Operational considerations for the application of remotely sensed forest data from LANDSAT or other airborne platforms

Research in the application of remotely sensed data from LANDSAT or other airborne platforms to the efficient management of a large timber based forest industry was divided into three phases: (1) establishment of a photo/ground sample correlation, (2) investigation of techniques for multi-spectral digital analysis, and (3) development of a semi-automated multi-level sampling system. To properly verify results, three distinct test areas were selected: (1) Jacksonville Mill Region, Lower Coastal Plain, Flatwoods, (2) Pensacola Mill Region, Middle Coastal Plain, and (3) Mississippi Mill Region, Middle Coastal Plain. The following conclusions were reached: (1) the probability of establishing an information base suitable for management requirements through a photo/ground double sampling procedure, alleviating the ground sampling effort, is encouraging, (2) known classification techniques must be investigated to ascertain the level of precision possible in separating the many densities involved, and (3) the multi-level approach must be related to an information system that is executable and feasible