGAFiReS, a wide range of remote sensing methods and data have been used in the project. Sensors such as NOAA-AVHRR, Landsat-TM, IRS-LISS, IRS-Wifs, ERS-SAR, and aerial photography have been tested in different territories of the study area. In some cases, integrated analysis between high and low-resolution sensors is proposed, showing the great interest of synergistic approaches. Methods of analysis explore both well known techniques, such as maximum likelihood classification and normalised vegetation indices, and more innovative procedures, like spectral mixture analysis, middle infrared vegetation indices, tasselled cap, logistic regression, and neural networks.