Perspectives of physics-based machine learning strategies for geoscientific applications governed by partial  differential equations

Degen, Denise; Caviedes Voullième, Daniel; Buiter, Susanne; Hendricks Franssen, Harrie-Jan; Vereecken, Harry; González-Nicolás, Ana; Wellmann, Florian

doi:https://doi.org/10.5194/gmd-16-7375-2023

Articles | Volume 16, issue 24

https://doi.org/10.5194/gmd-16-7375-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-16-7375-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 16, issue 24

Review and perspective paper

|

19 Dec 2023

Review and perspective paper |

| 19 Dec 2023

Perspectives of physics-based machine learning strategies for geoscientific applications governed by partial differential equations

Denise Degen, Daniel Caviedes Voullième, Susanne Buiter, Harrie-Jan Hendricks Franssen, Harry Vereecken, Ana González-Nicolás, and Florian Wellmann

Data sets

Non-Intrusive Reduced Basis Code for Geothermal, Geodynamic, and Hydrology Benchmarks Denise Degen, Daniel Caviedes Voullième, Susanne Buiter, Harrie-Jan Hendriks Franssen, Harry Vereecken, Ana González-Nicolás, and Florian Wellmann https://doi.org/10.5281/zenodo.8369108

Model code and software

Non-Intrusive Reduced Basis Code for Geothermal, Geodynamic, and Hydrology Benchmarks Denise Degen, Daniel Caviedes Voullième, Susanne Buiter, Harrie-Jan Hendriks Franssen, Harry Vereecken, Ana González-Nicolás, and Florian Wellmann https://doi.org/10.5281/zenodo.8369108

ASPECT v2.4.0 (v2.4.0) Wolfgang Bangerth, Juliane Dannberg, Menno Fraters, Rene Gassmoeller, Anne Glerum, Timo Heister, Robert Myhill, and John Naliboff https://doi.org/10.5281/zenodo.6903424

cgre-aachen/DwarfElephant: DwarfElephant 1.0 D. Degen, K. Veroy, and F. Wellmann https://doi.org/10.5281/zenodo.4074777

parflow/parflow: ParFlow Version 3.10.0 (v3.10.0) Steven Smith, reedmaxwell, i-ferguson, et al. https://doi.org/10.5281/zenodo.6413322

Short summary

In geosciences, we often use simulations based on physical laws. These simulations can be computationally expensive, which is a problem if simulations must be performed many times (e.g., to add error bounds). We show how a novel machine learning method helps to reduce simulation time. In comparison to other approaches, which typically only look at the output of a simulation, the method considers physical laws in the simulation itself. The method provides reliable results faster than standard.