226Radium-230Thorium disequilibrium in plagioclase from recent eruptions at Mount Hood volcano, Oregon: Constraints on crystal residence and magma storage timescales

Uranium-series crystal dating methods have been used to quantify the temporal and geochemical histories of igneous mineral phases, providing insight into the sub-surface histories of the magmas that produced them. In particular, 226Ra-230Th disequilibrium of bulk plagioclase separates can be used to constrain timescales of crystal residence and magma storage. Crystal size distributions (CSDs) have also been used for this purpose, and combining the two methods can reveal information about magmatic processes that neither can provide in isolation.;Mount Hood is a dominantly andesitic to dacitic stratovolcano characterized by abundant evidence for magma mixing. Two populations of plagioclase have been recognized in nearly all eruptive sequences and have been geochemically-linked to crystallization in mafic and silicic magmas prior to mixing. The two plagioclase populations are distinguishable on the basis of size, providing the opportunity to physically separate and date the populations using uranium-series bulk mineral dating methods.;In this study, we measured uranium-series element concentrations in bulk plagioclase separates of different sizes from the Timberline (eruption age 1,500 a) and Old Maid (eruption age 215 a) eruptive sequences. We calculated 226Ra-230Th bulk plagioclase crystallization ages of eruption age to 3,500 a for plagioclase from the mafic end-member magma, and ages of 8,000 to >10,000 a for plagioclase from the silicic end-member magma. These ages are consistent with a mafic recharge and mixing model for the production of intermediate Mount Hood magmas. The mafic recharge magmas were transient features in the sub-surface, stored in the crust for less than 3 ky before eruption. These magmas interacted with a long-lived silicic magma stored in a shallow-crustal reservoir over timescales greater than 7 ky before eruption.;Comparing plagioclase age data as measured by 226Ra- 230Th disequilibrium and CSD reveals that plagioclase crystals from the silicic reservoir grew for only short periods (maximum of 0.15%) of their overall average histories. These crystals must have experienced episodes of growth hiatus and/or dissolution, consistent with residence in a liquid-poor silicic crystal mush in the shallow crust, which may have spent long periods of its history completely crystalline.