Creation of compensatory wetlands has been required in the U.S. since the late 1980s in an attempt to offset the massive decline in freshwater wetlands. To meet permitting requirements, vegetation composition in mitigation wetlands must be monitored for a minimum of five years following creation. Unfortunately, mitigated wetlands often lack the functionality of natural wetlands and may form hotspots for invasive plant species. However, wetland assessment is a time-consuming process that may also disturb fragile nascent plant communities. Thus there is a need for approaches that minimize disturbance, but still enable the collection of data over large portions of the landscape. Remote sensing, using hyperspectral imagery augmented by field data collection is a potential tool for rapid ecosystem assessment. In July 2010, vegetation community composition, spectral signatures of individual plant species, and plant canopies, and an aerial hyperspectral imagery dataset were obtained from two natural and two mitigation wetlands on the Rochester Institute of Technology (RIT) campus, Rochester, NY. We were able to locate specific wavelengths for four invasive plant species spectra that can be used to classify and map these species on the RIT campus with an overall accuracy of 94.34%. Reed canarygrass had a higher reflectance than the other three species and differences along the red-edge and near-infrared regions also enabled differentiation between broadleaf cattail and narrowleaf cattail. Values within the blue, red, red-edge, and near-infrared regions are needed to create a multi-spectral sensor with a larger emphasis on the red-edge and near-infrared regions. Such a sensor would be more readily available for land managers for classification and analysis of large plots of land, limiting the amount of time, personnel and funding needed to process the imagery and allowing managers to more rapidly identify patches of invasive plant species with minimal intrusion on sensitive wetland environments.
Library of Congress Subject Headings
Wetland ecology--New York (State)--Rochester; Wetland management--New York (State)--Rochester; Invasive plants--Research; Invasive plants--Remote sensing; Multispectral imaging
van Aardt, Jan
Dutcher, Nicole, "Towards a characterization of wetland invasive vegetation using a combination of field and remote sensing techniques" (2013). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus