Geoinformatics Facility: Integration of alphaMELTS petrologic software with flexible modeling environments

California Institute Of Technology, Pasadena CA

Investigators

Abstract

Scientists investigate the generation and evolution of magma (molten rock) with a range of tools such as field-based studies, experiments at high pressures and temperatures, and models of the physical and chemical behavior of partially molten rocks. The modeling component is particularly important for situations where field work is too hazardous or inaccessible and where it would take too many experiments to fully describe the processes leading to a catastrophic volcanic eruption. This work focuses on continuing development of a flexible, powerful, and easy-to-use package of modeling software used by hundreds of Earth scientists. The alphaMELTS 2 software can model complex scenarios with changing pressure, temperature, and chemical composition while monitoring energy and mass of major and trace elements and gaseous species. This project will increase the mineral systems alphaMELTS 2 can model, support users, extend applications of the software in teaching and training, and facilitate outreach workshops to train new users. These workshops will reach several dozen users, mostly early-career scientists, and equip them with quantitative tools for modern petrology research. There are several particular advances on which this effort will focus, mostly new thermodynamic model calibrations for use with alphaMELTS 2 and related tools: modeling of CO2 in mantle magmatic systems; the amphibole mineral group; the biotite mineral group; more flexible treatment of trace elements; better integration with MATLAB, Python, and other development environments; and user interaction and training through the online forum and dedicated training workshops. Extension of the calibration strategy to exploit new categories of experiments such as two-liquid equilibria, will allow the pMELTS code for high-pressure melting to describe carbon-bearing systems such as subducted carbonate or upwelling carbonated mantle rocks. The most widespread and important igneous minerals that are not well-modeled in current generations of MELTS are amphibole and biotite; a new strategy will be implemented that can tackle the great complexity of naturally occurring amphiboles and make an accurate yet tractable calibration. Dynamically loadable libraries that allow for seamless incorporation of all alphaMELTS 2 functionality into any Fortran, C++, Matlab or Python code will enable coupling with other codes and flexible scripting. Finally, a large expansion of the phases and trace elements that can be treated with variable and modular partition coefficients will allow users to customize their trace element calculations for greater accuracy. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.