Forest cover trends from time series Landsat data for the Australian continent

摘要: Abstract In perennial and natural vegetation systems, monitoring changes in over time is of fundamental interest for identifying quantifying impacts management processes. Subtle cover can be identified by calculating the trends a density index time. this paper, we apply such an index-trends approach, which has been developed applied to series Landsat imagery rangeland woodland environments, continental-scale disturbances within forested regions Australia. This paper describes operational methods used generation National Forest Trend (NFT) information, time-series summary providing visual indication within-forest (disturbance recovery) at 25 m resolution. result based on national archive calibrated TM/ETM+ data from 1989 2006 produced Australia's Carbon Accounting System (NCAS). The NCAS was designed 1999 initially provide consistent fine-scale classifications forest extent (i.e. land use change) Australian continent using imagery. NFT information identifies more subtle areas provides capacity identify processes affecting forests are primary ecologists managers. product relies identification appropriate Landsat-based (defined as linear combination spectral image bands) that sensitive density. values location, derived imagery, represents surrogate track changes. To produce statistical summaries response (such slope quadratic curvature) calculated. These calculated responses woody then displayed maps where different colours indicate approximate timing, direction (decline or increase), magnitude spatial cover. trend images self-contained easily interpretable change scales relevant resource (NRM) environmental reporting.