Forest Cover Change in Siberia - Results from the Siberia-II Project

* Corresponding author. Abstract – In this paper results of an object oriented image analysis approach for forest change detection based on spectral, contextual and object specific attributes are presented. A change detection classification system is developed that derives a change map using 1989 and the status in 2000 for selected forest inventory test territories in Siberia with extensive ground truth information from Russian forest inventory enterprises. Multitemporal Landsat satellite data sets are co-registered and atmosphere corrected. A hierarchical object-based and multi temporal class description is developed and applied to different multitemporal data sets. This work presents the methodology and results of a combination of a direct two-date change detection and a post classification procedure that also uses landuse change information.