Date of Award


Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Master of Science (MS)


Forest Resources


Steven A. Sader

Second Committee Member

Alan White

Third Committee Member

William B. Krohn


This thesis demonstrates the use of economical, medium spatial resolution Landsat satellite imagery to map and monitor forest conditions and changes on a statewide scale. Twenty-nine Landsat scenes were used to develop a statewide forest change map and fourteen Landsat scenes were used to create a 2007 forest cover type map of Maine. Chapter 1 demonstrates in a straightforward, reproducible method to create a statewide forest change and cover type map. The cover type map was evaluated for agreement with U. S. Forest Service Forest Inventory and Analysis (FIA) ground plot data (64% overall agreement) while the change map was evaluated using visual interpretation of random sample locations on multiple years of Landsat data and aerial photos (84% overall accuracy). Forest disturbance rates and 2007 forest cover type distributions were aggregated and reported for multiple biophysical and administrative units (biophysical region, sub-basin watersheds, counties, townships, elevation and slope) to determine patterns of harvesting trends. The previous era of forest industry landowners in Maine has been replaced with corporate structures as Real Estate Investment Trusts and Timber Management Investment Organizations, who now own the majority (41.2%) of Maine's commercial forest lands. Between 2000 and 2007, four million acres of commercial forestland changed ownership. The statewide forest disturbance rates and cover type composition of various landowner groups were evaluated. The investment landowner types were found to have the highest disturbance rates (18%) over the seven year study period, and significantly higher than public and conservation forest land groups. Forestlands that changed hands had similar disturbance rates to stable landowners. Disturbance rates on certified and easement forestlands, compared to non-certified and non-easement lands, indicated no significant differences.