Abstract
Vegetation boundaries such as riparian zones or forest stand edges are inherent features of landscapes that play more than one functional role in terrestrial ecosystems dynamics (Hansen and Di Castri 1992). The delineation of ecotones (i.e., boundary areas), either sharp or gradual, between adjacent vegetation communities is important in several aspects of land use management planning (e.g., the delineation of forest stands for estimating wood volume available for harvesting). The location and accuracy of the delimited vegetation boundaries, however, depend on the spatial and temporal resolution of the data available as well as the subsequent statistical methods used to detect them. In the context of forest stand delineation, various data types (e.g., field data, aerial photographs and remotely sensed images) are available. The type of data determines which edge detection methods (e.g., statistical methods, photointerpretation, or filter kernels) should be used. Accuracy of the delineated vegetation boundaries therefore depends both on the accuracy of the data and the sensitivity of the edge detection methods.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have