Introducing BurnMan 1.0.1

Overview

BurnMan is an open source mineral physics and seismological toolkit written in Python which can enable a user to calculate (or fit) the physical and chemical properties of endmember minerals, fluids/melts, solid solutions, and composite assemblages.

Properties which BurnMan can calculate include:

  • the thermodynamic free energies, allowing phase equilibrium calculations, endmember activities, chemical potentials and oxygen (and other) fugacities.

  • entropy, enabling the user to calculate isentropes for a given assemblage.

  • volume, to allow the user to create density profiles.

  • seismic velocities, including Voigt-Reuss-Hill and Hashin-Strikman bounds and averages.

The toolkit itself comes with a large set of classes and functions which are designed to allow the user to easily combine mineral physics with geophysics, and geodynamics. The features of BurnMan include:

  • the full codebase, which includes implementations of many static and thermal equations of state (including Vinet, Birch Murnaghan, Mie-Debye-Grueneisen, Modified Tait), and solution models (ideal, symmetric, asymmetric, subregular).

  • popular endmember and solution datasets already coded into burnman-usable format (including [HollandPowell11], [SLB05] and [SLB11])

  • Optimal least squares fitting routines for multivariate data with (potentially correlated) errors in pressure and temperature. As an example, such functions can be used to simultaneously fit volumes, seismic velocities and enthalpies.

  • a “Planet” class, which self-consistently calculates gravity profiles, mass, moment of inertia of planets given the chemical and temperature structure of a planet

  • published geotherms

  • a tutorial on the basic use of BurnMan

  • a large collection of annotated examples

  • a set of high-level functions which create files readable by seismological and geodynamic software, including: Mineos [MWF11], AxiSEM [NissenMeyervanDrielStahler+14] and ASPECT

  • an extensive suite of unit tests to ensure code functions as intended

  • a series of benchmarks comparing BurnMan output with published data

  • a directory containing user-contributed code from published papers

BurnMan makes extensive use of SciPy, NumPy and SymPy which are widely used Python libraries for scientific computation. Matplotlib is used to display results and produce publication quality figures. The computations are consistently formulated in terms of SI units.

The code documentation including class and function descriptions can be found online at http://burnman.readthedocs.io.

This software has been designed to allow the end-user a great deal of freedom to do whatever calculations they may wish and to add their own modules. The underlying Python classes have been designed to make new endmember, solid solution and composite models easy to read and create. We have endeavoured to provide examples and benchmarks which cover the most popular uses of the software, some of which are included in the figure below. This list is certainly not exhaustive, and we will definitely have missed interesting applications. We will be very happy to accept contributions in form of corrections, examples, or new features.

Structure

_images/structure.png

Installation

Requirements

  • Python 3.7+

  • Python modules: NumPy, SciPy, SymPy, Matplotlib

Source code

The source code can be found at https://github.com/geodynamics/burnman.

Install under Ubuntu

  1. Install dependencies using apt by opening a terminal window and entering sudo apt-get install python python-scipy python-numpy python-sympy python-matplotlib git

  2. Clone the BurnMan repository git clone https://github.com/geodynamics/burnman.git

  3. Go to the Burnman examples directory and type: python example_beginner.py Figures should show up, indicating that it is working.

Install on a Mac

  1. get Xcode

  2. If you don’t have Python yet, download it (for free) from python.org/download . Make sure to use Python 3.7+. To check your version of python, type the following in a terminal: python --version

  3. Install the latest Numpy version from http://sourceforge.net/projects/numpy/files/NumPy/

  4. Install the latest Scipy from http://sourceforge.net/projects/scipy/files/

  5. Install the latest Sympy from http://sourceforge.net/projects/sympy/files/

  6. Install the latest Matplotlib from http://sourceforge.net/projects/matplotlib/files/matplotlib/matplotlib-1.1.1/

  7. Clone the BurnMan repository git clone https://github.com/geodynamics/burnman.git

  8. Go to the Burnman examples directory and type python example_beginner.py Figures should show up, indicating that it is working.

Install under Windows

To get Python running under Windows:

  1. Download Python from http://www.python.org/ and install

  2. Go to http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy, download and install

  3. Go to http://www.lfd.uci.edu/~gohlke/pythonlibs/#scipy, download and install

  4. Go to http://www.lfd.uci.edu/~gohlke/pythonlibs/#sympy, download and install

  5. Go to http://www.lfd.uci.edu/~gohlke/pythonlibs/#matplotlib, download and install

  6. Download BurnMan from github (https://github.com/geodynamics/burnman)

  7. Open Python Shell (IDLE Python GUI)

  8. File – Open – find one of the example files

  9. Run the module (or press F5)

Citing BurnMan

If you use BurnMan in your work, we ask that you cite the following publications:

  • Myhill, R., Cottaar, S., Heister, T., Rose, I., and Unterborn, C. (2021): BurnMan v1.0.1 [Software]. Computational Infrastructure for Geodynamics. Zenodo. (https://doi.org/10.5281/zenodo.5552756)

  • Cottaar S., Heister, T., Rose, I., and Unterborn, C., 2014, BurnMan: A lower mantle mineral physics toolkit, Geochemistry, Geophysics, and Geosystems, 15(4), 1164-1179 (https://doi.org/10.1002/2013GC005122)

Contributing to BurnMan

If you would like to contribute bug fixes, new functions or new modules to the existing codebase, please contact us at info@burnman.org or make a pull request at https://github.com/geodynamics/burnman.

BurnMan also includes a contrib directory that contains python and ipython scripts used to reproduce published results. We welcome the submission of new contributions to this directory. As with the contribution of code, please contact us at info@burnman.org or make a pull request at https://github.com/geodynamics/burnman.

Acknowledgement and Support

  • This project was initiated at, and follow-up research support was received through, Cooperative Institute of Deep Earth Research, CIDER (NSF FESD grant 1135452) – see www.deep-earth.org

  • We thank all the members of the CIDER Mg/Si team for their input: Valentina Magni, Yu Huang, JiaChao Liu, Marc Hirschmann, and Barbara Romanowicz. We also thank Lars Stixrude for providing benchmarking calculations and Zack Geballe, Motohiko Murakami, Bill McDonough, Quentin Williams, Wendy Panero, and Wolfgang Bangerth for helpful discussions.

  • We thank CIG (www.geodynamics.org) for support and accepting our donation of BurnMan as an official project.