Use of satellite imagery to measure cover of prairie vegetation for the detection of change

• Main Author: Hurst, Rebecca Jeanne
• Language: en
• Published: 2006
• Online Access: http://etd.lib.montana.edu/etd/2006/hurst/HurstR0506.pdf

Adaptive resource management requires a cost effective, easily repeatable tool for measurement of vegetation quality and comparison of management treatment effects across a natural area and through time. High-resolution satellite imagery may be used as a guide for treatment by facilitating either the comparison of differentially treated units or the measurement of change in vegetation through time. Demonstration of successful application of remote sensing for measurement of vegetation to recognize trends in vegetation quality will help adaptive managers both to apply and to improve the methods for management of vegetation. Space Imaging's IKONOS satellite imagery was used first to map mixed grass prairie communities in the Missouri Coteau region of North Dakota and Montana. Vegetation was classified hierarchically into 5 classes (wetland, tree, shrub, grass/dwarf shrub, and pure grass) with an overall accuracy of 72%. The resultant map was used to sample vegetation of units with various management histories in Lostwood National Wildlife Refuge to measure differences across fire and grazing treatments. Differences across treatments were slight. Satellite imagery may provide the best tool for resampling needed in this and other systems. Sampling with remote sensing may be more expedient than ground surveying, but classifications must have higher accuracies and less bias to be useful to managers. Therefore, satellite imagery needs further development to support on-going adaptive resource management.