Monitoring Sparse Perennial Vegetation Cover

In Australia, remotely sensed Landsat TM data is routinely used for mapping and monitoring the change in extent of woody perennial vegetation cover providing information for conservation, land management and for modelling carbon flux and water use. For a variety of reasons including land cover definitions and sensor spectral limitations, studies have tended to focus on reporting changes in land covers defined as having a canopy cover greater than some nominal value, say 20%. Increasingly there is a requirement to map and monitor current and historical changes in areas of sparse perennial vegetation having 5% to 20% canopy cover. Using Landsat TM data, sparse perennial vegetation cannot be reliably discriminated from other ground cover types using just the spectral information for a given pixel, however by including information on texture the results can be significantly improved. In this paper we describe the development of a methodology and its application to mapping and monitoring sparse perennial vegetation. Examples will be presented from a collaborative project conducted in NSW with the Australian Greenhouse Office and the NSW Department of Infrastructure, Natural Resources and Planning.