Abstract
Bentonites serve as an important stratigraphic tool for facies correlation due to their instantaneous deposition and preservation across multiple depositional environments. Correlating one bed to another can be a complex issue limited by the 3-dimensional availability of outcrops and core. This process is reliant on the assumption that the geochemical signatures of ash beds do not vary enough to statistically alter the correlation coefficients that tie one ash-fall to another. This assumption may be faulty, however, and the problem compounds as bentonite bed thickness increases because the chemical composition varies through an eruption due to stratification of the magma chamber and through physical transport. Linear discriminant analysis (LDA) can significantly aid with this issue, allowing for the utilization of a multivariate dataset to statistically separate chemically-similar groups. Here we present a statistical approach to assessing the primary discriminatory variables on a suite of 87 Cretaceous bentonites deposited in the Bighorn Basin, Wyoming, USA, that span five geologic stages and seven formations, from the Cloverly Formation (110 Ma) to the Meeteetse Formation (70 Ma), and includes nine named ash beds. Using LDA, bentonites were sorted into groups based on geologic stage, formation, and unique beds using key elemental concentrations and radiogenic isotopes (e.g., 87Sr/86Sr, TiO2, Sc, Na, and Th). Additionally, distinct similarities in the discrimination pattern between formations reveal two periods of magmatism, the development of the Idaho batholith and the migration and development of the Boulder batholith and surrounding related plutons. With well-preserved, temporally-constrained ash, LDA can correlate ancient tephra deposits while also shedding light on the cyclical nature of geochemical trends in a migrating subduction zone.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have