Change Vector Analysis technique to monitor selective logging activities in Amazon

Abstract

The lack of sustainability in the exploitation of the tropical Amazon forest has caused severely impacted environment and biodiversity. In this context, the objective of this study is to detect, characterize and quantify the forest areas affected by timber exploitation in the years 2001 and 2002. This research was developed in the north of the Mato Grosso state, Brazil, utilizing fraction images that were generated by a linear spectral mixture model referring to the mentioned years. Change vector analysis (CVA) was applied to the difference image of the fraction images from the two years. One image representing magnitude of change, and two images representing the angles of the change vectors were generated. These images are efficient for detection of intensity and type of change that occurred in the forested areas, which are subject to selective logging.