<![CDATA[Geological mapping for volcano products that has just erupted will give important basic information to understand the eruption characteristic and the hazard mitigation that can be done. Difficult terrain and high danger around active volcano cause geological mapping hard to be done using direct observation method. To give another solution for the problem, we try to explain digital volcano products mapping using supervised classification method from field data that still be reached. Image data selection with atmospheric correction is the important factor to get high map accuracy. Supervised classification method that used in this research is Minimum Distance Classification. This method uses average vector for each ROI and calculates Euclidean distance from every pixel for each class, so every class can be distinguished well and precisionly. Geological observation is done before classification process by choosing 140 observation point that spread in the east of Agung Volcano. Geological data consists of lithology, correlation, and the mechanism of the rocks that used as the reference in the determination of class for classification from Landsat 8-OLI/TORS image. The research area is located on the east of Agung Volcano which has 60 km2. It is located between Kubu District, Abang District, and Bebandem District, Karangasem Regency, Bali Province. Supervised classification can map products of Agung Volcano for area that can’t be reached around 324 km2. Verification process is done by comparing supervised classification result in research area with observation point at the field using Classification Accuracy Assessment method with result reaching 90%.]]>
