GEORGIA: A Graph Neural Network Based EmulatOR for Glacial Isostatic Adjustment. Lin, Y., Whitehouse, P. L., Valentine, A. P., & Woodroffe, S. A. Geophysical Research Letters, 50(18):e2023GL103672, 2023. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1029/2023GL103672![GEORGIA: A Graph Neural Network Based EmulatOR for Glacial Isostatic Adjustment [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex Glacial isostatic adjustment (GIA) modeling is not only useful for understanding past relative sea-level change but also for projecting future sea-level change due to ongoing land deformation. However, GIA model predictions are subject to a range of uncertainties, most notably due to uncertainty in the input ice history. An effective way to reduce this uncertainty is to perform data-model comparisons over a large ensemble of possible ice histories, but this is often impossible due to computational limitations. Here we address this problem by building a deep-learning-based GIA emulator that can mimic the behavior of a physics-based GIA model while being computationally cheap to evaluate. Assuming a single 1-D Earth rheology, our emulator shows 0.54 m mean absolute error on 150 out-of-sample testing data with \textless0.5 s emulation time. Using this emulator, two illustrative applications related to the calculation of barystatic sea level are provided for use by the sea-level community.
@article{lin_georgia_2023,
title = {{GEORGIA}: {A} {Graph} {Neural} {Network} {Based} {EmulatOR} for {Glacial} {Isostatic} {Adjustment}},
volume = {50},
copyright = {© 2023. The Authors. Geophysical Research Letters published by Wiley Periodicals LLC on behalf of American Geophysical Union.},
issn = {1944-8007},
shorttitle = {{GEORGIA}},
url = {https://onlinelibrary.wiley.com/doi/abs/10.1029/2023GL103672},
doi = {10.1029/2023GL103672},
abstract = {Glacial isostatic adjustment (GIA) modeling is not only useful for understanding past relative sea-level change but also for projecting future sea-level change due to ongoing land deformation. However, GIA model predictions are subject to a range of uncertainties, most notably due to uncertainty in the input ice history. An effective way to reduce this uncertainty is to perform data-model comparisons over a large ensemble of possible ice histories, but this is often impossible due to computational limitations. Here we address this problem by building a deep-learning-based GIA emulator that can mimic the behavior of a physics-based GIA model while being computationally cheap to evaluate. Assuming a single 1-D Earth rheology, our emulator shows 0.54 m mean absolute error on 150 out-of-sample testing data with {\textless}0.5 s emulation time. Using this emulator, two illustrative applications related to the calculation of barystatic sea level are provided for use by the sea-level community.},
language = {en},
number = {18},
urldate = {2024-01-29},
journal = {Geophysical Research Letters},
author = {Lin, Yucheng and Whitehouse, Pippa L. and Valentine, Andrew P. and Woodroffe, Sarah A.},
year = {2023},
note = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1029/2023GL103672},
keywords = {glacial isostatic adjustment, machine learning, sea-level change, statistical emulator},
pages = {e2023GL103672},
}
Downloads: 0
{"_id":"5JbTQB39vSSTYygCq","bibbaseid":"lin-whitehouse-valentine-woodroffe-georgiaagraphneuralnetworkbasedemulatorforglacialisostaticadjustment-2023","author_short":["Lin, Y.","Whitehouse, P. L.","Valentine, A. P.","Woodroffe, S. A."],"bibdata":{"bibtype":"article","type":"article","title":"GEORGIA: A Graph Neural Network Based EmulatOR for Glacial Isostatic Adjustment","volume":"50","copyright":"© 2023. The Authors. Geophysical Research Letters published by Wiley Periodicals LLC on behalf of American Geophysical Union.","issn":"1944-8007","shorttitle":"GEORGIA","url":"https://onlinelibrary.wiley.com/doi/abs/10.1029/2023GL103672","doi":"10.1029/2023GL103672","abstract":"Glacial isostatic adjustment (GIA) modeling is not only useful for understanding past relative sea-level change but also for projecting future sea-level change due to ongoing land deformation. However, GIA model predictions are subject to a range of uncertainties, most notably due to uncertainty in the input ice history. An effective way to reduce this uncertainty is to perform data-model comparisons over a large ensemble of possible ice histories, but this is often impossible due to computational limitations. Here we address this problem by building a deep-learning-based GIA emulator that can mimic the behavior of a physics-based GIA model while being computationally cheap to evaluate. Assuming a single 1-D Earth rheology, our emulator shows 0.54 m mean absolute error on 150 out-of-sample testing data with \\textless0.5 s emulation time. Using this emulator, two illustrative applications related to the calculation of barystatic sea level are provided for use by the sea-level community.","language":"en","number":"18","urldate":"2024-01-29","journal":"Geophysical Research Letters","author":[{"propositions":[],"lastnames":["Lin"],"firstnames":["Yucheng"],"suffixes":[]},{"propositions":[],"lastnames":["Whitehouse"],"firstnames":["Pippa","L."],"suffixes":[]},{"propositions":[],"lastnames":["Valentine"],"firstnames":["Andrew","P."],"suffixes":[]},{"propositions":[],"lastnames":["Woodroffe"],"firstnames":["Sarah","A."],"suffixes":[]}],"year":"2023","note":"_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1029/2023GL103672","keywords":"glacial isostatic adjustment, machine learning, sea-level change, statistical emulator","pages":"e2023GL103672","bibtex":"@article{lin_georgia_2023,\n\ttitle = {{GEORGIA}: {A} {Graph} {Neural} {Network} {Based} {EmulatOR} for {Glacial} {Isostatic} {Adjustment}},\n\tvolume = {50},\n\tcopyright = {© 2023. The Authors. Geophysical Research Letters published by Wiley Periodicals LLC on behalf of American Geophysical Union.},\n\tissn = {1944-8007},\n\tshorttitle = {{GEORGIA}},\n\turl = {https://onlinelibrary.wiley.com/doi/abs/10.1029/2023GL103672},\n\tdoi = {10.1029/2023GL103672},\n\tabstract = {Glacial isostatic adjustment (GIA) modeling is not only useful for understanding past relative sea-level change but also for projecting future sea-level change due to ongoing land deformation. However, GIA model predictions are subject to a range of uncertainties, most notably due to uncertainty in the input ice history. An effective way to reduce this uncertainty is to perform data-model comparisons over a large ensemble of possible ice histories, but this is often impossible due to computational limitations. Here we address this problem by building a deep-learning-based GIA emulator that can mimic the behavior of a physics-based GIA model while being computationally cheap to evaluate. Assuming a single 1-D Earth rheology, our emulator shows 0.54 m mean absolute error on 150 out-of-sample testing data with {\\textless}0.5 s emulation time. Using this emulator, two illustrative applications related to the calculation of barystatic sea level are provided for use by the sea-level community.},\n\tlanguage = {en},\n\tnumber = {18},\n\turldate = {2024-01-29},\n\tjournal = {Geophysical Research Letters},\n\tauthor = {Lin, Yucheng and Whitehouse, Pippa L. and Valentine, Andrew P. and Woodroffe, Sarah A.},\n\tyear = {2023},\n\tnote = {\\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1029/2023GL103672},\n\tkeywords = {glacial isostatic adjustment, machine learning, sea-level change, statistical emulator},\n\tpages = {e2023GL103672},\n}\n\n","author_short":["Lin, Y.","Whitehouse, P. L.","Valentine, A. P.","Woodroffe, S. A."],"key":"lin_georgia_2023","id":"lin_georgia_2023","bibbaseid":"lin-whitehouse-valentine-woodroffe-georgiaagraphneuralnetworkbasedemulatorforglacialisostaticadjustment-2023","role":"author","urls":{"Paper":"https://onlinelibrary.wiley.com/doi/abs/10.1029/2023GL103672"},"keyword":["glacial isostatic adjustment","machine learning","sea-level change","statistical emulator"],"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://api.zotero.org/users/3220083/collections/G3U77W5V/items?key=xBJ1ETD6lXd0fvRF3JiRItrt&format=bibtex&limit=100","dataSources":["faCBEcxzpEm3MqKhw","xxp2MFsBLWJmqgZcw"],"keywords":["glacial isostatic adjustment","machine learning","sea-level change","statistical emulator"],"search_terms":["georgia","graph","neural","network","based","emulator","glacial","isostatic","adjustment","lin","whitehouse","valentine","woodroffe"],"title":"GEORGIA: A Graph Neural Network Based EmulatOR for Glacial Isostatic Adjustment","year":2023}