Empirical methods for the estimation of Southern Ocean CO2: support vector and random forest regression
The Southern Ocean accounts for 40% of oceanic CO2 uptake, but the estimates are bound by large uncertainties due to a paucity in observations. Gap-filling empirical methods have been used to good effect to approximate pCO(2) from satellite observable variables in other parts of the ocean, but many...
Published in: | Biogeosciences |
---|---|
Main Authors: | , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Copernicus Gesellschaft Mbh
2017
|
Subjects: | |
Online Access: | https://archimer.ifremer.fr/doc/00662/77388/79018.pdf https://archimer.ifremer.fr/doc/00662/77388/79019.pdf https://doi.org/10.5194/bg-14-5551-2017 https://archimer.ifremer.fr/doc/00662/77388/ |
id |
ftarchimer:oai:archimer.ifremer.fr:77388 |
---|---|
record_format |
openpolar |
spelling |
ftarchimer:oai:archimer.ifremer.fr:77388 2023-05-15T18:24:49+02:00 Empirical methods for the estimation of Southern Ocean CO2: support vector and random forest regression Gregor, Luke Kok, Schalk Monteiro, Pedro M. S. 2017-12 application/pdf https://archimer.ifremer.fr/doc/00662/77388/79018.pdf https://archimer.ifremer.fr/doc/00662/77388/79019.pdf https://doi.org/10.5194/bg-14-5551-2017 https://archimer.ifremer.fr/doc/00662/77388/ eng eng Copernicus Gesellschaft Mbh https://archimer.ifremer.fr/doc/00662/77388/79018.pdf https://archimer.ifremer.fr/doc/00662/77388/79019.pdf doi:10.5194/bg-14-5551-2017 https://archimer.ifremer.fr/doc/00662/77388/ info:eu-repo/semantics/openAccess restricted use Biogeosciences (1726-4170) (Copernicus Gesellschaft Mbh), 2017-12 , Vol. 14 , N. 23 , P. 5551-5569 text Publication info:eu-repo/semantics/article 2017 ftarchimer https://doi.org/10.5194/bg-14-5551-2017 2021-09-23T20:36:29Z The Southern Ocean accounts for 40% of oceanic CO2 uptake, but the estimates are bound by large uncertainties due to a paucity in observations. Gap-filling empirical methods have been used to good effect to approximate pCO(2) from satellite observable variables in other parts of the ocean, but many of these methods are not in agreement in the Southern Ocean. In this study we propose two additional methods that perform well in the Southern Ocean: support vector regression (SVR) and random forest regression (RFR). The methods are used to estimate Delta pCO(2) in the Southern Ocean based on SOCAT v3, achieving similar trends to the SOM-FFN method by Landschitzer et al. (2014). Results show that the SOM-FFN and RFR approaches have RMSEs of similar magnitude (14.84 and 16.45 mu atm, where 1 atm D 101 325 Pa) where the SVR method has a larger RMSE (24.40 mu atm). However, the larger errors for SVR and RFR are, in part, due to an increase in coastal observations from SOCAT v2 to v3, where the SOM-FFN method used v2 data. The success of both SOM-FFN and RFR depends on the ability to adapt to different modes of variability. The SOM-FFN achieves this by having independent regression models for each cluster, while this flexibility is intrinsic to the RFR method. Analyses of the estimates shows that the SVR and RFR's respective sensitivity and robustness to out-liers define the outcome significantly. Further analyses on the methods were performed by using a synthetic dataset to assess the following: which method (RFR or SVR) has the best performance? What is the effect of using time, latitude and longitude as proxy variables on Delta pCO(2)? What is the impact of the sampling bias in the SOCAT v3 dataset on the estimates? We find that while RFR is indeed better than SVR, the ensemble of the two methods outperforms either one, due to complementary strengths and weaknesses of the methods. Results also show that for the RFR and SVR implementations, it is better to include coordinates as proxy variables as RMSE scores are lowered and the phasing of the seasonal cycle is more accurate. Lastly, we show that there is only a weak bias due to undersampling. The synthetic data provide a useful framework to test methods in regions of sparse data coverage and show potential as a useful tool to evaluate methods in future studies. Article in Journal/Newspaper Southern Ocean Archimer (Archive Institutionnelle de l'Ifremer - Institut français de recherche pour l'exploitation de la mer) Southern Ocean Biogeosciences 14 23 5551 5569 |
institution |
Open Polar |
collection |
Archimer (Archive Institutionnelle de l'Ifremer - Institut français de recherche pour l'exploitation de la mer) |
op_collection_id |
ftarchimer |
language |
English |
description |
The Southern Ocean accounts for 40% of oceanic CO2 uptake, but the estimates are bound by large uncertainties due to a paucity in observations. Gap-filling empirical methods have been used to good effect to approximate pCO(2) from satellite observable variables in other parts of the ocean, but many of these methods are not in agreement in the Southern Ocean. In this study we propose two additional methods that perform well in the Southern Ocean: support vector regression (SVR) and random forest regression (RFR). The methods are used to estimate Delta pCO(2) in the Southern Ocean based on SOCAT v3, achieving similar trends to the SOM-FFN method by Landschitzer et al. (2014). Results show that the SOM-FFN and RFR approaches have RMSEs of similar magnitude (14.84 and 16.45 mu atm, where 1 atm D 101 325 Pa) where the SVR method has a larger RMSE (24.40 mu atm). However, the larger errors for SVR and RFR are, in part, due to an increase in coastal observations from SOCAT v2 to v3, where the SOM-FFN method used v2 data. The success of both SOM-FFN and RFR depends on the ability to adapt to different modes of variability. The SOM-FFN achieves this by having independent regression models for each cluster, while this flexibility is intrinsic to the RFR method. Analyses of the estimates shows that the SVR and RFR's respective sensitivity and robustness to out-liers define the outcome significantly. Further analyses on the methods were performed by using a synthetic dataset to assess the following: which method (RFR or SVR) has the best performance? What is the effect of using time, latitude and longitude as proxy variables on Delta pCO(2)? What is the impact of the sampling bias in the SOCAT v3 dataset on the estimates? We find that while RFR is indeed better than SVR, the ensemble of the two methods outperforms either one, due to complementary strengths and weaknesses of the methods. Results also show that for the RFR and SVR implementations, it is better to include coordinates as proxy variables as RMSE scores are lowered and the phasing of the seasonal cycle is more accurate. Lastly, we show that there is only a weak bias due to undersampling. The synthetic data provide a useful framework to test methods in regions of sparse data coverage and show potential as a useful tool to evaluate methods in future studies. |
format |
Article in Journal/Newspaper |
author |
Gregor, Luke Kok, Schalk Monteiro, Pedro M. S. |
spellingShingle |
Gregor, Luke Kok, Schalk Monteiro, Pedro M. S. Empirical methods for the estimation of Southern Ocean CO2: support vector and random forest regression |
author_facet |
Gregor, Luke Kok, Schalk Monteiro, Pedro M. S. |
author_sort |
Gregor, Luke |
title |
Empirical methods for the estimation of Southern Ocean CO2: support vector and random forest regression |
title_short |
Empirical methods for the estimation of Southern Ocean CO2: support vector and random forest regression |
title_full |
Empirical methods for the estimation of Southern Ocean CO2: support vector and random forest regression |
title_fullStr |
Empirical methods for the estimation of Southern Ocean CO2: support vector and random forest regression |
title_full_unstemmed |
Empirical methods for the estimation of Southern Ocean CO2: support vector and random forest regression |
title_sort |
empirical methods for the estimation of southern ocean co2: support vector and random forest regression |
publisher |
Copernicus Gesellschaft Mbh |
publishDate |
2017 |
url |
https://archimer.ifremer.fr/doc/00662/77388/79018.pdf https://archimer.ifremer.fr/doc/00662/77388/79019.pdf https://doi.org/10.5194/bg-14-5551-2017 https://archimer.ifremer.fr/doc/00662/77388/ |
geographic |
Southern Ocean |
geographic_facet |
Southern Ocean |
genre |
Southern Ocean |
genre_facet |
Southern Ocean |
op_source |
Biogeosciences (1726-4170) (Copernicus Gesellschaft Mbh), 2017-12 , Vol. 14 , N. 23 , P. 5551-5569 |
op_relation |
https://archimer.ifremer.fr/doc/00662/77388/79018.pdf https://archimer.ifremer.fr/doc/00662/77388/79019.pdf doi:10.5194/bg-14-5551-2017 https://archimer.ifremer.fr/doc/00662/77388/ |
op_rights |
info:eu-repo/semantics/openAccess restricted use |
op_doi |
https://doi.org/10.5194/bg-14-5551-2017 |
container_title |
Biogeosciences |
container_volume |
14 |
container_issue |
23 |
container_start_page |
5551 |
op_container_end_page |
5569 |
_version_ |
1766205738418438144 |