| Summary: | viii, 75 p. This paper examines data collection through traditional methods as compared to remote sensing for the purposes of vegetation analysis in a coastal wet-sedge ecosystem in Barrow, Alaska. Vegetation mapping of a 100 meter transect was used to ground-truth the remotely sensed hyperspectral data, with a high degree of relatedness resulting. Although mosses, lichens, and vascular plants show significant change in percent cover over the 2001 summer season, the normalized difference vegetation index (NDVI), a spectral reflection index describing pigments, shows the vascular plants to be most highly correlated to the seasonal progression of the ecosystem. Mosses show high NDVI values and lichens low NDVI values throughout the season, both with little alteration, while vascular plants have much larger fluctuations, and correspond to seasonal ecosystem NDVI. Vascular plants also show the strongest correlation between NDVI and photosynthesis, whereas mosses and lichens have weaker correlation and lower rates of photosynthesis. Furthermore, functional type and NDVI differ with respect to microtopography; higher areas are often drier and covered with lichens and non-graminoid vascular plants while lower areas are wetter and dominated by mosses and graminoids, and generally give higher NDVI values. Therefore, vascular plants, which tend to reside in wetter areas, are more affected by seasonal warming with respect to percent cover, NDVI, and photosynthesis. Thus, future remote sensing employing hyperspectral bands should focus on the role of vascular plants, as they may be the most prominent indicators of ecosystem change, particularly in the case of global warming. Department of Biology and Microbiology. California State University. Los Angeles, California. University of Maryland, Baltimore County. Baltimore, Maryland.
|
|---|