Spatial Component Analysis to Improve Mineral Estimation Using Sentinel-2 Band Ratio: Application to a Greek Bauxite Residue

Remote sensing can be fruitfully used in the characterization of metals within stockpiles and tailings, produced from mining activities. Satellite information, form band ratio, act as an auxiliary variable, with a certain correlation ground primary data. In presence this modeled nested structures, spatial components without filtered out, so that useful data grows. This paper investigates possibility to substitute co-kriging system, whole ratio only correlated components. The method has been applied over bauxite residues case study presents three estimation alternatives: ordinary kriging, co-kriging, component co-kriging. Results have shown how using most reduces variance improves results. general terms, when good samples exists, satellite band-ratio Component reconstruction mineral grade distribution, thus affecting selectivity. On other hand, use approach exalts distance variability.