
![]()
The primary data source was multi-date NOAA satellite LAC (Local Area Coverage-1km) resolution imagery acquired during the summer months. A classification scheme was selected that described the diversity of the land cover. Particular attention was given to the classification of the forest. Automated image analysis was conducted using both maximum Likelihood and parallelepiped methods. Training sites were selected with the aid of LANDSAT data, maps, and forest inventory maps.
The land cover classes that were identified were: Mixedwood Forest, Deciduous Forest, Water, Transitional Forest, Coniferous Forest, Artic/Alpine Tundra, Barren Land, Perennial Snow/Ice, Agricultural Cropland, Rangeland and Pasture, Built-Up Land, USA, Boundary Lines, and Permanent Polar Ice.