ESTIMATING WATER CONTENTS OF BISHOP TUFF MAGMA FROM HYDROXYL CONCENTRATIONS IN FELDSPAR PHENOCRYSTS

The Bishop Tuff resulted from the largest of >200 eruptions in the Long Valley volcanic field over the past 4.5 Ma (Hildreth 2004). Roughly 766 ky ago, >600 km³ of rhyolitic magma was explosively erupted as pyroclastic fall and flow deposits over a period of approximately six days (Wilson & Hildreth 1997). The primary trigger of such explosive supereruptions is often considered to be magmatic overpressure due to the exsolution of volatiles from magma as it is stored and evolves. However, external tectonic triggers such as earthquakes and faulting can also initiate supereruptions by fracturing the walls or roof of the magma chamber, effectively destabilizing it. By measuring structural hydroxyl (OH) concentrations in plagioclase phenocrysts, magmatic water contents during the latest stages of accumulation, ascent and eruption can be constrained (Johnson & Rossman 2013). Individual phenocrysts were picked from the Bishop fall deposits (which are divided into units F1−F9), and were analyzed using a Fourier Transform Infrared Spectrometer (FTIR) at JMU. Polarized infrared spectra of absorption were obtained from each phenocryst and the OH concentration (in ppm) was calculated using the modified Beer−­Lambert Law and the calibration of Mosenfelder et al. (2015). Preliminary results show OH concentrations of 40−88 ppm in plagioclase phenocrysts from the earliest erupted material (F1) and 57−60 ppm in plagioclase phenocrysts from the last fall deposit (F9). Additional analyses of F1 plagioclase phenocrysts will constrain the range of OH concentrations in this unit. We will use our results together with melt inclusion data for the H₂O concentrations in F1 and F9 magma (Wallace et al. 1999; Myers, unpublished data) and similar data from the Huckleberry Ridge Tuff (Yellowstone) to calibrate an empirical relationship between magmatic water concentration and OH in plagioclase. Using this calibration, magmatic water concentration can be determined from OH in plagioclase even in samples lacking pristine melt inclusions.