Studying long-term vegetation dynamics using digital processing of historical aerial photographs

Plant ecologists have long recognized the importance of aerial photographs as a data source for studies of vegetation dynamics. Recent advances in com puter-aided technology (digital photogrammetry, computerized image processi ng, and geographical information systems) have opened new possibilities for the extraction of data on vegetation changes from aerial photographs. In t his study we describe a computer-based approach for studying landscape-scal e, long-term vegetation dynamics, using historical aerial photographs as a major data source. The method we employ consists of four main steps: 1) ima ge scanning and preprocessing (rectification, georeferencing, spectral corr ections and mosaicking), 2) image classification and construction of vegeta tion maps, 3) field validation, and 4) statistical analysis of vegetation c hanges. We applied our approach by analyzing changes in tree cover over a p eriod of 32 years in a mountainous landscape dominated by Mediterranean maq uis in northern Israel and discuss the main limitations and potential error sources of each stage of our analysis. We conclude that digital processing of historical aerial photographs may serve as a powerful tool for the dete ction, quantification, and analysis of landscape-scale patterns of vegetati on dynamics. This conclusion is a important because aerial photographs prov ide the largest source of information available today for research of long- term vegetation dynamics, and are the only source of information on vegetat ion dynamics that combines high spatial resolution, large spatial extent, a nd long-term coverage. (C) Elsevier Science Inc., 1999.