Digital Twin Earth - Coasts: Developing a fast and physics-informed surrogate model for coastal floods via neural operators.
View/ Open
Average rating
votes
Date
2021Author
Jiang, Peishi
Meinert, Nis
Jordão, Helga
Weisser, Constantin
Holgate, Simon
Lavin, Alexander
Lütjens, Björn
Newman, Dava
Wainwright, Haruko
Walker, Catherine
Barnard, Patrick
Status
PublishedMetadata
Show full item recordAbstract
Developing fast and accurate surrogates for physics-based coastal and ocean models
is an urgent need due to the coastal flood risk under accelerating sea level rise,
and the computational expense of deterministic numerical models. For this purpose,
we develop the first digital twin of Earth coastlines with new physics-informed
machine learning techniques extending the state-of-art Neural Operator. As a
proof-of-concept study, we built Fourier Neural Operator (FNO) surrogates on the
simulations of an industry-standard coastal and ocean model – Nucleus for European
Modelling of the Ocean (NEMO). The resulting FNO surrogate accurately
predicts the sea surface height in most regions while achieving upwards of 45x
acceleration of NEMO. We delivered an open-source CoastalTwin platform in an
end-to-end and modular way, to enable easy extensions to other simulations and
ML-based surrogate methods. Our results and deliverable provide a promising
approach to massively accelerate coas.....
Resource URL
https://neurips.cc/Conferences/2021Title of Book
Machine Learning and the Physical Sciences Workshop at the 35th Conference on Neural Information Processing Systems (NeurIPS) December 13, 2021. [Online]Page Range
7pp.Publisher
Neural Information Processing Systems FoundationDocument Language
enSustainable Development Goals (SDG)
14.aMaturity Level
MatureDOI Original
https://doi.org/10.48550/arXiv.2110.07100Citation
Jiang, P., et al (2021) Digital Twin Earth - Coasts: Developing a fast and physics-informed surrogate model for coastal floods via neural operators. In: Machine Learning and the Physical Sciences Workshop at the 35th Conference on Neural Information Processing Systems (NeurIPS) December 13, 2021, 7pp.(arXiv:2110.07100v1 [physics.ao-ph] for this version). DOI: https://doi.org/10.48550/arXiv.2110.07100Collections