A multitemporal land-cover change analysis tool using change vector and principal components analysis

Scientists are devoting a lot of energy to monitor and to assess the impact of fires and to measure the evolution on land cover in tropical savannas. A key element in a monitoring system is the land cover changes analysis tool. While some land cover changes are caused by natural processes, such as long term changes of the climate or shorter-term vegetation successions, human activity modifies surface cover through direct actions, such as deforestation, farming activities, or indirectly, through man-induced climatic change. This paper presents a method that applies principal components analysis for land-cover change detection with multitemporal SPOT data.