Assessing improvements in global ocean pCO<sub>2</sub> machine learning reconstructions with Southern Ocean autonomous sampling

Heimdal, Thea Hatlen; McKinley, Galen A.; Sutton, Adrienne J.; Fay, Amanda R.; Gloege, Lucas

doi:https://doi.org/10.5194/bg-2023-160

Preprints

https://doi.org/10.5194/bg-2023-160

Preprints

10 Oct 2023

| 10 Oct 2023

Status: a revised version of this preprint was accepted for the journal BG and is expected to appear here in due course.

Assessing improvements in global ocean pCO₂ machine learning reconstructions with Southern Ocean autonomous sampling

Thea Hatlen Heimdal, Galen A. McKinley, Adrienne J. Sutton, Amanda R. Fay, and Lucas Gloege

Abstract. The Southern Ocean plays an important role in the exchange of carbon between the atmosphere and oceans, and is a critical region for the ocean uptake of anthropogenic CO₂. However, estimates of the Southern Ocean air-sea CO₂ flux are highly uncertain due to limited data coverage. Increased sampling in winter and across meridional gradients in the Southern Ocean may improve machine learning (ML) reconstructions of global surface ocean pCO₂. Here, we use a Large Ensemble Testbed (LET) of Earth System Models and the pCO₂-Residual reconstruction method to assess improvements in pCO₂ reconstruction fidelity that could be achieved with additional autonomous sampling in the Southern Ocean added to existing Surface Ocean CO₂ Atlas (SOCAT) observations. The LET allows us to robustly evaluate the skill of pCO₂ reconstructions in space and time through comparison to ‘model truth’. With only SOCAT sampling, Southern Ocean and global pCO₂ are overestimated, and thus the ocean carbon sink is underestimated. Incorporating Uncrewed Surface Vehicle (USV) sampling increases the spatial and seasonal coverage of observations within the Southern Ocean, leading to a decrease in the overestimation of pCO₂. A modest number of additional observations in southern hemisphere winter and across meridional gradients in the Southern Ocean leads to improvement in reconstruction bias and root-mean squared error (RMSE) can be improved by as much as 65 % and 19 %, respectively, as compared to using SOCAT sampling alone. Lastly, the large decadal variability of air-sea CO₂ fluxes shown by SOCAT-only sampling, may be partially attributable to undersampling of the Southern Ocean.

Received: 13 Sep 2023 – Discussion started: 10 Oct 2023

Download & links

Preprint (PDF, 9421 KB)

Supplement (14679 KB)

Download & links

Preprint (9421 KB)
Metadata XML
Supplement (14679 KB)
BibTeX
EndNote

Thea Hatlen Heimdal, Galen A. McKinley, Adrienne J. Sutton, Amanda R. Fay, and Lucas Gloege

Status: closed

RC1:
'Comment on bg-2023-160', Anonymous Referee #1, 09 Nov 2023
General comments:
Heildal et al., assess how much our estimate of the annual ocean CO₂ sink would improve if we had regular autonomous sampling of the partial pressure of CO₂ in surface seawater (pCO₂) from Saildrones in the Southern Ocean. To do this, observing system simulation experiments were carried out with a large ensemble testbed (LET) of Earth System Models and the authors’ observation-based method for estimating the ocean CO₂ sink. In other words, they are evaluating the influence of the availability of observations on the performance of their method.
Such analysis is particularly relevant because we are increasingly relying on observation-based methodologies to estimate the annual ocean CO₂ sink, and these methods therefore need to be evaluated more systematically. There is also an alarming decline in the ocean CO₂ observing capacity that autonomous platforms (such as Saildrone) should help to counter.
However, (i) I have some reservations about the way the authors have used their method, (ii) I would expect a more quantitative estimate of the magnitude of the differences potentially detected between the different experiments performed, and (iii) the results and discussion of the decadal variability of the oceanic CO₂ sink need to be improved (see specific comments). As a result, the paper is likely to be a significant scientific contribution with some revisions.
Specific comments:
1) To calculate the residual pCO₂, the authors used the equation in line 160. In this equation, the term pCO₂^mean came from “surface ocean pCO₂ […] using all 1°x1° grid cells from the testbed” (line 162). But in the original publication of the methodology (equation 2 in Bennington et al., 2022a), the term pCO₂^mean comes from an initial reconstruction of pCO₂. Therefore, in this submitted manuscript, by using model outputs instead of an initial reconstruction of pCO₂ fields, the authors assumed that their method would be able to perfectly reconstruct the long-term average pCO₂ at each grid cell. It seems to me that to obtain a more accurate evaluation of their method (i.e., more accurate observing system simulation experiments) the authors should follow the steps as they were originally published. If this is not possible, could the authors explain why and prove that this assumption does not influence their results.
2) To calculate the net sea–air CO₂ flux, authors used (line 272): “EN4.2.2 salinity (Good et al., 2013), SST and ice fraction from NOAA Optimum Interpolation Sea Surface Temperature V2 (OISSTv2) (Reynolds et al., 2002), and surface winds and associated wind scaling factor from the European Centre for Medium-Range Weather Forecasts (ECMWF ERA5 sea level pressure (Hersbach et al., 2020)”. But as mentioned in line 95: “The goal here is to assess the accuracy with which an ML algorithm can reconstruct the ‘model truth’”. Therefore, I would expect the model outputs (some of which have already been used for pCO₂ reconstruction) to be used to calculate the CO₂ flux, rather than observational data which may have different variabilities and/or trends to those simulated. The authors could then compare this calculated CO₂ flux to the simulated CO₂ flux (and not to a "model truth" CO₂ flux from simulated pCO₂ fields mixed with observational data). This is particularly important when the authors are discussing the ability of their method to reproduce CO₂ flux variability (see my next comment).
3) Authors wrote line 531: “The SOCAT baseline demonstrates a weakening of the global and Southern Ocean carbon sink in the 2000s (Figs. 10, S12), which is in agreement with various data products using real-world SOCAT data”. The weakening of the Southern Ocean carbon sink occurred in the 1990s (Le Quéré et al., 2007), while a reinvigoration of the sink was observed during the 2000s (Landschützer et al., 2015). The authors therefore need to revise their text. More importantly, this study focuses on the ability of the authors' method to reproduce "model-truth" variability and not the "real-world" variability. Consequently, I would suggest calculating certain metrics of variability (for example, the size of decadal variability or trends) from simulated CO₂ fluxes (and not from recalculated "model-truth" CO₂ fluxes, see my previous comment) and comparing the values of these metrics with the values that would be obtained when reconstructed CO₂ fluxes are used. Because, otherwise, it assumes that all models perfectly reproduce the variability of the 'real world', which might not be the case.
4) The LET has 75 members (i.e., simulations). For each experiment, the values given in the manuscript and in the figures are for the most part averages calculated over the 75 members of the ensemble. But no information is given on the dispersion (or confidence interval) around these averages. It is therefore not possible to assess whether the differences mentioned between the experiments are significant or not.
For example:
The interpretation of Figure 5 (line 335): “The ‘one-latitude’ ‘high-sampling’ run ‘x13_10Y_J-A’ (44,250 observations) show similar bias or is outperformed by all ‘zigzag’ runs as well as the ‘one-latitude’-runs that restrict sampling to southern hemisphere winter months (i.e., ‘x5_5Y_W’ and ‘x13_10Y_W’).” How similar or superior is the performance? Is it true for all members?

Line 346: “Run ‘Z_x10_5Y_W’, which has the lowest bias out of the ‘zigzag’ runs (Fig. 5), shows improvement even further back in time, until the beginning of the testbed period (Fig. S6).” Is it really significant?

I would therefore suggest not only reporting the averages over the 75 members, but also taking advantage of the study of the spread around these averages.
Technical corrections:
5) Line 37, Please explain the acronym "fCO₂". Note that the term "pCO₂" is also used in the manuscript. Although understandable to researchers working on this topic, it is less clear to a wider audience. Therefore, authors should be more careful about the terms they use, especially in the Abstract and Introduction sections.
6) Line 178, Why aren't the number of decision trees and depth levels different for each reconstruction?
7) Line 188, After reading this sentence, I wasn't sure whether the authors were always going to use the "unseen" values. Could the authors be clearer?
8) Line 203: “2) potential future meridional USV observations (‘zigzag’ track)”. Are they realistic? I found some elements of response later in the text, but it would be good to know here whether all the experiments are realistic or not.
9) Table 1: I suggest replacing table 1 with table S1. This is because the information in table 1 is repeated in table S1, and table S1 contains important values that the reader should be able to access easily.
10) Line 268, Why not use the same method across all models to calculate pCO₂^atm? Do all the values obtained take into account the contribution of water vapor pressure?
11) Line 293: “where algorithm generally overestimates pCO₂”. This is not the case for the Atlantic sector of the Southern Ocean.
12) Figure 3, colour scale: The colour scales need to be harmonised. In panel a, a white colour means a good value, whereas in panel b, it means a bad value.
13) Figure 3, line 301 to 307: All this information is already present in the text. Please write shorter figure captions. This is a general comment, not just on figure 3.
14) Line 318 and wherever necessary in the text: “…where the baseline reconstruction…” Please, use the expression "SOCAT baseline" that was introduced in the method section.
15) Line 384, Please delete the reference to "bias". This was introduced in the previous section.
16) Line 493, why not excluding the hypothetical data points that would be covered by sea ice?
17) Figure 10: The figure starts in 1985 and not 1982, why?
18) Figure S3: Because you focused on the open-ocean (line 123), non-open-ocean data should be removed as they were not use for the training, is it right? Does this drastically modified the data availability and explain why better results are obtained from 1990?
Citation: https://doi.org/10.5194/bg-2023-160-RC1
- AC1: 'Reply on RC1', Thea Hatlen Heimdal, 21 Dec 2023
  
  See attached PDF
  
  Citation: https://doi.org/10.5194/bg-2023-160-AC1
RC2:
'Comment on bg-2023-160', Anonymous Referee #2, 15 Nov 2023

Please kindly find Reviewer's comments in the file attached.

Citation: https://doi.org/10.5194/bg-2023-160-RC2
- AC2: 'Reply on RC2', Thea Hatlen Heimdal, 21 Dec 2023
  
  See attached PDF.
  
  Citation: https://doi.org/10.5194/bg-2023-160-AC2

Status: closed

RC1:
'Comment on bg-2023-160', Anonymous Referee #1, 09 Nov 2023
General comments:
Heildal et al., assess how much our estimate of the annual ocean CO₂ sink would improve if we had regular autonomous sampling of the partial pressure of CO₂ in surface seawater (pCO₂) from Saildrones in the Southern Ocean. To do this, observing system simulation experiments were carried out with a large ensemble testbed (LET) of Earth System Models and the authors’ observation-based method for estimating the ocean CO₂ sink. In other words, they are evaluating the influence of the availability of observations on the performance of their method.
Such analysis is particularly relevant because we are increasingly relying on observation-based methodologies to estimate the annual ocean CO₂ sink, and these methods therefore need to be evaluated more systematically. There is also an alarming decline in the ocean CO₂ observing capacity that autonomous platforms (such as Saildrone) should help to counter.
However, (i) I have some reservations about the way the authors have used their method, (ii) I would expect a more quantitative estimate of the magnitude of the differences potentially detected between the different experiments performed, and (iii) the results and discussion of the decadal variability of the oceanic CO₂ sink need to be improved (see specific comments). As a result, the paper is likely to be a significant scientific contribution with some revisions.
Specific comments:
1) To calculate the residual pCO₂, the authors used the equation in line 160. In this equation, the term pCO₂^mean came from “surface ocean pCO₂ […] using all 1°x1° grid cells from the testbed” (line 162). But in the original publication of the methodology (equation 2 in Bennington et al., 2022a), the term pCO₂^mean comes from an initial reconstruction of pCO₂. Therefore, in this submitted manuscript, by using model outputs instead of an initial reconstruction of pCO₂ fields, the authors assumed that their method would be able to perfectly reconstruct the long-term average pCO₂ at each grid cell. It seems to me that to obtain a more accurate evaluation of their method (i.e., more accurate observing system simulation experiments) the authors should follow the steps as they were originally published. If this is not possible, could the authors explain why and prove that this assumption does not influence their results.
2) To calculate the net sea–air CO₂ flux, authors used (line 272): “EN4.2.2 salinity (Good et al., 2013), SST and ice fraction from NOAA Optimum Interpolation Sea Surface Temperature V2 (OISSTv2) (Reynolds et al., 2002), and surface winds and associated wind scaling factor from the European Centre for Medium-Range Weather Forecasts (ECMWF ERA5 sea level pressure (Hersbach et al., 2020)”. But as mentioned in line 95: “The goal here is to assess the accuracy with which an ML algorithm can reconstruct the ‘model truth’”. Therefore, I would expect the model outputs (some of which have already been used for pCO₂ reconstruction) to be used to calculate the CO₂ flux, rather than observational data which may have different variabilities and/or trends to those simulated. The authors could then compare this calculated CO₂ flux to the simulated CO₂ flux (and not to a "model truth" CO₂ flux from simulated pCO₂ fields mixed with observational data). This is particularly important when the authors are discussing the ability of their method to reproduce CO₂ flux variability (see my next comment).
3) Authors wrote line 531: “The SOCAT baseline demonstrates a weakening of the global and Southern Ocean carbon sink in the 2000s (Figs. 10, S12), which is in agreement with various data products using real-world SOCAT data”. The weakening of the Southern Ocean carbon sink occurred in the 1990s (Le Quéré et al., 2007), while a reinvigoration of the sink was observed during the 2000s (Landschützer et al., 2015). The authors therefore need to revise their text. More importantly, this study focuses on the ability of the authors' method to reproduce "model-truth" variability and not the "real-world" variability. Consequently, I would suggest calculating certain metrics of variability (for example, the size of decadal variability or trends) from simulated CO₂ fluxes (and not from recalculated "model-truth" CO₂ fluxes, see my previous comment) and comparing the values of these metrics with the values that would be obtained when reconstructed CO₂ fluxes are used. Because, otherwise, it assumes that all models perfectly reproduce the variability of the 'real world', which might not be the case.
4) The LET has 75 members (i.e., simulations). For each experiment, the values given in the manuscript and in the figures are for the most part averages calculated over the 75 members of the ensemble. But no information is given on the dispersion (or confidence interval) around these averages. It is therefore not possible to assess whether the differences mentioned between the experiments are significant or not.
For example:
The interpretation of Figure 5 (line 335): “The ‘one-latitude’ ‘high-sampling’ run ‘x13_10Y_J-A’ (44,250 observations) show similar bias or is outperformed by all ‘zigzag’ runs as well as the ‘one-latitude’-runs that restrict sampling to southern hemisphere winter months (i.e., ‘x5_5Y_W’ and ‘x13_10Y_W’).” How similar or superior is the performance? Is it true for all members?

Line 346: “Run ‘Z_x10_5Y_W’, which has the lowest bias out of the ‘zigzag’ runs (Fig. 5), shows improvement even further back in time, until the beginning of the testbed period (Fig. S6).” Is it really significant?

I would therefore suggest not only reporting the averages over the 75 members, but also taking advantage of the study of the spread around these averages.
Technical corrections:
5) Line 37, Please explain the acronym "fCO₂". Note that the term "pCO₂" is also used in the manuscript. Although understandable to researchers working on this topic, it is less clear to a wider audience. Therefore, authors should be more careful about the terms they use, especially in the Abstract and Introduction sections.
6) Line 178, Why aren't the number of decision trees and depth levels different for each reconstruction?
7) Line 188, After reading this sentence, I wasn't sure whether the authors were always going to use the "unseen" values. Could the authors be clearer?
8) Line 203: “2) potential future meridional USV observations (‘zigzag’ track)”. Are they realistic? I found some elements of response later in the text, but it would be good to know here whether all the experiments are realistic or not.
9) Table 1: I suggest replacing table 1 with table S1. This is because the information in table 1 is repeated in table S1, and table S1 contains important values that the reader should be able to access easily.
10) Line 268, Why not use the same method across all models to calculate pCO₂^atm? Do all the values obtained take into account the contribution of water vapor pressure?
11) Line 293: “where algorithm generally overestimates pCO₂”. This is not the case for the Atlantic sector of the Southern Ocean.
12) Figure 3, colour scale: The colour scales need to be harmonised. In panel a, a white colour means a good value, whereas in panel b, it means a bad value.
13) Figure 3, line 301 to 307: All this information is already present in the text. Please write shorter figure captions. This is a general comment, not just on figure 3.
14) Line 318 and wherever necessary in the text: “…where the baseline reconstruction…” Please, use the expression "SOCAT baseline" that was introduced in the method section.
15) Line 384, Please delete the reference to "bias". This was introduced in the previous section.
16) Line 493, why not excluding the hypothetical data points that would be covered by sea ice?
17) Figure 10: The figure starts in 1985 and not 1982, why?
18) Figure S3: Because you focused on the open-ocean (line 123), non-open-ocean data should be removed as they were not use for the training, is it right? Does this drastically modified the data availability and explain why better results are obtained from 1990?
Citation: https://doi.org/10.5194/bg-2023-160-RC1
- AC1: 'Reply on RC1', Thea Hatlen Heimdal, 21 Dec 2023
  
  See attached PDF
  
  Citation: https://doi.org/10.5194/bg-2023-160-AC1
RC2:
'Comment on bg-2023-160', Anonymous Referee #2, 15 Nov 2023

Please kindly find Reviewer's comments in the file attached.

Citation: https://doi.org/10.5194/bg-2023-160-RC2
- AC2: 'Reply on RC2', Thea Hatlen Heimdal, 21 Dec 2023
  
  See attached PDF.
  
  Citation: https://doi.org/10.5194/bg-2023-160-AC2

Thea Hatlen Heimdal, Galen A. McKinley, Adrienne J. Sutton, Amanda R. Fay, and Lucas Gloege

Supplement

https://doi.org/10.5194/bg-2023-160-supplement

Thea Hatlen Heimdal, Galen A. McKinley, Adrienne J. Sutton, Amanda R. Fay, and Lucas Gloege

Viewed

Total article views: 562 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
401	139	22	562	32	19	19

HTML: 401
PDF: 139
XML: 22
Total: 562
Supplement: 32
BibTeX: 19
EndNote: 19

Views and downloads (calculated since 10 Oct 2023)

Month	HTML	PDF	XML	Total
Oct 2023	139	45	6	190
Nov 2023	44	19	1	64
Dec 2023	34	19	10	63
Jan 2024	30	11	1	42
Feb 2024	43	18	0	61
Mar 2024	46	14	0	60
Apr 2024	65	13	4	82

Cumulative views and downloads (calculated since 10 Oct 2023)

Month	HTML	PDF	XML	Total
Oct 2023	139	45	6	190
Nov 2023	44	19	1	64
Dec 2023	34	19	10	63
Jan 2024	30	11	1	42
Feb 2024	43	18	0	61
Mar 2024	46	14	0	60
Apr 2024	65	13	4	82

Viewed (geographical distribution)

Total article views: 571 (including HTML, PDF, and XML) Thereof 571 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Apr 2024

Download

Preprint (9421 KB)
Metadata XML

Short summary

Measurements of ocean carbon are limited in time and space. Machine learning algorithms are therefore used to reconstruct ocean carbon where observations do not exist. Improving these reconstructions is important in order to accurately estimate how much carbon the ocean absorbs from the atmosphere. In this study, we find that that a small addition of observations from the Southern Ocean, obtained by autonomous sampling platforms, could significantly improve the reconstructions.


Total:	0
HTML:	0
PDF:	0
XML:	0

Assessing improvements in global ocean pCO2 machine learning reconstructions with Southern Ocean autonomous sampling

Supplement

Viewed

Viewed (geographical distribution)

Assessing improvements in global ocean pCO₂ machine learning reconstructions with Southern Ocean autonomous sampling