The manner in which magma moves through the Earth is incomplete understood, partly because we currently lack computer models that can simulate the full magmatic system, from its origin in the mantle (at depths > 150 km) until its arrival in the magma chambers underneath volcanoes at a few kilometers depth.

MAGMA is a Consolidator Grant research project funded by the European Research Council. In this project, a team of researchers, are working on developing new multi-phase and multi-chemistry computer models to simulate how magma migrates from the mantle to the upper crust. By comparing the results of such computer simulations with the wealth of existing petrological, geochemical and geophysical data, we hope to obtain new insights in the physics of magmatic processes. The project started in October 2018 and will run until September 2023.



A typical problem in geodynamic modelling of natural systems is to create model setups that are consistent with a wide range of published geological and geophysical data. In some cases, seismic tomography models are available in digital format, but in many other cases not and we are left with cross-sections in published papers.

Recent advances in modelling magmatic systems

Tobias Keller laid the foundations for the MAGMA project with the work in his PhD thesis. After postdocs in Oxford and Stanford he is now a lecturer in Glasglow and in this outreach article, he explains some of the advances he has made in more recent years.

EGU2020 presentation by Arne on modelling of active magmatic systems

As EGU2020 will be an online-only event, Arne prepared a video that summarizes his findings on 3D modelling of the Puna-Altiplano magmatic system. Have a look at it here.

Related Publications

A Multiphysics Approach to Constrain the Dynamics of the Altiplano-Puna Magmatic System.

We present combined geodynamic/gravity modelling to interpret the Puna magmatic system, which also showcases the application of sensitivity testing tools to determine the key model parameters.

Inferring rheology and geometry of subsurface structures byadjoint-based inversion of principal stress directions

We present a way to use principal stress directions and strain rate components for geodynamic inversions and have implemented this, as well as scaling laws, in our open-source 3D geodynamic code LaMEM. This might sound technical but has significant implications for modelling actual case studies.

An autonomous phase diagram database for geodynamic simulations of magmatic systems.

A method is proposed with which we can pre-compute phase diagrams on parallel computers and construct a database that allows to compute the chemical evolution of magmatic system.

Plume-lithosphere interactions during the Archean and implications for the generation of early continental terranes

We perform 3D geodynamic models that include a simplified melt migration and differentiation algorithm to study how plume lithosphere interaction results on the formation of continental terranes.

Insights into the compositional evolution of crustal magmatic systems from coupled petrological-geodynamical models.

The interaction between deformation of the lithosphere and the chemical evolution of migrating melts is incompletely understood. Here, we describe an approach in which we couple a large amount of precomputed phase diagrams with a lithospheric deformation code that includes a simple diking parameterization. The simulations are shown to give detailed, yet realistic, predictions that compare reasonably well with observations.

Project Goals

  1. Create new computer models to simulate two-phase flow through a deforming thermo-mechanical visco-elasto-plastic lithosphere. Both mode-1 and mode-2 plasticity models will be taken into account. We plan to utilize PETSc for this purpose together with the DMSTAG framework.

  2. Develop new approaches to couple state-of-the-art thermodynamic melting models with geodynamic models to simulate the evolving chemistry of crystallizing melts on its way from the mantle or lower crust to shallower levels in the lithosphere. Recently, hydrous mafic melting models have been developed that can, in principle, simulate the chemical differentiation of a hydrous mafic melt to more felsic compostions such as granite or granodiorite. Yet, coupling such thermodynamic models with geodynamic models remains a challenging task as the thermodynamic computations are too time-consuming to be performed on the fly. We will work on different approaches to overcome this.

  3. Perform geodynamic inverse models of active magmatic systems, for which geophysics data is available. This allows interpreting the data in a mechanically consistent manner.

  4. Perform systematic numerical simulations to better understand the physics of these models and compare it with observations.

As of now, work concentrates on the first three tasks. The software that will be developed in MAGMA will be released as open-source software, after we publish the methodology.


Project leader


Boris Kaus

Professor of Geodynamics and Geophysics, JGU Mainz

Computational Geodynamics, Tectonics, Magmatic Processes



Daniel Kiss

Postdoc, JGU

Couple geodynamic and thermodynamic models


Nicolas Riel

Postdoc, JGU

New thermodynamic solvers, Modelling magmatic systems

PhD students


Arne Spang

PhD student, JGU

Geodynamic inverse modelling, Mechanical interpretation of active magmatic systems


Mara Arts

PhD student, JGU

Developing software for modelling two-phase flow


Nicolas Berlie

PhD student, JGU

Developing software for modelling two-phase flow



Tanja Eich

Administrative wizzard, JGU

JGU collaborators


Andrea Picollo

Postdoc, JGU

Dynamics of the Early Earth, magmatic processes


Anton Popov

Lecturer, JGU

Software development for computational geodynamics, Geomechanics


Evangelos Moulas

Junior Professor Metamorphic Processes, JGU

Quantitative thermodynamics, Geomechanics


Georg Reuber

PhD student, JGU

Geodynamic inverse modelling approaches, Adjoint based inversion


Tobias Baumann

Lecturer, JGU

Geodynamic inverse modelling

External collaborators


Eleanor Green

Senior Lecturer, University of Melbourne

thermodynamic melting models,


Patrick Sanan

Postdoc, ETH Zurich

PETSc software development, high-performance computing


Richard White

Professor of Earth Science, University of St. Andrews

thermodynamics, melt migration



Lisa Rummel


Combining thermodynamic and geodynamic models to simulate magmatic systems

Contact us

  • +49 6131 392 4527
  • Johann-Joachim-Becher-Weg 21, Mainz, 55128
  • We are at the ground floor of the building. Enter through the main entrance and take a left at the elevators (which have a useful map to show our offices).
  • Skype us