On the potential of Robust Satellite Techniques (RST-FIRES) for forest fire detection and monitoring

Satellite remote sensing has become an important tool for fire detection and monitoring, thanks to several satellite platforms orbiting around Earth providing repetitive information at a global scale and generally at low cost. Several algorithms have been proposed in order to detect forest fires from space. Multi-channel threshold algorithms use the MIR channel (around 3.5-4.0mm) of satellite sensors like AVHRR (Advanced Very High Resolution Radiometer) in order to identify potential fires, the TIR channel (around 11 μm) in order to remove possible clouds, and spectral difference temperature T=TMIR-TTIR in order to isolate fires from background. Contextual algorithms use, instead, initial thresholds in order to identify potential fires, and compute the spatial average and standard deviation of the spectral difference TMIR-TTIR for background pixels , in order to confirm or reject fires. These methods perform well under specific obervational conditons, but generally show several limitations. In this paper an original multi-temporal approach for forest fire detection and monitoring, named RST-FIRES, will be tested in different fire regimes (winter/summer fires), in comparison with traditional satellite methods. Moreover, its potential in timely detecting the beginning of fires, using data provided by geostationary satellites like MSG-SEVIRI (Meteosat Second Generation-Spinning Enhanced Visible and Infrared Imager), will also be assessed and discussed.