Using neural network ensembles to separate ocean biogeochemical and physical drivers of phytoplankton biogeography in Earth system models

Holder, Christopher; Gnanadesikan, Anand; Aude-Pradal, Marie

doi:https://doi.org/10.5194/gmd-15-1595-2022

Articles | Volume 15, issue 4

https://doi.org/10.5194/gmd-15-1595-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

Benchmark datasets and machine learning algorithms for Earth...

https://doi.org/10.5194/gmd-15-1595-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 4

Methods for assessment of models

| Highlight paper

|

23 Feb 2022

Methods for assessment of models | Highlight paper |

| 23 Feb 2022

Using neural network ensembles to separate ocean biogeochemical and physical drivers of phytoplankton biogeography in Earth system models

Christopher Holder, Anand Gnanadesikan, and Marie Aude-Pradal

Download

Final revised paper (published on 23 Feb 2022)
Preprint (discussion started on 19 Aug 2021)

Interactive discussion

Status: closed

RC1:
'Comment on gmd-2021-167', Anonymous Referee #1, 08 Oct 2021

The comment was uploaded in the form of a supplement: https://gmd.copernicus.org/preprints/gmd-2021-167/gmd-2021-167-RC1-supplement.pdf

Citation: https://doi.org/10.5194/gmd-2021-167-RC1
- AC1: 'Reply on RC1', Christopher Holder, 21 Dec 2021
  
  The comment was uploaded in the form of a supplement: https://gmd.copernicus.org/preprints/gmd-2021-167/gmd-2021-167-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/gmd-2021-167-AC1
CEC1:
'Comment on gmd-2021-167', Juan Antonio Añel, 12 Oct 2021

Dear authors,
I have checked the 'Code and Data Availability' section of your manuscript, and I would like to request you three minor fixes:
- First, please, add in the mentioned section a statement specific about the availability of data;
- Second, in the Zenodo repository, add a Readme file with an explanation about what is each file;
- Make a clear distinction between input files and output files.
Juan A. Añel

Geosc. Mod. Dev. Exec. Editor

Citation: https://doi.org/10.5194/gmd-2021-167-CEC1
- AC2: 'Reply on CEC1', Christopher Holder, 21 Dec 2021
  
  Author Responses Addressing Review from Juan Antonio Añel for “Using Neural Network Ensembles to Separate Biogeochemical and Physical Components in Earth System Models” by Holder et al.
  For these responses, we address each Referee comment individually and include our response below it. The Referee Comments are abbreviated as RC, while the Author Responses are abbreviated as AR.
  
  Dear authors,
  I have checked the 'Code and Data Availability' section of your manuscript, and I would like to request you three minor fixes:
  
  RC1: First, please, add in the mentioned section a statement specific about the availability of data.
  AC1: Added a statement in the updated manuscript specific to the availability of the data.
  
  RC2: Second, in the Zenodo repository, add a Readme file with an explanation about what is each file.
  AC2: Added a Readme file in to the Zenodo repository with file explanations.
  
  RC3: Make a clear distinction between input files and output files.
  AC3: As part of the Readme file, we included information about which files were input and output files.
  
  Juan A. Añel
  
  Geosc. Mod. Dev. Exec. Editor
  
  Citation: https://doi.org/10.5194/gmd-2021-167-AC2
RC2:
'Comment on gmd-2021-167', Anonymous Referee #2, 15 Nov 2021

See attached.

Citation: https://doi.org/10.5194/gmd-2021-167-RC2
- AC3: 'Reply on RC2', Christopher Holder, 21 Dec 2021
  
  The comment was uploaded in the form of a supplement: https://gmd.copernicus.org/preprints/gmd-2021-167/gmd-2021-167-AC3-supplement.pdf
  
  Citation: https://doi.org/10.5194/gmd-2021-167-AC3

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Christopher Holder on behalf of the Authors (21 Dec 2021) Author's response Author's tracked changes Manuscript

ED: Publish as is (07 Jan 2022) by Richard Mills

AR by Christopher Holder on behalf of the Authors (14 Jan 2022) Author's response Manuscript

Short summary

It can be challenging to understand why Earth system models (ESMs) produce specific results because one can arrive at the same result simply by changing the values of the parameters. In our paper, we demonstrate that it is possible to use machine learning to figure out how and why particular components of an ESM (such as biology or ocean circulations) affect the output. This work could be applied to observations to improve the accuracy of the formulations used in ESMs.