Multi-temporal Remote Sensing of Rangeland Vegetation for Investigation of Fire-related Ecology at Canadian Forces Base Suffield, Alberta

Date
2013-05-27
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Canadian Forces Base (CFB) Suffield, located in southeastern Alberta, faces pressure from a variety of competing land uses and requires geospatial tools to quantify and manage the effect of human activities (particularly military training-related fire) on ecosystem functions. I used multi-temporal remote-sensing techniques to model plant functional types (PFT; C3 vs. C4 grasses), as an indicator of ecosystem state. The best-performing model (overall accuracy = 74 %, weighted kappa = 0.53) was compared against a spatial fire-history database digitized from the Landsat archive (1972 to 2007). Probit regression results revealed statistically significant relationships between PFT-derived ecosystem states and fire history (P < .001), but succession processes were different between ecological units. In general, this ecosystem is sensitive to repeated fire, with recovery taking decades. This research provides novel contributions to ecological knowledge in northern mixedgrass prairie, and outlines specific management actions required to maintain ecologically sustainable fire frequency.
Description
Keywords
Range Management, Ecology, Remote Sensing
Citation
Smith, B. (2013). Multi-temporal Remote Sensing of Rangeland Vegetation for Investigation of Fire-related Ecology at Canadian Forces Base Suffield, Alberta (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/27041