LSCE-FFNN-v1: a two-step neural network model for the reconstruction of surface ocean p CO 2 over the global ocean
A new feed-forward neural network (FFNN) model is presented to reconstruct surface ocean partial pressure of carbon dioxide ( p CO 2 ) over the global ocean. The model consists of two steps: (1) the reconstruction of p CO 2 climatology, and (2) the reconstruction of p CO 2 anomalies with respect to...
Published in: | Geoscientific Model Development |
---|---|
Main Authors: | , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Copernicus Publications
2019
|
Subjects: | |
Online Access: | https://doi.org/10.5194/gmd-12-2091-2019 https://doaj.org/article/5ad5231d8e26418dad07cfcadc405c62 |
id |
ftdoajarticles:oai:doaj.org/article:5ad5231d8e26418dad07cfcadc405c62 |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:5ad5231d8e26418dad07cfcadc405c62 2023-05-15T18:25:52+02:00 LSCE-FFNN-v1: a two-step neural network model for the reconstruction of surface ocean p CO 2 over the global ocean A. Denvil-Sommer M. Gehlen M. Vrac C. Mejia 2019-05-01T00:00:00Z https://doi.org/10.5194/gmd-12-2091-2019 https://doaj.org/article/5ad5231d8e26418dad07cfcadc405c62 EN eng Copernicus Publications https://www.geosci-model-dev.net/12/2091/2019/gmd-12-2091-2019.pdf https://doaj.org/toc/1991-959X https://doaj.org/toc/1991-9603 doi:10.5194/gmd-12-2091-2019 1991-959X 1991-9603 https://doaj.org/article/5ad5231d8e26418dad07cfcadc405c62 Geoscientific Model Development, Vol 12, Pp 2091-2105 (2019) Geology QE1-996.5 article 2019 ftdoajarticles https://doi.org/10.5194/gmd-12-2091-2019 2022-12-31T14:46:35Z A new feed-forward neural network (FFNN) model is presented to reconstruct surface ocean partial pressure of carbon dioxide ( p CO 2 ) over the global ocean. The model consists of two steps: (1) the reconstruction of p CO 2 climatology, and (2) the reconstruction of p CO 2 anomalies with respect to the climatology. For the first step, a gridded climatology was used as the target, along with sea surface salinity (SSS), sea surface temperature (SST), sea surface height (SSH), chlorophyll a (Chl a ), mixed layer depth (MLD), as well as latitude and longitude as predictors. For the second step, data from the Surface Ocean CO 2 Atlas (SOCAT) provided the target. The same set of predictors was used during step (2) augmented by their anomalies. During each step, the FFNN model reconstructs the nonlinear relationships between p CO 2 and the ocean predictors. It provides monthly surface ocean p CO 2 distributions on a <math xmlns="http://www.w3.org/1998/Math/MathML" id="M12" display="inline" overflow="scroll" dspmath="mathml"><mrow><mn mathvariant="normal">1</mn><msup><mi/><mo>∘</mo></msup><mo>×</mo><mn mathvariant="normal">1</mn><msup><mi/><mo>∘</mo></msup></mrow></math> <svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="34pt" height="11pt" class="svg-formula" dspmath="mathimg" md5hash="adb4f21844a1bb199b77c90adc5df69c"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="gmd-12-2091-2019-ie00001.svg" width="34pt" height="11pt" src="gmd-12-2091-2019-ie00001.png"/></svg:svg> grid for the period from 2001 to 2016. Global ocean p CO 2 was reconstructed with satisfying accuracy compared with independent observational data from SOCAT. However, errors were larger in regions with poor data coverage (e.g., the Indian Ocean, the Southern Ocean and the subpolar Pacific). The model captured the strong interannual variability of surface ocean p CO 2 with reasonable skill over the ... Article in Journal/Newspaper Southern Ocean Directory of Open Access Journals: DOAJ Articles Indian Pacific Southern Ocean Geoscientific Model Development 12 5 2091 2105 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
Geology QE1-996.5 |
spellingShingle |
Geology QE1-996.5 A. Denvil-Sommer M. Gehlen M. Vrac C. Mejia LSCE-FFNN-v1: a two-step neural network model for the reconstruction of surface ocean p CO 2 over the global ocean |
topic_facet |
Geology QE1-996.5 |
description |
A new feed-forward neural network (FFNN) model is presented to reconstruct surface ocean partial pressure of carbon dioxide ( p CO 2 ) over the global ocean. The model consists of two steps: (1) the reconstruction of p CO 2 climatology, and (2) the reconstruction of p CO 2 anomalies with respect to the climatology. For the first step, a gridded climatology was used as the target, along with sea surface salinity (SSS), sea surface temperature (SST), sea surface height (SSH), chlorophyll a (Chl a ), mixed layer depth (MLD), as well as latitude and longitude as predictors. For the second step, data from the Surface Ocean CO 2 Atlas (SOCAT) provided the target. The same set of predictors was used during step (2) augmented by their anomalies. During each step, the FFNN model reconstructs the nonlinear relationships between p CO 2 and the ocean predictors. It provides monthly surface ocean p CO 2 distributions on a <math xmlns="http://www.w3.org/1998/Math/MathML" id="M12" display="inline" overflow="scroll" dspmath="mathml"><mrow><mn mathvariant="normal">1</mn><msup><mi/><mo>∘</mo></msup><mo>×</mo><mn mathvariant="normal">1</mn><msup><mi/><mo>∘</mo></msup></mrow></math> <svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="34pt" height="11pt" class="svg-formula" dspmath="mathimg" md5hash="adb4f21844a1bb199b77c90adc5df69c"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="gmd-12-2091-2019-ie00001.svg" width="34pt" height="11pt" src="gmd-12-2091-2019-ie00001.png"/></svg:svg> grid for the period from 2001 to 2016. Global ocean p CO 2 was reconstructed with satisfying accuracy compared with independent observational data from SOCAT. However, errors were larger in regions with poor data coverage (e.g., the Indian Ocean, the Southern Ocean and the subpolar Pacific). The model captured the strong interannual variability of surface ocean p CO 2 with reasonable skill over the ... |
format |
Article in Journal/Newspaper |
author |
A. Denvil-Sommer M. Gehlen M. Vrac C. Mejia |
author_facet |
A. Denvil-Sommer M. Gehlen M. Vrac C. Mejia |
author_sort |
A. Denvil-Sommer |
title |
LSCE-FFNN-v1: a two-step neural network model for the reconstruction of surface ocean p CO 2 over the global ocean |
title_short |
LSCE-FFNN-v1: a two-step neural network model for the reconstruction of surface ocean p CO 2 over the global ocean |
title_full |
LSCE-FFNN-v1: a two-step neural network model for the reconstruction of surface ocean p CO 2 over the global ocean |
title_fullStr |
LSCE-FFNN-v1: a two-step neural network model for the reconstruction of surface ocean p CO 2 over the global ocean |
title_full_unstemmed |
LSCE-FFNN-v1: a two-step neural network model for the reconstruction of surface ocean p CO 2 over the global ocean |
title_sort |
lsce-ffnn-v1: a two-step neural network model for the reconstruction of surface ocean p co 2 over the global ocean |
publisher |
Copernicus Publications |
publishDate |
2019 |
url |
https://doi.org/10.5194/gmd-12-2091-2019 https://doaj.org/article/5ad5231d8e26418dad07cfcadc405c62 |
geographic |
Indian Pacific Southern Ocean |
geographic_facet |
Indian Pacific Southern Ocean |
genre |
Southern Ocean |
genre_facet |
Southern Ocean |
op_source |
Geoscientific Model Development, Vol 12, Pp 2091-2105 (2019) |
op_relation |
https://www.geosci-model-dev.net/12/2091/2019/gmd-12-2091-2019.pdf https://doaj.org/toc/1991-959X https://doaj.org/toc/1991-9603 doi:10.5194/gmd-12-2091-2019 1991-959X 1991-9603 https://doaj.org/article/5ad5231d8e26418dad07cfcadc405c62 |
op_doi |
https://doi.org/10.5194/gmd-12-2091-2019 |
container_title |
Geoscientific Model Development |
container_volume |
12 |
container_issue |
5 |
container_start_page |
2091 |
op_container_end_page |
2105 |
_version_ |
1766207555878518784 |