Ice-flow model emulator based on physics-informed deep learning
International audience Convolutional neural networks (CNN) trained from high-order ice-flow model realisations have proven to be outstanding emulators in terms of fidelity and computational performance. However, the dependence on an ensemble of realisations of an instructor model renders this strate...
Published in: | Journal of Glaciology |
---|---|
Main Authors: | , |
Other Authors: | , , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
HAL CCSD
2023
|
Subjects: | |
Online Access: | https://hal.science/hal-04232949 https://hal.science/hal-04232949v2/document https://hal.science/hal-04232949v2/file/PINN_ice_flow_emulator.pdf https://doi.org/10.1017/jog.2023.73 |
id |
ftunivcotedazur:oai:HAL:hal-04232949v2 |
---|---|
record_format |
openpolar |
spelling |
ftunivcotedazur:oai:HAL:hal-04232949v2 2024-02-04T10:01:20+01:00 Ice-flow model emulator based on physics-informed deep learning Jouvet, Guillaume Cordonnier, Guillaume Université de Lausanne = University of Lausanne (UNIL) GRAPHics and DEsign with hEterogeneous COntent (GRAPHDECO) Inria Sophia Antipolis - Méditerranée (CRISAM) Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria) ANR-22-CE33-0012,INVTERRA,Contrôle inverse de terrains cohérents physiquement(2022) 2023 https://hal.science/hal-04232949 https://hal.science/hal-04232949v2/document https://hal.science/hal-04232949v2/file/PINN_ice_flow_emulator.pdf https://doi.org/10.1017/jog.2023.73 en eng HAL CCSD International Glaciological Society info:eu-repo/semantics/altIdentifier/doi/10.1017/jog.2023.73 hal-04232949 https://hal.science/hal-04232949 https://hal.science/hal-04232949v2/document https://hal.science/hal-04232949v2/file/PINN_ice_flow_emulator.pdf doi:10.1017/jog.2023.73 http://creativecommons.org/licenses/by/ info:eu-repo/semantics/OpenAccess ISSN: 0022-1430 EISSN: 1727-5652 Journal of Glaciology https://hal.science/hal-04232949 Journal of Glaciology, 2023, pp.1-15. ⟨10.1017/jog.2023.73⟩ [SDU.STU.GL]Sciences of the Universe [physics]/Earth Sciences/Glaciology [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI] info:eu-repo/semantics/article Journal articles 2023 ftunivcotedazur https://doi.org/10.1017/jog.2023.73 2024-01-09T23:43:58Z International audience Convolutional neural networks (CNN) trained from high-order ice-flow model realisations have proven to be outstanding emulators in terms of fidelity and computational performance. However, the dependence on an ensemble of realisations of an instructor model renders this strategy difficult to generalise to a variety of ice-flow regimes found in the nature. To overcome this issue, we adopt the approach of physics-informed deep learning, which fuses traditional numerical solutions by finite differences/elements and deep-learning approaches. Here, we train a CNN to minimise the energy associated with high-order ice-flow equations within the time iterations of a glacier evolution model. As a result, our emulator is a promising alternative to traditional solvers thanks to its high computational efficiency (especially on GPU), its high fidelity to the original model, its simplified training (without requiring any data), its capability to handle a variety of ice-flow regimes and memorise previous solutions, and its relatively simple implementation. Embedded into the ‘Instructed Glacier Model’ (IGM) framework, the potential of the emulator is illustrated with three applications including a large-scale high-resolution (2400x4000) forward glacier evolution model, an inverse modelling case for data assimilation, and an ice shelf. Article in Journal/Newspaper Ice Shelf Journal of Glaciology HAL Université Côte d'Azur Journal of Glaciology 1 15 |
institution |
Open Polar |
collection |
HAL Université Côte d'Azur |
op_collection_id |
ftunivcotedazur |
language |
English |
topic |
[SDU.STU.GL]Sciences of the Universe [physics]/Earth Sciences/Glaciology [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI] |
spellingShingle |
[SDU.STU.GL]Sciences of the Universe [physics]/Earth Sciences/Glaciology [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI] Jouvet, Guillaume Cordonnier, Guillaume Ice-flow model emulator based on physics-informed deep learning |
topic_facet |
[SDU.STU.GL]Sciences of the Universe [physics]/Earth Sciences/Glaciology [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI] |
description |
International audience Convolutional neural networks (CNN) trained from high-order ice-flow model realisations have proven to be outstanding emulators in terms of fidelity and computational performance. However, the dependence on an ensemble of realisations of an instructor model renders this strategy difficult to generalise to a variety of ice-flow regimes found in the nature. To overcome this issue, we adopt the approach of physics-informed deep learning, which fuses traditional numerical solutions by finite differences/elements and deep-learning approaches. Here, we train a CNN to minimise the energy associated with high-order ice-flow equations within the time iterations of a glacier evolution model. As a result, our emulator is a promising alternative to traditional solvers thanks to its high computational efficiency (especially on GPU), its high fidelity to the original model, its simplified training (without requiring any data), its capability to handle a variety of ice-flow regimes and memorise previous solutions, and its relatively simple implementation. Embedded into the ‘Instructed Glacier Model’ (IGM) framework, the potential of the emulator is illustrated with three applications including a large-scale high-resolution (2400x4000) forward glacier evolution model, an inverse modelling case for data assimilation, and an ice shelf. |
author2 |
Université de Lausanne = University of Lausanne (UNIL) GRAPHics and DEsign with hEterogeneous COntent (GRAPHDECO) Inria Sophia Antipolis - Méditerranée (CRISAM) Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria) ANR-22-CE33-0012,INVTERRA,Contrôle inverse de terrains cohérents physiquement(2022) |
format |
Article in Journal/Newspaper |
author |
Jouvet, Guillaume Cordonnier, Guillaume |
author_facet |
Jouvet, Guillaume Cordonnier, Guillaume |
author_sort |
Jouvet, Guillaume |
title |
Ice-flow model emulator based on physics-informed deep learning |
title_short |
Ice-flow model emulator based on physics-informed deep learning |
title_full |
Ice-flow model emulator based on physics-informed deep learning |
title_fullStr |
Ice-flow model emulator based on physics-informed deep learning |
title_full_unstemmed |
Ice-flow model emulator based on physics-informed deep learning |
title_sort |
ice-flow model emulator based on physics-informed deep learning |
publisher |
HAL CCSD |
publishDate |
2023 |
url |
https://hal.science/hal-04232949 https://hal.science/hal-04232949v2/document https://hal.science/hal-04232949v2/file/PINN_ice_flow_emulator.pdf https://doi.org/10.1017/jog.2023.73 |
genre |
Ice Shelf Journal of Glaciology |
genre_facet |
Ice Shelf Journal of Glaciology |
op_source |
ISSN: 0022-1430 EISSN: 1727-5652 Journal of Glaciology https://hal.science/hal-04232949 Journal of Glaciology, 2023, pp.1-15. ⟨10.1017/jog.2023.73⟩ |
op_relation |
info:eu-repo/semantics/altIdentifier/doi/10.1017/jog.2023.73 hal-04232949 https://hal.science/hal-04232949 https://hal.science/hal-04232949v2/document https://hal.science/hal-04232949v2/file/PINN_ice_flow_emulator.pdf doi:10.1017/jog.2023.73 |
op_rights |
http://creativecommons.org/licenses/by/ info:eu-repo/semantics/OpenAccess |
op_doi |
https://doi.org/10.1017/jog.2023.73 |
container_title |
Journal of Glaciology |
container_start_page |
1 |
op_container_end_page |
15 |
_version_ |
1789967109539758080 |