Title

Compositional variation within thick (>10 m) flow units of Mauna Kea Volcano cored by the Hawaii Scientific Drilling Project

Document Type

Article

Abstract

Geochemical analyses of stratigraphic sequences of lava flows are necessary to understand how a volcano works. Typically one sample from each lava flow is collected and studied with the assumption that this sample is representative of the flow composition. This assumption may not be valid. The thickness of flows ranges from <1 to >100 m. Geochemical heterogeneity in thin flows may be created by interaction with the surficial environment whereas magmatic processes occurring during emplacement may create geochemical heterogeneities in thick flows. The Hawaii Scientific Drilling Project (HSDP) cored ∼3.3 km of basalt erupted at Mauna Kea Volcano. In order to determine geochemical heterogeneities in a flow, multiple samples from four thick (9.3–98.4 m) HSDP flow units were analyzed for major and trace elements. We found that major element abundances in three submarine flow units are controlled by the varying proportion of olivine, the primary phenocryst phase in these samples. Post-magmatic alteration of a subaerial flow led to loss of SiO2, CaO, Na2O, K2O and P2O5, and as a consequence, contents of immobile elements, such as Fe2O3 and Al2O3, increase. The mobility of SiO2 is important because Mauma Kea shield lavas divide into two groups that differ in SiO2 content. Post-magmatic mobility of SiO2 adds complexity to determining if these groups reflect differences in source or process. The most mobile elements during post-magmatic subaerial and submarine alteration are K and Rb, and Ba, Sr and U were also mobile, but their abundances are not highly correlated with K and Rb. The Ba/Th ratio has been used to document an important role for a plagioclase-rich source component for basalt from the Galapagos, Iceland and Hawaii. Although Ba/Th is anomalously high in Hawaiian basalt, variation in Ba abundance within a single flow shows that it is not a reliable indicator of a deep source component. In contrast, ratios involving elements that are typically immobile, such as La/Nb, La/Th, Nb/Th, Ce/Pb, Sr/Nd, La/Sm, Sm/Yb, Nb/Zr, Nb/Y and La/Yb, are uniform within the units, and they can be used to constrain petrogenetic processes. Nevertheless all elements are mobile under some conditions. For example, a surprising result is that relative to other samples, the uppermost sample collected from subaerial flow Unit 70, less than 1 m below the flow surface, is depleted in P, HREE and Y relative to all other samples from this flow unit. This result is complementary to the P, REE and Y enrichment found in subaerial lava flows from several Hawaiian shields, e.g., Kahoolawe and Koolau Volcanoes. These enrichments require mobilization of REE and followed by deposition a P-rich mineral. © 2016 Elsevier Ltd