PARENT SESSION
Poster Session #18: Remote Sensing and GIS.
Tuesday, August 7, 2001. Presentation from 3:30 PM to 5:00 PM. Exhibition Hall
94
Using geospatial technologies for ecological classification of longleaf pine sites in southern Alabama.
MACKENZIE, MARK1, CARTER, ROBERT2, 1 2
ABSTRACT- Previous field research in the Conecuh National Forest and the Solon Dixon Forestry Education Center in south Alabama identified seven landscape scale Landtypes (LTs) on upland sites. Each LT had a unique suite of plant species as well as unique soil and landform variables. The research presented in this poster addresses the question of whether geospatial technologies (i.e., remote sensing, GIS, and GPS) can be used to extrapolate field based, sample classification results across the entire landscape. A landscape scale classification would be extremely useful to resource managers (e.g., the USDA Forest Service) in making management decisions, especially in regard to longleaf pine (Pinus palustris)restoration. A spatial database was created in the context of a GIS. Variables included in the database were based on those found to be discriminating variables determined through field sampling (e.g., soil texture, topography, landform, and land cover). Current land cover was classified using satellite imagery (i.e., Landsat TM and ETM). Landtypes within the entire study area will be identified using discriminant functions developed from the field sampling. Preliminary results indicate that LTs can be classified within the context of a GIS but, to date, accuracy assessment based on ground truthing has not been performed.
KEY WORDS: Pinus palustris, GIS, remote sensing, discrimnant analysis