No abstract

No abstract

Graeme Behn , a research officer with WA Department of Conservation and Land Management, is also based at the Leeuwin Centre (E-mail: graemeb@calm.wa.gov.au). The work described here is based on their experiences over many years in a range of environments in methods development and applications using satellite imagery to provide environmental monitoring information.Summary Vegetation changes over time are important indicators of condition, and are particularly important as targets or triggers for management. Satellite image data have unique capacities to provide information on changes in vegetation. In particular, Landsat imagery has the spatial resolution and a historical archive that make it relevant to providing information for understanding and management of native vegetation at a range of scales from small remnant to region. Regional and national vegetation monitoring programs based on time series Landsat imagery are now operational in Australia. These programs and their data have huge potential to provide information for conservation and natural resource management questions. They have already found multiple applications, including applications to biodiversity assessment and planning. This paper presents some examples of the delivery and application of satellite image monitoring information in the context of vegetation management.

The southwest of Western Australia is affected by dryland salinity that results in the loss of previously productive agricultural land, damage to buildings, roads, and other infrastructure, decline in pockets of remnant vegetation and biodiversity, and reduction in water quality. Accurate information on the location and rate of change of the extent of saline land over the region is required by resource managers. For the first time, comprehensive, spatially explicit maps of dryland salinity and its change over approximately 10 yr for the southwest agricultural region of Western Australia have been produced operationally in the 'Land Monitor' project. The methods rely on an integrated analysis of long-term sequences of Landsat TM satellite image data together with variables derived from digital elevation models (DEMs). Understanding of the physical process and surface expression of salinity provided by experts was used to guide the analyses. Ground data-the delineation of salt-affected land by field experts-was collected for training and validation. The results indicate that the land area currently affected by salinity in Western Australia's southwest is about 1 million hectares (in 1996) and the annual rate of increase is about 14,000 ha. This is a lesser extent than many previous estimates and lower rate of change than generally predicted from limited hydrological data. The results are widely distributed and publicly available. The key to providing accurate mapping and monitoring information was the incorporation of time series classification of a sequence of images over several years combined with landform information.

THEME: BIOD / Biodiversity. Special session: "Trends in operational land cover mapping" KEY WORDS: Multi-temporal, Landsat, Land Cover, Monitoring, Forestry, Change Detection ABSTRACT:Land use and forest change, in particular deforestation, have contributed the largest proportion of Indonesia's estimated greenhouse gas emissions. Indonesia's remaining forests store globally significant carbon stocks, as well as biodiversity values. In 2010, the Government of Indonesia entered into a REDD+ partnership. A spatially detailed monitoring and reporting system for forest change which is national and operating in Indonesia is required for participation in such programs, as well as for national policy reasons including Monitoring, Reporting, and Verification (MRV), carbon accounting, and land-use and policy information.Indonesia's National Carbon Accounting System (INCAS) has been designed to meet national and international policy requirements. The INCAS remote sensing program is producing spatially-detailed annual wall-to-wall monitoring of forest cover changes from time-series Landsat imagery for the whole of Indonesia from 2000 to the present day. Work on the program commenced in 2009, under the Indonesia-Australia Forest Carbon Partnership. A principal objective was to build an operational system in Indonesia through transfer of knowledge and experience, from Australia's National Carbon Accounting System, and adaptation of this experience to Indonesia's requirements and conditions. A semi-automated system of image pre-processing (ortho-rectification, calibration, cloud masking and mosaicing) and forest extent and change mapping (supervised classification of a 'base' year, semi-automated single-year classifications and classification within a multi-temporal probabilistic framework) was developed for Landsat 5 TM and Landsat 7 ETM+. Particular attention is paid to the accuracy of each step in the processing. With the advent of Landsat 8 data and parallel development of processing capability, capacity and international collaboration s within the LAPAN Data Centre this processing is being increasingly automated. Research is continuing into improved processing methodology and integration of information from other data sources. This paper presents technical elements of the INCAS remote sensing program and some results of the 2000 -2012 mapping.* Corresponding author. This is useful to know for communication with the appropriate person in cases with more than one author.