Reservoir Computing as a Tool for Climate Predictability Studies
Abstract Reduced‐order dynamical models play a central role in developing our understanding of predictability of climate irrespective of whether we are dealing with the actual climate system or surrogate climate models. In this context, the linear inverse modeling (LIM) approach, by capturing a few...
Published in: | Journal of Advances in Modeling Earth Systems |
---|---|
Main Author: | |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
American Geophysical Union (AGU)
2021
|
Subjects: | |
Online Access: | https://doi.org/10.1029/2020MS002290 https://doaj.org/article/72edb6b614be428390a8b62a2cdbc1ab |
id |
ftdoajarticles:oai:doaj.org/article:72edb6b614be428390a8b62a2cdbc1ab |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:72edb6b614be428390a8b62a2cdbc1ab 2023-05-15T17:35:39+02:00 Reservoir Computing as a Tool for Climate Predictability Studies Balasubramanya T. Nadiga 2021-04-01T00:00:00Z https://doi.org/10.1029/2020MS002290 https://doaj.org/article/72edb6b614be428390a8b62a2cdbc1ab EN eng American Geophysical Union (AGU) https://doi.org/10.1029/2020MS002290 https://doaj.org/toc/1942-2466 1942-2466 doi:10.1029/2020MS002290 https://doaj.org/article/72edb6b614be428390a8b62a2cdbc1ab Journal of Advances in Modeling Earth Systems, Vol 13, Iss 4, Pp n/a-n/a (2021) climate echo state networks linear inverse modeling machine learning predictability reservoir computing Physical geography GB3-5030 Oceanography GC1-1581 article 2021 ftdoajarticles https://doi.org/10.1029/2020MS002290 2022-12-31T06:33:26Z Abstract Reduced‐order dynamical models play a central role in developing our understanding of predictability of climate irrespective of whether we are dealing with the actual climate system or surrogate climate models. In this context, the linear inverse modeling (LIM) approach, by capturing a few essential interactions between dynamical components of the full system, has proven valuable in providing insights into predictability of the full system. We demonstrate that reservoir computing (RC), a form of learning suitable for systems with chaotic dynamics, provides an alternative nonlinear approach that improves on the predictive skill of the LIM approach. We do this in the example setting of predicting sea surface temperature in the North Atlantic in the preindustrial control simulation of a popular earth system model, the Community Earth System Model so that we can compare the performance of the new RC‐based approach with the traditional LIM approach both when learning data are plentiful and when such data are more limited. The improved predictive skill of the RC approach over a wide range of conditions—larger number of retained EOF coefficients, extending well into the limited data regime, etc.—suggests that this machine‐learning technique may have a use in climate predictability studies. While the possibility of developing a climate emulator—the ability to continue the evolution of the system on the attractor long after failing to be able to track the reference trajectory—is demonstrated in the Lorenz‐63 system, it is suggested that further development of the RC approach may permit such uses of the new approach in more realistic predictability studies. Article in Journal/Newspaper North Atlantic Directory of Open Access Journals: DOAJ Articles Journal of Advances in Modeling Earth Systems 13 4 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
climate echo state networks linear inverse modeling machine learning predictability reservoir computing Physical geography GB3-5030 Oceanography GC1-1581 |
spellingShingle |
climate echo state networks linear inverse modeling machine learning predictability reservoir computing Physical geography GB3-5030 Oceanography GC1-1581 Balasubramanya T. Nadiga Reservoir Computing as a Tool for Climate Predictability Studies |
topic_facet |
climate echo state networks linear inverse modeling machine learning predictability reservoir computing Physical geography GB3-5030 Oceanography GC1-1581 |
description |
Abstract Reduced‐order dynamical models play a central role in developing our understanding of predictability of climate irrespective of whether we are dealing with the actual climate system or surrogate climate models. In this context, the linear inverse modeling (LIM) approach, by capturing a few essential interactions between dynamical components of the full system, has proven valuable in providing insights into predictability of the full system. We demonstrate that reservoir computing (RC), a form of learning suitable for systems with chaotic dynamics, provides an alternative nonlinear approach that improves on the predictive skill of the LIM approach. We do this in the example setting of predicting sea surface temperature in the North Atlantic in the preindustrial control simulation of a popular earth system model, the Community Earth System Model so that we can compare the performance of the new RC‐based approach with the traditional LIM approach both when learning data are plentiful and when such data are more limited. The improved predictive skill of the RC approach over a wide range of conditions—larger number of retained EOF coefficients, extending well into the limited data regime, etc.—suggests that this machine‐learning technique may have a use in climate predictability studies. While the possibility of developing a climate emulator—the ability to continue the evolution of the system on the attractor long after failing to be able to track the reference trajectory—is demonstrated in the Lorenz‐63 system, it is suggested that further development of the RC approach may permit such uses of the new approach in more realistic predictability studies. |
format |
Article in Journal/Newspaper |
author |
Balasubramanya T. Nadiga |
author_facet |
Balasubramanya T. Nadiga |
author_sort |
Balasubramanya T. Nadiga |
title |
Reservoir Computing as a Tool for Climate Predictability Studies |
title_short |
Reservoir Computing as a Tool for Climate Predictability Studies |
title_full |
Reservoir Computing as a Tool for Climate Predictability Studies |
title_fullStr |
Reservoir Computing as a Tool for Climate Predictability Studies |
title_full_unstemmed |
Reservoir Computing as a Tool for Climate Predictability Studies |
title_sort |
reservoir computing as a tool for climate predictability studies |
publisher |
American Geophysical Union (AGU) |
publishDate |
2021 |
url |
https://doi.org/10.1029/2020MS002290 https://doaj.org/article/72edb6b614be428390a8b62a2cdbc1ab |
genre |
North Atlantic |
genre_facet |
North Atlantic |
op_source |
Journal of Advances in Modeling Earth Systems, Vol 13, Iss 4, Pp n/a-n/a (2021) |
op_relation |
https://doi.org/10.1029/2020MS002290 https://doaj.org/toc/1942-2466 1942-2466 doi:10.1029/2020MS002290 https://doaj.org/article/72edb6b614be428390a8b62a2cdbc1ab |
op_doi |
https://doi.org/10.1029/2020MS002290 |
container_title |
Journal of Advances in Modeling Earth Systems |
container_volume |
13 |
container_issue |
4 |
_version_ |
1766134885205934080 |