Fire Detection and Monitoring (FDeM)

NRT Product Available since Apr 2010


Product Documentation

This operational product is documented in the Algorithm Theoretical Basis Document (ATBD), Product User Manual document (PUM) and the Product Output Format document (POF) The validation results for this product are available in the (VR) document.
Data Policy
The use of LSA SAF products in publications is kindly requested to be duly acknowledged:
FDeM was provided by the EUMETSAT Satellite Application Facility on Land Surface Analysis (LSA SAF; Trigo et al., 2011)

Trigo, I. F., C. C. DaCamara, P. Viterbo, J.-L. Roujean, F. Olesen, C. Barroso, F. Camacho-de Coca, D. Carrer, S. C. Freitas, J. García-Haro, B. Geiger, F. Gellens-Meulenberghs, N. Ghilain, J. Meliá, L. Pessanha, N. Siljamo, and A. Arboleda, 2011: The Satellite Application Facility on Land Surface Analysis. Int. J. Remote Sens., 32, 2725-2744, doi: 10.1080/01431161003743199
Fire Detection and Monitoring (FD&M) explores the capability of SEVIRI/MSG to detect and monitor active fires, particularly over Africa and Europe.
Fires are an important and highly variable source of air pollution emissions in many regions of the world and they constitute a significant factor controlling the interannual variability of the atmospheric composition. Remote sensing data has long been recognized as an effective way to detect fires over large areas. The monitoring of large fires along with a full characterization of fire occurrence at various time scales (from sub-daily to seasonal) is likely to be best served by geostationary satellites than by polar-orbiters, despite the better spatial resolution provided by the latter.
The LSA SAF makes available identification of wild fires for individual SEVIRI/MSG observations, i.e., every a 15-minute.
Product Description
FD&M is based on the algorithm FIDALGO (Fire Detection Algorithm) developed within the LSA SAF (Amraoui et al., 2010) to identify SEVIRI/Meteosat pixels potentially contaminated by fires. The algorithm is applied to 15-minute SEVIRI observation, providing a list of identified fire events, along with a quality flag covering all pixels. This contains information on the scene classification (e.g., cloudy, clear sky land, clear sky water) and a quality indication on the pixel classification.
Detection and systematic monitoring of active fires over the African continent is essential for an accurate assessment of the overall fire activity, namely in protected areas e.g. national parks, reserves and hunting concessions. It also allows the characterization of fire regimes those areas as well as their impacts on the natural habitats and biodiversity. Active fire detection over Europe is in turn essential for early fire warning and for fire prevention, namely in what respects to a proper calibration of risk of fire indices, namely those that integrate the Risk of Fire Mapping (RFM) product currently being developed by the LSA SAF.
Algorithm Description
Depending on whether they are smouldering or flaming, most wild fires burn at temperatures between 500 and 1200 K. Higher temperatures may even be reached in some forested areas. At these temperatures, and according to the Planck law, there is a very strong emission in the middle-infrared, at wavelengths between 3 and 5 µm. The signature of fires, even with scales considerably smaller than one pixel, is then capture by SEVIRI channel centred at 3.9 µm. Emission by fires strongly contrasts with a non-burning background, which presents emission peaks in the thermal infra-red around 10 µm. The algorithm used for FD&M retrieval, FIDALGO, is a contextual algorithm where values of thresholds are dynamically derived using appropriate statistics obtained from the neighbouring pixels. A fire event is then identified within a given pixel, when the contrast between that pixel and its surroundings is high enough.
Data Characteristics
The FD&M product is derived every 15 min at the native SEVIRI pixel resolution. The disseminated product includes a list of pixels where a fire event was identified and also a matrix with a quality indicator by pixel.
Product Uncertainties
The FD&M is assessed via comparisons to near-simultaneous results derived from MODIS.
Amraoui, M., C.C. DaCamara, J.M.C. Pereira, (2010). Detection and monitoring of African vegetation fires using MSG-SEVIRI imagery, Remote Sens Environ., 114, 1038-1052. doi:10.1016/j.rse.2009.12.019.

Example of Product