A Combination of Geostatistical Methods and Principal Components Analysis for Detection of Mineralized Zones in Seafloor Hydrothermal Systems

Seafloor massive sulfide (SMS) deposits are typical of submarine mineral resources and generally rich in base metals (Cu, Pb, Zn); however, their distribution, configuration, and formation mechanism, especially sub-seafloor mineralization, remain poorly understood because of scant drilling and geophysical data. To address this problem, this study aims to identify and characterize mineralized zones in seafloor hydrothermal areas using limited metal content data from sparse drilling sites. We use principal component analysis to decrease the dimensionality of the content data. High metal content zones are delineated using principal component values by three geostatistical methods: (1) spatial estimation using ordinary kriging; (2) turning bands simulations (TBSIM); and (3) sequential Gaussian simulations. We selected an active seafloor vent area at 1570 m below sea level in the Okinawa Trough, southwest Japan, as a case study. Results from the three methods show two types of high metal content zones: One is around a sulfide mound, and the other is layered in association with lateral flow of hydrothermal fluids from the bottom of the mound. TBSIM is the most effective under scarce data conditions because the model yields the smallest cross-validation error, decreases the smoothing effect, and corresponds well to a conceptual deposit model that shows a stockwork below the sulfide mound. The results contribute to better understanding the formation mechanism of SMS deposits as well as constraining submarine metal reserves and mining.