SUPERSATURATION NUCLEATION AND GROWTH OF PLAGIOCLASE IN RESPONSE TO ERUPTIVE DECOMPRESSION (Invited Presentation)

Magma is forced into disequilibrium during decompression and returns to equilibrium through nucleation of microlites and growth on preexisting crystals. The resulting textures can be used to infer the decompression pathway of explosive and effusive eruptions, especially when compared with high temperature experiments. We performed a series of experiments using starting material from the climactic 1991 Pinatubo eruption. We decompressed experiments using continuous rates ranging from 100 to 0.3 MPa h^-1. Plagioclase microlite number densities range from 10⁷ to 10⁸ cm^-3. Plagioclase microlite area increased systematically from ~20 μm² to ~1000 μm² with increasing experiment duration. Average nucleation and areal growth rates of plagioclase are highest in the fastest decompressions (~10^7.5 cm^-3 h^-1 and ~10 μm² h^-1, respectively) and more than an order of magnitude lower in the slowest experiments (~10^5.5 cm^-3 h^-1 and ~0.8 μm² h^-1, respectively). We used these results to develop Supersaturation Nucleation and Growth of Plagioclase (SNGPlag), a numerical model that predicts the evolution of plagioclase crystallization in a decompressing magma. SNGPlag uses the MELTS web service to determine equilibrium plagioclase mode as a function of pressure and temperature. At each time step the model evaluates the difference between the calculated crystallinity and equilibrium crystallinity for a given pressure and temperature to determine the degree of supersaturation, which then sets plagioclase nucleation and growth rates. Growth rates are used to grow existing crystals whereas nucleation rates add new microlites. SNGPlag is thus experimentally and thermodynamically calibrated and produces results that can be used as a tool to quantify textures in natural volcanic rocks. We used SNGPlag to reproduce textures observed from the 2008 Chaitén, 1902 Santa Maria, 1980-1986 Mt St Helens, and 1800 BP Ksudach eruptions. Textural similarity to natural samples, or lack thereof, isolates the complex interplay and relative importance of variables and changing conditions. SNGPlag commonly predicts decompression rates that are slower than those indicated by previous work, likely caused by reliance on stepped decompressions or for failing to account for the effects of preexisting crystals on microlite nucleation and growth.