Data-driven predictions of a multiscale Lorenz 96 chaotic system using machine-learning methods: reservoir computing, artificial neural network, and long short-term memory network

Chattopadhyay, Ashesh; Hassanzadeh, Pedram; Subramanian, Devika

doi:https://doi.org/10.5194/npg-27-373-2020

Articles | Volume 27, issue 3

https://doi.org/10.5194/npg-27-373-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/npg-27-373-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 27, issue 3

Research article

|

02 Jul 2020

Research article |

| 02 Jul 2020

Data-driven predictions of a multiscale Lorenz 96 chaotic system using machine-learning methods: reservoir computing, artificial neural network, and long short-term memory network

Ashesh Chattopadhyay, Pedram Hassanzadeh, and Devika Subramanian

Download

Final revised paper (published on 02 Jul 2020)
Preprint (discussion started on 02 Jan 2020)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review Summary', Anonymous Referee #1, 05 Mar 2020
RC2: 'Review of the manuscript "Data-driven prediction of a multi-scale Lorenz 96 chaotic system..."', Anonymous Referee #2, 20 Mar 2020
AC1: 'Response to reviewers' comments and manuscript with tracked changes', Pedram Hassanzadeh, 01 May 2020

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Pedram Hassanzadeh on behalf of the Authors (01 May 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (22 May 2020) by Zoltan Toth

RR by Istvan Szunyogh (24 May 2020)

RR by Anonymous Referee #2 (26 May 2020)

ED: Publish subject to technical corrections (27 May 2020) by Zoltan Toth

Dear Authors:
Thank you for your careful response to the two expert Reviewers and the additional experiments you performed. The revised manuscript was sent back for a second evaluation to the original reviewers, who were satisfied with your revisions. So I am glad to accept your manuscript for publication in NPG, subject to minor technical corrections, see editorial comments below. Congratulations and I hope you will consider NPG for publishing your future research results
Zoltan Toth - Editor

A few editorial comments:
The latter part of this sentence from the Abstract is awkward, can you please rephrase? "The RNN-LSTM and ANN show some prediction skills as well; with RNN-LSTM outperform- ing ANN.”

Wonder if you would consider including Table 2 from your response to Rev. 1’s comments with a short associated discussion into an Appendix?

Grammar needs adjustment in this sentence from Discussion: "We have also found that the prediction horizon, and in particular, the prediction accuracy, of RC-ESN to have a weak dependence on the size of the training set (Figure 7).”

l.324, change period to colon? ANN<… sentence is incomplete now.

l.326, “them” - grammatically not clear what this refers to, pls clarify

l.334-336 - lots of references but no indication of what possibilities may exist there or are being explored?

l.341 - parentheses not used properly

l.349 - “…coupling, see their Fig. 5"

l.351-352, sentence straddling 2 lines is grammatically incorrect

l.353 - replace variant w various?

l.355-356, avoid using 2nd level paranthesis, separate reference with comma instead

l.358-359 - Elevate text in parenthesis, make it a separate sentence; replace semicolon with comma

l.361, place only year of publication in parenthesis when citation is integral part of text, I believe that's the convention

l.374 - break up way too long sentence at colon

l.373-378 - nice closing thoughts

Can you please carefully read the rest of your manuscript before you submit the final version to check if there are any grammatical / stylistic corrections like those listed above needed?

Hide

AR by Pedram Hassanzadeh on behalf of the Authors (28 May 2020) Author's response Manuscript

Short summary

The performance of three machine-learning methods for data-driven modeling of a multiscale chaotic Lorenz 96 system is examined. One of the methods is found to be able to predict the future evolution of the chaotic system well from just knowing the past observations of the large-scale component of the multiscale state vector. Potential applications to data-driven and data-assisted surrogate modeling of complex dynamical systems such as weather and climate are discussed.