MAPPING VARIATIONS IN ALTERATION MINERALOGY WITH HYPERSPECTRAL REMOTE SENSING DATA

The goal of remote sensing geologists is to accurately map the locations of significant alteration minerals that will then be field checked. Principal Component (PC) analyses have the benefit of greatly increasing the contrast between different materials, leading to colorful images. However, the colors are often difficult to relate to specific mineral or rock types because the PC analysis is data dependent. The PC transformation varies with different surface materials. This paper discusses the use of a directed decorrelation stretch (DDS) to increase the contrast between surface materials and map variations in mineralogy and rock type. In this study hyperspectral airborne data were collected over two areas in Nevada. The data were converted to reflectance and absorption features of selected minerals of interest were analyzed. In spectroscopic analyses, the position and shape of absorption features are key. Images made using the DDS technique show detailed variation in mineral composition and distribution. These images can be a good complement to strict mineral identification images produced by typical spectral unmixing techniques. DDS and mineral mapping images produced for this project have helped focus exploration efforts and have also raised questions about lithologic variations.