| This thesis contains results of an evaluation of the spectral signatures of vegetation communities recognised on multispectral photography, at scales of 1:15,000 and 1:5,000, in North West Queensland, Australia, with reference to specific environmental parameters, notably bedrock mineralization, and microdensitometer optical density measurements. Detailed ground-truth information obtained for the evaluation included plant species frequency recording and plant and soil sampling along a series of transects located across known mineralized zones and across spectral anomalies detected on the photography. Overlays of the spectral signature units, including those subjected to enhancement techniques, were prepared and correlated with the ground-truth data and a number of microdensitometer scan lines measuring the optical density within each of the eight spectral bands of the multispectral photography. The spectral signatures of vegetation, and especially the distinct spectral signature of geobotanically anomalous plant communities forming surface expression of mineralized bedrock and related geochemical soil anomalies, are best displayed on false colour infra-red photography which covers the visible and near infra-red part of the spectrum in which plant reflectivity is most sensitive, consequently yielding most information. Ground-truth investigations, confirming the interpretation of infra-red spectral signatures, reveal the composition of anomalous and background vegetation units and indicatethe influence of the mineral status of the soil, relief, drainage and lithology on plant species distribution.Over mineralized terrain, higher reflectivity in all spectral bands and especially the true colour green, panchromatic green and false colour infra-red are recorded by microdensitometer optical density measurements which positively correlate with the above average concentrations of metal in the indicator plants and surface soils.In the semi-arid environment of North West Queensland, the use of multispectral photography, in conjunction with detailed field studies, is shown to have application in the study of vegetation associations and plant conmunities forming surface expression of characteristic ground conditions. The recognition of geobotanical anomalies and differences in soil and plant reflectivity indicative of bedrock mineralization offers considerable promise in mineral exploration for the detection of ore bodies. |
