Advancing Indonesian forest resource monitoring using multi-source remotely sensed imagery

Indonesia experiences the third highest rate of deforestation among tropical countries (FAO, 2005, 2010, Hansen et al 2008, 2009). Providing timely and accurate forest data at the national scale is crucial to combat deforestation and forest degradation in support climate change mitigation and biodiversity conservation policy initiatives. Indonesia's national communication to the United Nations Framework on Climate Change (UNFCCC) in 2009 reveals a substantial role of forests as well as peatlands in contributing projected national carbon emission under business as usual. Further, in 2009 the government of Indonesia pledged to reduce national carbon emissions from between 26% to 41% of business as usual by 2020 while maintaining annual economic growth at 7%. The forests and wetlands therefore are becoming major interest for climate change mitigation in Indonesia. Our work here documents timely natural primary forests extent and loss within various types of physiographical formations. We present a hybrid approach to quantify the extent and change of primary forests for different composition and structure using a per-pixel supervised classification mapping combined with a GIS-based fragmentation analysis (Margono et al 2012). We also demonstrate the mapping of Indonesia's wetlands cover as a single thematic class (including peatlands, freshwater wetlands and mangroves) using topographical indices (derived from SRTM elevation data) combined with optical (Landsat) and radar (ALOS PALSAR) image data. We provided analysis within different physiographic formations (uplands and lowlands) defined using DEM data. The loss of Indonesia's primary intact and primary degraded forests, within wetlands and non-wetlands covers, as well as in the upland was quantified to provide suitable information for the objectives of the UNFCCC Reducing Emission from Deforestation and Forest Degradation (REDD and REDD+) program.