Petrologic Controls on Explosive Eruptions from the Pacific-Antarctic Ridge

Monday, 30 January 2017
Marina/Gretel (Hobart Function and Conference Centre)
Madeline Lewis1, Paul D Asimow1 and David C Lund2, (1)California Institute of Technology, Pasadena, CA, United States, (2)University of CT at Avery Point, Groton, CT, United States
Abstract:
Intervals in sediment core OC170-26-159, retrieved from 14km west of the Pacific-Antarctic Ridge (PAR) at 39°S, contain abundant basaltic glass shards deposited at least 7km from their axial source. We interpret these shards as the products of large explosive submarine eruptions, with volcanic activity peaking at 148 and 130ka; the latter concurrent with glacial termination T-II and enhanced hydrothermal activity on the East Pacific Rise (Lund et. al, 2016). Using our analyses of the glass, we evaluate whether sea level driven changes in magmatic flux may be related to the frequency of large explosive submarine eruptions through emplacement of distinct compositions or volatile contents.

Major element compositional data show that the glasses fall within a narrow compositional range for all major components, with MgO ranging from 5.8 to 6.5wt%. The glasses are evolved relative to effusively erupted basalts on the PAR axis (PetDB) and are slightly enriched in H2O, yet appear consistent with the same liquid line of descent. H2O and CO2 saturation pressures range from seafloor pressure near the ridge axis to 860 bars (approximately 1.8km crustal depth). Using our maximum measured saturation pressure and fO2 values consistent with MORB upper mantle conditions, we implemented the MELTS thermodynamic model (Ghiorso and Sack 1995, Asimow and Ghiorso 1998) to successfully predict the composition and volatile content of the glass shards through fractional crystallization of an axial magma. Fractionation from the most primitive axial composition results in a viscosity increase from 101.37 Pa×s to 102.88 Pa×s. Our modeling confirms the relation of the shards to the axial magmatic system and suggests that the explosivity was caused by higher viscosity in the fractionated melt and variations in magmatic flux over time. The periods of energetic eruptions centered around 148 and 130ka require high magmatic flux to the ridge axis, while a period of fractionation between the explosive intervals indicates low magmatic flux. We propose that sea level changes alter the magmatic flux and, consequently, the eruption styles.