xii, 127 p. : ill. (some col.) / I developed software to incorporate the FlamMap fire model into an agent-based model, Envision, to enable the exploration of relationships between wildfire, land use, climate change, and vegetation dynamics in the Willamette Valley.
A dynamic-link library plug-in utilizing row-ordered compressed array lookup tables converts parameters between polygon-based Envision data and grid-based FlamMap data. Modeled fires are determined through Monte-Carlo draws against a set of possible fires by linking historic fire data to future climate projections.
I used classification and regression tree (CART) and logistic regression to relate ignitions to human and land use factors in the Willamette Valley above the valley floor from 2000-2009. Both methods showed decreasing distance to major and minor roads as key factors that increase ignition probability for human ignitions but not for lightning ignitions. The resulting statistical model is implemented in the FlamMap plug-in to provide a dynamic ignition probability map over time. / Committee in charge: Dr. Bart Johnson, Co-Chair;
Dr. Scott Bridgham ,Co-Chair;
Dr. John Bolte; Member
Identifer | oai:union.ndltd.org:uoregon.edu/oai:scholarsbank.uoregon.edu:1794/12184 |
Date | 12 1900 |
Creators | Sheehan, Timothy J. |
Publisher | University of Oregon |
Source Sets | University of Oregon |
Language | en_US |
Detected Language | English |
Type | Thesis |
Rights | rights_reserved |
Relation | University of Oregon theses, Dept. of Biology, M.S, 2011; |
Page generated in 0.0011 seconds