TomoSAR derived features for estimation of Forest Structure and Fuel load

Abstract:
Wildfire management/prediction require accurate mapping of forest fuel load, which depends primarily on trees' diameter, stem volume, and forest structure. Thus, to achieve reliable wildfire risk estimation using remote sensing imagery, this paper aims to establish a link between the estimated vertical forest structures derived from SAR tomography and the quantification of actual forest fuel load. In this context, some novel statistical features are introduced that have potential to indicate fuel load. Based on the results obtained over the tropical forest of Mondah, Gabon, it is expected that these features will be useful for developing predictive models of fuel load.