Vorster, Anthony, authorStohlgren, Thomas, advisorKumar, Sunil, advisorCheng, Tony, committee memberEvangelista, Paul, committee member2007-01-032007-01-032014http://hdl.handle.net/10217/84570The availability of remote sensing imagery before, during, and after the recent mountain pine beetle (Dendroctonus ponderosae Hopkins) epidemic in the southern Rocky Mountains presents exciting opportunities for assessing the current state of forests and how forest management in previous decades influenced outbreak severity across the landscape. I mapped outbreak severity at a 30-m resolution using integrative spatial modeling. I predicted that: 1) outbreak severity can be accurately predicted and mapped at Fraser Experimental Forest, Colorado using stand characteristics with a boosted regression tree model, Landsat imagery, geographic information system (GIS) data, and field data; and 2) forest stands that were unmanaged since the 1950s will have higher outbreak severity compared to stands that were treated since the 1950s. Outbreak severity, measured by the ratio of dead lodgepole pine (Pinus contorta) basal area to the basal area of all trees, was mapped across Fraser Experimental Forest with a cross-validation correlation of 0.86 and a Spearman correlation with independently observed values of 0.64. The outbreak severity at stands harvested between 1954 and 1985 was lower than comparable uncut stands. Lessons learned about past treatments will inform forest management for future mountain pine beetle outbreaks.born digitalmasters thesesengCopyright and other restrictions may apply. User is responsible for compliance with all applicable laws. For information about copyright law, please see https://libguides.colostate.edu/copyright.boosted regression treesLandsatmountain pine beetleremote sensingsilvicultureText