Reentrants: Reliable Recorders of Magma Decompression Rates?

Abstract:

We present a new experimental method to assess the fidelity of reentrants (unenclosed melt inclusions) as recorders of magma decompression history. During magma decompression, the solubilities of H₂O and CO₂ decrease in the melt, causing these volatile species to diffuse out of reentrants and into bubbles in the host magma. Under explosive ascent conditions (<10 minutes to a few hours), quenched reentrants preserve volatile concentration gradients that can be measured and modeled to infer decompression rates. This approach has been applied to natural samples but characterized by variable degrees of misfit between measured and modeled gradients. An experimental method is therefore befitting to determine the conditions under which reentrants reliably record decompression rates.

Our experimental method involves loading reentrant-bearing quartz crystals and ash from the Bishop Tuff with buffering components into gold capsules and subjecting them to known temperature-pressure-time paths using a cold-seal pressure vessel. Each experiment begins with a dwell period at 150 MPa to re-equilibrate the reentrants to storage conditions, followed by constant isothermal decompression. The first of nine preliminary experiments involved a dwell period followed by rapid quenching to confirm re-equilibration at 150 MPa, yielding starting conditions of ~2.3-2.5 wt. % H₂O and ~650-750 ppm CO₂. Two successful experiments with reentrants recovered, measured, and modeled were run at a constant decompression rate of 0.008 MPa s^-1, and the 1D diffusion model retrieves experimental conditions (within error). To better mimic the behavior of magma as it approaches the surface, we then experimentally simulated accelerating decompression, shifting the rate from 0.008 to 0.03 MPa s^-1 at 80 MPa. The best-fit model using the known starting conditions yields the slower, initial decompression rate (0.008 MPa s^-1), while using reentrant interior H₂O and CO₂ concentrations as starting conditions yields the faster, final rate (0.03 MPa s^-1). These results reveal the limitation of the current code in modeling more complex magma decompression scenarios. The goal of planned, future experiments is to further test how reentrants preserve accelerating decompression rates, and to guide revision of the current 1D diffusion code.