Abstract

Resource managers can benefit from improved methods for identifying invasive plant species. The utilization of hyperspectral remote sensing as a tool for species-level mapping has been increasing and techniques need to be explored for identifying species of interest. The overarching objective of this paper was to investigate three distinct processing methodologies (i.e., Derivatives, Continuum Removal, and Shape Filter) to explore their potential for delineating wetland invasive plant species within the spectral domain of typical airborne hyperspectral sensors. Field-level hyperspectral data (350-2500nm) were collected for twenty-two wetland plant species in a wetland located in the lower Muskegon River watershed in Michigan, USA. Generally, continuum removed spectra were more similar than raw reflectance for the invasive species of interest according to the Jeffries-Matusita distance measure. Second-derivative analysis showed that the wavelength locations of absorption and reflectance features were consistent for all species and emphasized the NIR region for separation. The shape-filter was useful as a method to identify invasive species and showed that useful wavelength regions can vary depending on the species of interest and approach utilized. Using the shape-filter, Lythrum salicaria, Phragmites australis, and Typha latifolia possessed maximum separation (distinguished from other species) at the red edge (700nm) and water absorption region (1350nm), the near-infrared down slope (1000 and 1100nm), and the visible/chlorophyll absorption region (500nm) and red edge (650nm), respectively.

Key words: invasive plant species, coastal wetlands, hyperspectral, derivatives, continuum removal, shape-filter

Was this article helpful?

0 0

Post a comment