DEVELOPING RHYOLITE-MELTS GEOBAROMETRY FOR MAFIC VOLCANIC SYSTEMS

Geobarometers are critical for constraining magmatic storage depths at volcanoes. One of the most commonly applied geobarometers in olivine-bearing systems is melt inclusion (MI) vapor saturation pressures. This approach is costly and time consuming, and volatile contents of MIs can be altered by post-entrapment processes. I aim to calibrate the rhyolite-MELTS geobarometer to estimate storage depths from the major element content of matrix glasses in volcanic systems with olivine, plagioclase, ± clinopyroxene in equilibrium with melt.

I targeted basaltic to basanitic glass compositions with independently determined pressure, H2O, and fO2 constraints from samples containing olivine, plagioclase, ± clinopyroxene. Initial isobaric crystallization models on natural matrix glass (Ross Island, Fo75-90) crystallized olivine with Fo40-50 and only crystallized plagioclase at low water content (0.5-1wt%). Additionally, pressure calculations were shallow relative to those calculated from melt inclusion vapor saturation pressures (99-350 vs 400). Matrix glass from Cero Negro did not replicate olivine compositions properly (Fo20 vs Fo70-81, respectively). All but one Cerro Negro olivine-hosted MI composition failed to crystallize olivine. All MI compositions were corrected for post-entrapment crystallization (PEC), but notably, the one MI composition that did successfully replicate the olivine and plagioclase compositions (Fo78 & An80+) did not require any PEC correction . Models run on glass compositions from phase-equilibria experiments of Medicine Lake basalts (>Fo86.5 & An93) produced both olivine and plagioclase compositions consistent with experimental samples, but the pressure estimated by the barometer is shallower than the experimental pressure (50-75 MPa vs 200 MPa).

Our preliminary results are consistent with other studies that suggest rhyolite-MELTS geobarometry results are sensitive to the input composition. Thus, alteration of materials or incorrectly estimated PEC corrections could result in an inappropriate input composition. The discrepancy in the pressure results from the rhyolite-MELTS geobarometer and phase-equilibria experiments indicates further work must be done to develop a proper mode of mafic geobarometry using olivine, plagioclase, ± clinopyroxene.