Show simple item record

dc.contributor.advisorEgbert, Stephen L.
dc.contributor.authorWamser, William Kyle
dc.date.accessioned2013-02-17T18:16:37Z
dc.date.available2013-02-17T18:16:37Z
dc.date.issued2012-12-31
dc.date.submitted2012
dc.identifier.otherhttp://dissertations.umi.com/ku:12572
dc.identifier.urihttp://hdl.handle.net/1808/10839
dc.description.abstractHyperspectral imagery and the corresponding ability to conduct analysis below the pixel level have tremendous potential to aid in landcover monitoring. During large ecosystem restoration projects, being able to monitor specific aspects of the recovery over large and often inaccessible areas under constrained finances are major challenges. The Civil Air Patrol's Airborne Real-time Cueing Hyperspectral Enhanced Reconnaissance (ARCHER) can provide hyperspectral data in most parts of the United States at relatively low cost. Although designed specifically for use in locating downed aircraft, the imagery holds the potential to identify specific aspects of landcover at far greater fidelity than traditional multispectral means. The goals of this research were to improve the use of ARCHER hyperspectral imagery to classify sub-canopy and open-area vegetation in coniferous forests located in the Southern Rockies and to determine how much fidelity might be lost from a baseline of 1 meter spatial resolution resampled to 2 and 5 meter pixel size to simulate higher altitude collection. Based on analysis comparing linear spectral unmixing with a traditional supervised classification, the linear spectral unmixing proved to be statistically superior. More importantly, however, linear spectral unmixing provided additional sub-pixel information that was unavailable using other techniques. The second goal of determining fidelity loss based on spatial resolution was more difficult to determine due to how the data are represented. Furthermore,the 2 and 5 meter imagery were obtained by resampling the 1 meter imagery and therefore may not be representative of the quality of actual 2 or 5 meter imagery. Ultimately, the information derived from this research may be useful in better utilizing hyperspectral imagery to conduct forest monitoring and assessment.
dc.format.extent72 pages
dc.language.isoen
dc.publisherUniversity of Kansas
dc.rightsThis item is protected by copyright and unless otherwise specified the copyright of this thesis/dissertation is held by the author.
dc.subjectGeography
dc.subjectRemote sensing
dc.titleUSING HYPERSPECTRAL IMAGERY TO ASSIST FEDERAL FOREST MONITORING AND RESTORATION PROJECTS IN THE SOUTHERN ROCKY MOUNTAINS, COLORADO
dc.typeThesis
dc.contributor.cmtememberSlocum, Terry A.
dc.contributor.cmtememberLi, Xingong
dc.thesis.degreeDisciplineGeography
dc.thesis.degreeLevelM.A.
kusw.oastatusna
kusw.oapolicyThis item does not meet KU Open Access policy criteria.
kusw.bibid8085687
dc.rights.accessrightsopenAccess


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record