The influence of magma ascent path on the texture, mineralogy, and formation of hornblende reaction rims
Although hornblende reaction rims are widely used as a tool for evaluating magma ascent during volcanic eruptions, very few studies constrain the manner in which they form. This study investigates the influence of magma ascent path on the formation of hornblende reaction rims. To do this, we conducted isothermal (840 °C) decompression experiments using dacite pumice samples erupted in December 1989 from Redoubt volcano, Alaska. Experiments were first held within the hornblende stability field determined through phase equilibria experiments for 3 to 5 days before being decompressed to different pressures ranging from 100 to 2 MPa for 1 to 30 days before being quenched. Decompression was performed in either multiple, equal steps (constant rate) or in one single step. Results from multi-step experiments show that reaction rims form preferentially at pressures from 10 to 40 MPa, and that this favorable pressure range narrows and decreases with increased decompression duration. Hornblendes in multi-step experiments are tightly enclosed by fine-grained reaction rims composed of plagioclase and orthopyroxene crystals with high aspect ratios ranging from 1 to 12. For single-step experiments, reaction rims also form preferentially within a narrow pressure range (60-70 MPa), where they consist of medium-grained plagioclase, titanomagnetite, and orthopyroxene crystals with aspect ratios ranging from 1 to 6 that broadly surround subrounded hornblendes. Hornblendes from single-step experiments dropped to lower pressures (< 40 MPa), however, are typically euhedral and tightly enclosed by fine-grained reaction rims composed of plagioclase and orthopyroxene with aspect ratios ranging from 1 to 12. Little or no titanomagnetite is observed in these rims. Reaction rim growth was not observed at pressures below 10 MPa, regardless of decompression style or experiment duration, suggesting that hornblende in magma stored at very shallow depth (< 200 m) will not develop reaction rims due to the limited hornblende dissolution combined with extremely high viscosity of the near-solidus interstitial melt. Finally, we describe what is expected from several different magma ascent paths with respect to the texture, mineralogy, and thickness of hornblende reaction rims, as well as changes in hornblende modal abundance. Observations from dome samples emplaced during the 1989-1990 eruption of Redoubt are consistent with the mixing of slowly rising and intermittently stalling batches of magma with lesser amounts of fresh, rapidly ascending magma. This interpretation is also consistent with seismic observations from the eruption.