Discrimination and identification of ferric oxides using satellite Thematic Mapper data: A Newman case study

Abstract A technique is described for identifying and mapping hematite and goethite on Thematic Mapper (TM) data over arid and semi-arid terrains. The method involves directing a principal components analysis at two selected input ratios, each chosen because of its effectiveness at highlighting and/or separating hematite, goethite and vegetation, in n-dimensional ratio-versus-ratio space. Once the relative positions of hematite, goethite and vegetation are established—a scene-independent criterion—the directions of the resulting eigenvectors indicate which component contains the hematite-goethite information and which is related to vegetation.