Remote sensing of vegetation with space-borne multispectral sensors has been successful in mapping vegetation communities on meso to broad scales, and in extrapolating other indices of plant function such as NDVI and VI. As much as multispectral remote sensing has succeeded in many areas it has been limited in others, such as fine scale discrimination of lower order taxonomic vegetation groups, or the detection of vegetation on a sub-pixel basis. Hyperspectral sensors have achieved this with some success; however there is a need to apply this capability to a wider range of vegetation types in a broad range of ecosystem settings.
This project aims to test the extent to which several Australian native vegetation species can be discriminated or spectrally separated with HyMap hyperspectral imagery. The Monarto study site, approximately 50km east of Adelaide, South Australia, presents a great diversity of endemic and naturalised tree species to test spectral diversity. Spectral variation with and between plant groups is being tested, with the ultimate aim of mapping distribution and heterogeneity in blocks of remnant and regenerating vegetation. The information derived will give an understanding of hyperspectral remote sensing’s ability to discriminate species and groups of species of native vegetation. The results of this study will have a wide range of benefits and applications including for ecological studies, conservation, forestry, and land management.