Ascertainment on abnormity component and classification on grades for alteration information extracted by principal component analysis from ETM data

摘要

Many ETM (Enhanced Thematic Mapper) digital image processing methods such as ratio and principal component analysis (PCA) have been developed in exploration with spectral feature reflected in ETM data. Unfortunately, by our knowledge, there is no clear idea in ascertainment on abnormality component in PCA and no quantitative scale to classify abnormality for alteration. In this study, to improve exploration efficiency, the rules of ascertainment on abnormality component and classification on abnormality for alteration were established. PCA with ETM bands 1, 4, 5, 7 and ETM bands 1, 3, 4, 5 with interferential factors masked for OH~- alteration and Fe~(2+) and Fe~(3+) alteration respectively were conducted. Meanwhile, the rules for ascertainment on alteration abnormality by PCA for OH" or Fe~(2+) and Fe~(3+) alteration were well established by the contribution of their diagnostic spectral bands. That is, the abnormity component of principal components (PCs) for OFT alteration can be ascertained by obtaining positive contribution from ETM band 5 but negative from ETM band 7, while that of Fe~(2+) and Fe~(3+) alteration obtained positive contribution from ETM band 3 but negative from ETM bands 1, 4 and 5. Alteration grades were delineated by the standard deviation σ and average uof pixel value of abnormality component which coincides with probability density function. Accordingly, drill holes conducted in potential areas of alteration, a famous mineralization belt in China, Northwestern Yunnan, also reveals that alteration-related information extracted from ETM data is practical and effective in mineral application.