Unmasking forest borderlines by an automatic delineation based on airborne laser scanner data

Abstract

ABSTRACTThis study presents an automatic methodology based on airborne laser scanner (ALS) data, which allows the mapping of forests, using quantitative criteria typical of forest definitions, i.e. minimum threshold for the height of trees, canopy cover, forest area size, and width. Interactions between forest and other land uses are explored by the methodology for the definition of forest borderlines using an additional criterion; this criterion is the distance-discontinuity (DD), which establishes a minimum width (MW) for portions of territory categorized by land uses different from the forest. The proposed forest mapping approach introduces also a fuzzy algorithm to assess the canopy cover, thereby enhancing the positional accuracy in the delineation of the forest borderline. This methodology has a very flexible mapping approach with the potential to address the many forest definitions existing worldwide. The evaluation and the improvement of the methodology are particularly encouraged by its release as an open source tool.