2007 GSA Denver Annual Meeting (28–31 October 2007)

Paper No. 47
Presentation Time: 8:00 AM-12:00 PM

REEXAMINATION OF THE 26.5 KA ORUANUI ERUPTION, TAUPO VOLCANO, NEW ZEALAND BY POLYTOPIC VECTOR ANALYSIS


TEFEND, Karen S.1, VOGEL, Thomas A.2, WILSON, Colin J.N.3, DEERING, Chad D.4, DONATELLE, Angela2 and HILDALGO, Paulo2, (1)Department of Geosciences, University of West Georgia, 1601 Maple Street, Carrollton, GA 30118, (2)Geological Sciences, Michigan State University, 206 Natural Sciences Bldg, East Lansing, MI 48824, (3)Department of Geology, Auckland University, Private Bag 92019, Auckland, 1020, New Zealand, (4)Department of Geological Sciences, University of Canterbury, Private Bag 4800, Christchurch, New Zealand, ktefend@westga.edu

Polytopic Vector Analysis (PVA) is the multivariate analytical approach of unmixing a dataset represented in this case by individual pumice clasts from volcanic deposits, allowing the interpretation of magmatic processes that occur prior to eruption. The 26.5 ka Oruanui ignimbrite, located in the Taupo Volcanic Zone of the North Island of New Zealand, is a well studied volcanic deposit that serves as an example of the advantage of multivariate analysis of pumice clasts. Previous studies of the Oruanui ignimbrite have established the presence of mafic, silicic (>71.7 wt% SiO2), and banded pumice clasts of intermediate composition; the chemical and mineralogical variation in these pumice clasts have been attributed to open-system processes such as magma mixing and assimilation of basement rock by mafic magma (Wilson et al., 2006). Based on major and trace element bivariate plots, a subgroup of silicic (rhyolite) pumice clasts with a high Mg/FeOt composition was identified and interpreted as the result of mixing between the mafic magmas and the more evolved rhyolite magmas at some point prior to eruption. Reexamination of the geochemical data by multivariate analysis of these pumice clasts using PVA yields an interpretation consistent with that of Wilson et al. (2006); thus, we interpret magma mixing between the rhyolite and mafic magmas. However, PVA has allowed us to refine the earlier interpretation; specifically, the ‘high-Mg group' of rhyolite pumice clasts defined by Wilson et al. (2006) is not recognized by PVA. Instead multivariate diagrams exhibit a new subgroup of rhyolite samples, referred to as the ‘PVA-generated subgroup'. Because of the identification of this subgroup by PVA, revised interpretations can be made; at least two mixing events occurred between the mafic and silicic magmas, one which occurred prior to eruption and explains in part the geochemical variation seen in the silicic pumice clasts, and a later mixing event which produced the banded pumice clasts. Furthermore, PVA results identify another end member present, which we interpret as assimilation of a Zr-rich, Ba-depleted crustal rock. This possible assimilant identified by polytopic vector analysis of the rhyolite samples was undetected by previous conventional analysis of silicic pumice clast geochemistry.