FIRE DETECTION AND MONITORING OVER AFRICA

Biomass burning is a significant global source of greenhouse gases (e.g. carbon dioxide and methane) as well as of nitric and carbon monoxides, methyl bromide and hydrocarbons that lead to acid rain and the photochemical production of tropospheric ozone and destruction of stratospheric ozone which impact global climate. Other impacts of biomass burning relate to the biogeochemical cycling of nitrogen and carbon compounds, the hydrological cycle, the reflectivity and emissivity of the land, the stability of ecosystems and ecosystem biodiversity. The potential of the SEVIRI instrument on-board the MSG series for applications related to fire detection and monitoring has long been recognized. We present an operational procedure for active fire detection based on information from Meteosat-8/SEVIRI, which is currently being developed within the framework of the Satellite Application Facility on Land Surface Analysis (LSA SAF). The procedure primarily relies on information from MSG channels (namely the 0.6 μm, 0.8 μm, 3.9 μm, 10.8 μm and 12.0 μm) together with information on illumination and viewing angles. The method is based on contextual algorithms that have been successfully developed for different sensors, namely GOES, NOAA-AVHRR and MODIS. A potential fire pixel is compared with the neighbouring ones and the decision is made based on relative thresholds as derived from the pixels in the neighbourhood. The algorithm is self-adaptive and has shown consistency over large areas and throughout the seasons. We will present an overview of results obtained for 2007, paying special attention to the duration of wildfires as well as to daily and weekly cycles of wildfire activity over different regions and for different types of land cover. Quality of results will be assessed by means of a set of tests applied under various conditions and to different ecosystems and by comparing obtained results with those from other sources, namely; i) the MODIS fire and thermal anomalies products, ii) burnt area maps from L3JRC of SPOT VGT and iii) the AFIS (Advanced Fire Information Service) from the CSIR-Meraka Institute (South Africa). INTRODUCTION The main purpose of the Satellite Application Facility (SAF) on Land Surface Analysis (LSA) is to take full advantage of remotely sensed data, particularly those available from EUMETSAT sensors, to describe/derive land surface properties/variables (LSA SAF, 2006). After successfully completing the Development Phase and the Initial Operational Phase, the LSA SAF has started, in March 2007, its continuous Development and Operations Phase (CDOP) that is planned to last until February 2012. The LSA SAF products are relevant to a wide range of applications and, since the beginning of its scientific and technical activities, particular attention has been devoted to the end-user community (Tavares and DaCamara, 1999). Although the Numerical Weather Prediction (NWP) community has been identified as having the greatest potential to fully exploit the LSA SAF products and was therefore assigned the highest priority during the Development Phase (DaCamara and Tavares, 2001), the LSA SAF addresses a much broader community (Tavares and DaCamara, 2002), including amongst others agricultural and forestry applications and natural hazard management. Whereas agricultural and forestry applications require information on soil and vegetation properties, natural hazard management requires frequent observations of terrestrial surfaces in both the solar and thermal bands together with merged information from