Assessing forest landscape structure using geographic windows

Landscape structure, interpreted as indicator of functional processes, has become a main attribute of multiresource forest inventories, enhancing its value with respect to society needs. This approach implies effective use of earth observation techniques and geographic information systems to obtain a global view of the inventoried landscapes and to understand the ecological functions of large spatially-heterogeneous landscape mosaics. Landscape structure often reveal extremely complex patterns that can only be very roughly characterized by methods of Euclidean geometry. Conversely, fractals can be applied to adequately describe many of the irregular, fragmented patterns found in nature. In this paper, a Fractal Analysis Software (FRACAS) for computing local fractal dimension on classified raster GIS layers with a non-regression technique is presented. FRACAS is based on the use of Merchant's adaptive geographic window, which is designed to operate on a neighborhood of patches instead of a fixed rectangular neighborhood of pixels (the conventional approach in image analysis).