LSA SAF Meteosat FRP products – Part 1: Algorithms, product contents, and analysis

摘要

Characterizing changes in landscape fire activity at better than hourlytemporal resolution is achievable using thermal observations of activelyburning fires made from geostationary Earth Observation (EO) satellites.Over the last decade or more, a series of research and/or operational"active fire" products have been developed from geostationary EO data, oftenwith the aim of supporting biomass burning fuel consumption and trace gasand aerosol emission calculations. Such Fire Radiative Power (FRP)products are generated operationally from Meteosat by the Land SurfaceAnalysis Satellite Applications Facility (LSA SAF) and are available freelyevery 15 min in both near-real-time and archived form. These productsmap the location of actively burning fires and characterize their rates ofthermal radiative energy release (FRP), which isbelieved proportional to rates of biomass consumption and smoke emission.The FRP-PIXEL product contains the full spatio-temporal resolution FRPdata set derivable from the SEVIRI (Spinning Enhanced Visible andInfrared Imager) imager onboard Meteosat at a 3 km spatialsampling distance (decreasing away from the west African sub-satellitepoint), whilst the FRP-GRID product is an hourly summary at 5°grid resolution that includes simple bias adjustments for meteorologicalcloud cover and regional underestimation of FRP caused primarily byunderdetection of low FRP fires. Here we describe the enhancedgeostationary Fire Thermal Anomaly (FTA) detection algorithm used to deliverthese products and detail the methods used to generate the atmosphericallycorrected FRP and per-pixel uncertainty metrics. Using SEVIRI scenesimulations and real SEVIRI data, including from a period of Meteosat-8"special operations", we describe certain sensor and data pre-processingcharacteristics that influence SEVIRI's active fire detection and FRPmeasurement capability, and use these to specify parameters in the FTAalgorithm and to make recommendations for the forthcoming Meteosat ThirdGeneration operations in relation to active fire measures. We show that thecurrent SEVIRI FTA algorithm is able to discriminate actively burning firescovering down to 10−4 of a pixel and that it appears more sensitive tofire than other algorithms used to generate many widely exploited activefire products. Finally, we briefly illustrate the information containedwithin the current Meteosat FRP-PIXEL and FRP-GRID products, providingexample analyses for both individual fires and multi-year regional-scalefire activity; the companion paper (Roberts et al., 2015) provides a fullproduct performance evaluation and a demonstration of product use withincomponents of the Copernicus Atmosphere Monitoring Service (CAMS).