Assessing the Suitability of Territories for Detailed Geological and Mineralogical Mapping Based on Statistical Methods for Processing Remote Sensing Data from the Landsat-8 Spacecraft: Using the Example of Southeastern Zabaykalsky Krai, Russia

This paper considers the suitability of using multispectral remote sensing data from the Landsat-8 satellite for conducting regional geological and mineralogical mapping of the territory of southeastern Zabaykalsky krai (Russia) under conditions of medium–low mountain relief and continental climate. To increase the spectral information content of Landsat-8 satellite data, statistical processing algorithms are used, including principal component analysis (PCA), minimum noise fraction (MNF), and independent component analysis (ICA). Statistical processing results are compared with the classes of geological materials: minerals of the oxide/hydroxide group containing transitional iron ions (Fe 3+ and Fe 3+ /Fe 2+ ); a group of clay minerals containing Al-OH and Fe, Mg-OH; and minerals containing divalent iron ion (Fe 2+ ) and vegetation cover. Pseudocolor RGB composites are generated and interpreted, reflecting the distribution and multiplication of classes of geological materials. To build a prospectivity scheme for the discovery of the minerals, the integration of informative thematic layers is carried out using a fuzzy logic model. The resulting scheme is compared with geological information, and positive conclusions are made about the suitability of territory for further remote studies on mapping hydrothermally altered zones and hypergenesis products in order to localize areas that are promising for revealing hydrothermal–metasomatic mineralization.