Characterizing the spatio-temporal fire regime in Ethiopia using the MODIS-active fire product: a replicable methodology for country-level fire reporting

Abstract

In many regions of the world, fire is an integral part of land-use practices. The accurate spatio-temporal characterization of the fire regime can, therefore, inform land-use policy at many scales. Satellite-based fire detections can be manipulated with GIS methodologies to investigate the spatio-temporal patterns of fire across a landscape. However, caveats and accuracy limitations of data and analysis methodologies must be understood in order to avoid misrepresentation of the fire regime and its impacts. This research uses moderate resolution imaging spectroradiometer (MODIS) active fire detections (MCD14ML) together with land cover data (MOD12), (MOD44B), population data (Afripop) and information on land use drawn from the literature. A case study is presented for Ethiopia reporting on a 7-year period. Results show that 91% of fires occur in the woody savanna and savanna biomes, and fire activity is inversely correlated with population density. A 0.05° latitude/longitude grid is used to report fire density and indicated as more adequate than the existing 0.5° MODIS Climate Modelling Grid. Fire occurs with highest density in north-western Ethiopia, where smaller clusters of high fire activity are pointed out. Caveats and lessons learned are discussed in order to provide a best-practice methodology for country-level fire reporting.

Keywords:

• Fire
• MODIS
• Ethiopia
• Africa
• land cover