Date of Award

Spring 8-1-1988

Document Type

Thesis

Degree Name

Master of Arts (MA)

Department

Geography, Geology, and Anthropology

First Advisor

Paul Mausel

Second Advisor

William Brooks

Third Advisor

Kamlesh Lulla

Abstract

Three dates of Landsat digital data were classified and then analyzed using a geographic information system (GIS). The intent of this study was to determine if differences in burn severity relate to later vegetative cover within a northern pines forest. The May 5, 1980, Mack Lake Fire in the Huron National Forest, Michigan, was selected as the study site for this research. This wildfire burned over 9000 hectares of jack pine, red pine, and mixed deciduous forest within a six-hour period. Landsat MSS data from June 1973 and TM data from October 1982 were classified using an unsupervised approach to create pre-fire and post-fire landcover maps of the study area. These maps were then compared by means of a GIS, and a map of vegetation change was created. An infrared/red band ratio of Landsat MSS data from June 1980 was classified to create a map of three degrees of burn severity, and this was then compared to the map of vegetation change by means of the GIS. Analysis of the pine and deciduous forest revealed that the greatest change in vegetation was found to occur in areas that had been subjected to the most intense burn and, conversely, the least amount of change occurred within lightly burned areas. A large part of this change was attributed to two classes defined a? regenerating forest, with the remainder to shrub vegetation or another forest class. The shrub vegetation class, in contrast, showed the most change in lightly burned areas, and least in the severely burned, which correlates with conclusions drawn by other investigations. Results from this research indicate this technique is valid for vegetation change detection, and could be applied to other studies of change detection as well.

Download

Share

COinS