ABSTRACT
Mediterranean landscapes are characterized by high stability to fire since regeneration of pre-existing spatial relationships among distinctive vegetation patches is quite rapid. The aim of this paper is to introduce a method based on percolation theory to quantitatively estimate structural variations in remotely sensed biomass data linked to post-fire dynamics of Mediterranean vegetation. The ability of the proposed method to estimate fire-induced variations in Mediterranean vegetation patterns made it appropriate to the monitoring of post five vegetation regrowth in Mediterranean landscapes at the Landsat TM scale.