Akanvaara reciprocal mixing model

Abstract

Akanvaara reciprocal mixing model Publication code version: v1.0.0-akanvaaraDOI: 10.5281/zenodo.19823404 This Zenodo bundle contains only three files:- akanvaara_reciprocal_mixing.py- plot_reciprocal_mixing_dashboards.py- README.txt The modelling script is self-contained with respect to the parent meltcompositions. The starting compositions for `Loljunmaa` and `O16` are stored asdictionaries inside `akanvaara_reciprocal_mixing.py`. When the model is run, it first generates parent fractionation paths internallywith MAGEMin/PetThermoTools, writes those paths into`reciprocal_mixing_results/parent_fractionation_paths/`, samples evolved parentliquids from them, and then runs the reciprocal mixing grid. Environment - Python 3.11- Python packages: numpy, pandas, matplotlib, petthermotools- Julia 1.11.x- Thermodynamic backend: MAGEMin/Green2025 through the Julia environment used by PetThermoTools Model setup - Pressure is controlled by `P_BAR` in the modelling script; the default is 2000 bar- Thermodynamic model: `Green2025` through PetThermoTools- Redox: `FMQ`, offset `0`- Parent paths are generated by liquidus search from `1400 C`, then fractional crystallisation down to `800 C`- Parent liquid sampling grid: nominal `1200 to 800 C` in `50 C` steps- Fresh-melt additions: `25`, `50`, and `75 wt%`- Cr-spinel threshold: `15 wt% Cr2O3` The grid metric is cumulative Cr-spinel mass in grams across the wholecrystallisation path. Running Run the model:python3 akanvaara_reciprocal_mixing.py Then regenerate the figures:python3 plot_reciprocal_mixing_dashboards.py The Python scripts were developed with assistance from OpenAI Codex, then reviewed, modified and tested by the authors.