Box 1. Generating Natural Forest Cover Data 2013
Interpretation of satellite imageries 2013
Natural forest cover data 2013 obtained from the interpretation of Landsat ETM 7 images interpretation and also Landsat 8 with coverage time between 2012 and 2013. The span of this coverage time is intended to get the image data that are net of clouds. The satellite images were downloaded from the www.glovis.usgs.gov site.
The satellite images that have been downloaded, then extracted and composited. The channel used for the Landsat ETM 7 images is channel 1 to 5, while for Landsat 8, using channel 4 to 6. The image interpretation conducted with the combination of channel 543 for Landsat 7 and combination of 654 for Landsat 8.
The image interpretation done visually on the computer screen (on screen digitizing) to delineate the covers that are still natural forests. The interpretation results of images covered in 2009 were used as the initial reference for interpretation of images covered in 2013. The minimum digitizing scale used is 1:50.000.
The accuracy of the fial result of forest cover interpretation tested using the method of contingency matrix accuracy test (confusion matrix) and kappa coeffiient. Total of sampling point of fild test used in the accuracy test is 6.086 points. From the resulf ot contigency matrix, the accuracy score is 82 percent and the kappa coeffiient score is 0,614.
Analyses of natural forest cover data 2013 and the forest loss (deforestation)
Indonesia’s natural forest is assumed as the forests that grows naturally and not from the planting process. Therefore, plantation forest is not categorized as f assumption is that there is no forest cover from the reforestation process during the period of 2009-2013. Based on those assumptions, the result of interpretation of forest cover 2013 is the result of a combination between the data on forest cover 2009 and the result of interpretation of forest cover 2013. This is also intended to improve the previous data on forest cover in 2009, which caused by the widespread of cloud appearance and error in interpretation.
The forest loss (deforestation) in 2009-2013 is the analyses of changes in natural forest cover during the period of 2009-2013 using the spatial data on forest area designation with the administrative area spatial data until the district level as the baseline. Deforestation in 2009-2013 is the condition of the area that in 2009 was still forested and in 2013 is not forest anymore, and disjoint with the designation of administrative areas