Hyperspectral remote sensing of marsh species and plant vigour gradient in the New Jersey Meadowlands

A study was conducted to map the plant vigour gradient using hyperspectral imagery combined with field-collected seasonal reflectance spectra of marsh species in a fragmented coastal wetland. Marsh surface types were identified by classifying a low-tide hyperspectral image of New Jersey Meadowlands acquired by Airborne Imaging Spectroradiometer for Applications (AISA) in October 2000. Reflectance spectra of dominant marsh species and seasonal spectra of Phragmites australis from April to October in 2003 were measured in the field. The separability of field-collected spectra of marsh species was tested and determined using non-parametric U-test. The spectra of Phragmites in greening-up phases were used as the surrogate to determine the vigour gradient of common reed stands. The results demonstrate that the reflectance spectra of one healthy stand of Phragmites sampled across the growing season did provide equivalent signatures of pure physiognomic types, which can be used to determine the plant vigour gradient.