Reliability of Raman analyses of CO2-rich fluid inclusions as a geobarometer at Kīlauea

Abstract

Interpreting signals of volcanic unrest requires knowledge of the architecture of the magmatic system, particularly the depths at which magmas are stored. Such information can be vital to help predict changes in eruptive style and vigour. However, popular petrological tools to assess magma storage depths (e.g., melt inclusions) are costly, present large uncertainties, and are too slow for real time monitoring. Here, we evaluate the reliability of Raman Spectroscopy measurements of CO 2dominated fluid inclusions as a geobarometer relative to microthermometry and melt inclusion barometry. We calculate storage pressures for 102 olivine-hosted fluid inclusions from the 2018 Lower East Rift Zone eruption of Kīlauea, which are statistically indistinguishable to those determined from melt inclusions. We show that calibrated Raman spectroscopy yields densities within 5-10 % of microthermometry for CO 2 -dominated fluid inclusions (<10 mol % H 2 O) but is a far more suitable method for systems like Kīlauea dominated by shallow magma storage. Overall, pressures determined from fluid inclusions by Raman spectroscopy are robust and require only a fraction of the time and resources of melt inclusion studies.

DiadFit: An open-source Python3 tool for peak fitting of Raman data from silicate melts and CO2 fluids

“…For example, to align with FAIR principles (Findable, Accessible, Interoperable, Reproducible), users can publish a supporting folder containing their spectra, relevant metadata, and Jupyter Notebooks showing exactly how spectra were fitted and subsequent calculations performed [e.g. Dayton et al 2023;DeVitre and Wieser 2024].…”

“…Fitting these spectra took ∼8 hrs using Fityk, and ∼15 minutes using DiadFit on a typical laptop with 16 GB of RAM and an i7 processor. Given the potential for fluid inclusion analyses to provide rapid estimates of magma storage depths during volcanic crises [Dayton et al 2023;DeVitre and Wieser 2024;Zanon et al 2024], it is vital to speed up data processing as much as possible to reap the full benefits of this speedy technique. We anticipate that users who are not familiar with Python will simply use the provided Jupyter Notebooks and narrated YouTube videos, changing simple parameters like the path to their files and peak fit parameters to adjust for the different appearance of spectral peaks on different Raman instruments.…”

“…DeVitre and Wieser [2024] implement a Python3 version of the mechanical re-equilibration model for olivine-hosted fluid inclusions of Wanamaker and Evans [1989] in their package RelaxiFI. We incorporate this methodology into DiadFit for ease of installation and maintenance, because it uses the underlying EOS and density-depth conversions from DiadFit to track changes in fluid inclusion volume, density, and pressure for different stalling, ascent, and quenching paths.…”

“…Since 2014, there has been a growing body of literature using Raman Spectroscopy to measure the density of CO 2 -rich fluids in melt inclusion vapour bubbles to more accurately obtain the total CO 2 content of the melt inclusion, and thus the magma storage depth [Hartley et al 2014;Moore et al 2015;Lamadrid et al 2017;Allison et al 2021;Wieser et al 2021;DeVitre et al 2023b]. Raman spectroscopy also shows enormous potential to quantify the densities of CO 2 -rich fluid inclusions [Wang et al 2011;Kobayashi et al 2012], allowing rapid and precise estimates of magma storage depths [Dayton et al 2023;DeVitre and Wieser 2024;Zanon et al 2024].…”