;;;;;;; General comment

In this manuscript the authors present new data obtained from two Saildrones Uncrewed Surface Vehicles (USV) in the Indian circumpolar region in the frame of the now called Southern Ocean Saildrone (SOS) mission. After a short (but important) history of the deployments and decisions during the SO-CHIC project (COVID issues, batteries, etc…), authors describe the properties measured that include T, S, O2, pCO2, Fluo-chla, atmospheric records, wind, and ADCP. However, they do not detail the accuracy/precision of the data and this should be added (e.g. in Table 1). They present preliminary interesting results of the pCO2 distribution related to eddies and/or Chl-a. Analysis of the processes that drive the pCO2 and air-sea CO2 flux variability in this complex frontal zone would be presented in other dedicated papers. Authors focused the results on eddies (primary goal of the project), but I guess all data including in “calm” regions (if any along the USV tracks) and for different months are interesting especially for data-based methods that use gridded SOCAT products (monthly at 1degx1deg) to reconstruct pCO2 and air-sea CO2 flux fields (e.g. SOMFFN); the SOCAT gridded product is somehow a smooth field compared to the high frequency and high resolution data presented here. Specifically, authors showed that large variability could occur at small scale (up to 40 µatm), suggesting that data-based methods should be adapted to better represent such variability. Apart the link with eddies, an interesting result, not discussed or highlighted in the MS, is that pCO2 was lower than in the atmosphere for almost all periods (Figure 4), i.e. a CO2 sink; this is different than observed by Sutton et al (2021) using data from the first circumpolar USV and who observed CO2 sources in winter. I was wondering why authors did not compare their data with the 2019 SD (Sutton et al, 2021). Is it because the tracks were not in the same region ?.  Note also that fCO2 data from several cruises are available in the investigated region in 2022-2024. Although the periods are not exactly the same (i.e. no direct cross-over), comparisons might highlight what the USV offers that could be not interpreted from shipboard data. Authors might also recall that, in the frame of the SO-CHIC project, a CARIOCA surface drifter was deployed (S.A AgulhasII cruise) and recorded pCO2 data in the South Indian as well (see Expocode 91AH20220123 in SOCAT). From these data Naëck et al (2025) focused the interpretation in the Atlantic sector but present also the data in the Indian sector where they estimated a CO2 sink in summer and near equilibrium in June (see their Figure S2). Finally, as the USV also recorded skin temperature, it might be useful to present these data along with SST to evaluate their differences and discuss potential bias on the air-sea CO2 fluxes estimates (e.g. Ford et al, 2024). Is there a link of the SST-Tskin in specific regions (e.g. in eddies) or this has no impact on CO2 fluxes in the investigated region.

The new data in 2022-2023 presented here are complementary to other cruises in the south Indian Ocean most of them being conducted in austral summer, and few cruises in autumn/winter/spring in recent years. For one USV (SD1039), the fCO2 data have been included in SOCAT data-base (and QCed flag C in version V2025) and this should be notice in the manuscript. It would be useful that data from USV SD1038 be also included in SOCAT to offer new constraints for the methods that attempt to reconstruct the fCO2 and CO2 fluxes fields (e.g. Neural Network methods used each year for the Global Carbon Budget, Friedlingstein et al, 2025).

The manuscript is suitable for publication in ESSD after revision. Below are listed specific comments.

;;;;;;;;;; Specific comments:

C-01: Line 24: For the CO2 sink in the SO maybe refer also to Hauck et al, (2023)

C-02: Line 27: For the decadal changes of anthropogenic CO2 maybe refer also to Muller et al, (2023)

C-03: Line 41-42: Authors refer to Fig. 3 from Bakker et al., (2016) who present data for 1957-2014. Since 2014, millions of new data have been included in SOCAT and authors may show maps (e.g. 4 seasons) of the most recent SOCAT version (v2025, Bakker et al, 2025) and inform that much less data were obtained in austral winter as noticed in line 42. In such maps the tracks of the UVS would be highlighted (suggestion).

C-04: Line 51 and 61: Sutton et al (2023) not listed in references.

C-05: Line 74: “downstream of the Kerguelen Plateau”. For reader not familiar with this region, please add on the map (figure 1) the names of some locations (Kerguelen, Kerguelen Plateau, Crozet etc…).

C-06: Lines 77 and 101: Figure 1: The track of SD1039 seems to be located in the SAZ between the STF and the SAF. I would suggest add the location of the STF (and AGF) on the map. Also, I think the PF should be south of Kerguelen.

C-07: Line 136: Chiodi et al (2023) not listed in references.

C-08: Line 150: Table 1: Please add information on accuracy /precision of the probes.

C-09: Line 174: In addition to NCEI, authors may also indicate that fCO2 data for one UVS are available in SOCAT-v2025 (Bakker et al, 2025).

C-10, Line 254: Authors write: “The most obvious signal in measured pCO2 by SD1038 during its June-July transit from South Africa is an increase in pCO2 values from ~350 µatm at 35° S to ~405 µatm at 50°S as the vehicle moved southwards (Fig. 4a and 4b). These values are within an expected range, as Shadwick et al. (2023) documented seasonal variations at a similar latitude in a mooring south of Tasmania with a peak (380-400 µatm) around July/August”.  As the SD1038 was in the western Indian sector, maybe compare with other results in this region (e.g. figure 3 in Metzl et al, 2006 for summer and winter) not only south of Tasmania. It might be also useful to compare the SD data with the climatology from Takahashi et al (2009) or from Fay et al (2024), especially because the SD1038 data were not used to compose the climatology.

C-11, Line 260: Authors notice a bias between SD data and the climatology from Landschutzer et al (Figure 5). What was the reference year for this climatology ? Would the bias explained by the increase of oceanic pCO2 ? Would be better to show the difference for DpCO2. Why not using data from data-based products for the same period as for SOS ?

C-12: Line 282: Authors write: ”This suggests the potential for outgassing of CO2 from the ocean to the atmosphere during these periods, but more work would be required to fully quantify this.” This is somehow like the results presented by Sutton et al (2021) and this is why I suggested in the general comment to compare with the USV data recorded in 2019.

C-13: Line 284: In September-October previous cruises (e.g. 35MF19950928, 35MF20050919 or 35MV20161006 in SOCAT) also indicate low pCO2 (<340 µatm) around 45°E or 65°E that might be associated to higher Chla in the frontal zone.

C-14: Line 285: Could you recall how the Chl-a in µg/l was derived from fluorescence data.

C-15: Line 286: To separate the effect of biology it might be useful to show the SST in Figure 6 and pCO2 normalized to SST at 13 or 15°C.

C-16: Line 309: Is the largest pCO2 change linked to eddy or meander in the frontal structure. Like for the SST (comment above), it would be interesting to show the sea surface salinity records (rapid change of salinity between 35.5 and 34 at the SAF is common in this region).

;;;;;;; Notes on figures

C-17: Figures 1: Add names of some geographical locations Indian Ocean, etc…). Would it be possible to add bathymetry to highlight the Agulhas or the Kerguelen Plateau.

C-18: Figures 2: Add on the map the Eddy number as listed in Figure 7. Would Figure 2 moved juste before figure 7

C-19: Figures 4: Atmospheric pCO2 presented in the plots range between 400-420 (in µatm). Recall in the text that pCO2atm here is at atmospheric pressure, i.e. xCO2 in ppm would be about 412-416 ppm in the southern hemisphere in 2022-2023. By the way, a comparison with atmospheric CO2 recorded at Stations Crozet would also validate the atmospheric pCO2 data from the USV.

C-19b: Figures 5: In captions: “Also shown are values from a mean seasonal climatology (red, data from Landschützer et al., 2020a,b).” Should be corrected ?: “… from a mean monthly climatology…”

C-20: Figures 5: Present DpCO2 on the plot (better to explain the bias in this comparison)

C-21: Figures 6: Would be useful to show pCO2 normalized at temperature (at 10 or 15°C).

;;;;;;; Notes in References

C-22: Line 427: Reference Landschützer et al., 2015 for the SO flux (cited line 26): should be corrected:

Landschutzer, P., et al, 2015. The Reinvigoration of the Southern Ocean Carbon Sink. Science, Vol. 349 no. 6253 pp. 1221-1224, DOI:10.1126/science.aab2620

C-23: Line 372 and 374: Carter et al not cited in the MS

C-24: Line 375: Chai et al not cited in the MS

C-25: Line 401: Frenger et al not cited in the MS

C-26: Line 403: Friedlingstein et al not cited in the MS

C-27: Line 405 and 408: Gaube et al not cited in the MS

C-28: Line 411: Gray et al not cited in the MS

C-29: Line 414: Gregor et al not cited in the MS

C-30: Line 419 and 422: Johnson et al not cited in the MS

C-31: Line 441: McWilliams not cited in the MS

C-32: Line 448: Mongwe et al not cited in the MS

C-33: Line 451: Nevison et al not cited in the MS

C-34: Line 454: Omand et al not cited in the MS

C-35: Line 459: Resplandy et al not cited in the MS

C-36: Line 483: Talley et al not cited in the MS

C-37: Line 487: Viglione et al not cited in the MS

C-38: Line 490 and 494: Wanninkhof et al not cited in the MS

C-39: Line 496: Whalen et al not cited in the MS

C-40: Line 498: Williams et al 2016 not cited in the MS

C-41: Line 504: Williams et al 2018 not cited in the MS

;;;;;; Reference added in this review not listed in the Manuscript:

Bakker, Dorothee C. E. et al (2025). Surface Ocean CO2 Atlas Database Version 2025 (SOCATv2025) (NCEI Accession 0304549). NOAA National Centers for Environmental Information. https://doi.org/10.25921/648f-fv35.

Fay, A. R., Munro, D. R., McKinley, G. A., Pierrot, D., Sutherland, S. C., Sweeney, C., and Wanninkhof, R.: Updated climatological mean ΔfCO2 and net sea–air CO2 flux over the global open ocean regions, Earth Syst. Sci. Data, 16, 2123–2139, https://doi.org/10.5194/essd-16-2123-2024, 2024.

Ford, D.J., Shutler, J.D., Blanco-Sacristán, J. et al. Enhanced ocean CO2 uptake due to near-surface temperature gradients. Nat. Geosci. (2024). https://doi.org/10.1038/s41561-024-01570-7

Friedlingstein, P., et al,: Global Carbon Budget 2024, Earth Syst. Sci. Data, 17, 965–1039, https://doi.org/10.5194/essd-17-965-2025, 2025.

Hauck, J., Gregor, L., Nissen, C., Patara, L., Hague, M., Mongwe, P., et al. (2023). The Southern Ocean carbon cycle 1985–2018: Mean, seasonal cycle, trends, and storage. Global Biogeochemical Cycles, 37, e2023GB007848. https://doi.org/10.1029/2023GB007848

Metzl, N., C.Brunet, Jabaud-Jan A, A.Poisson and B.Schauer, 2006. Summer and winter air-sea CO2 fluxes in the Southern Ocean Deep Sea Res I, 53, 1548-1563, doi:10.1016/j.dsr.2006.07.006

Müller, J. D., Gruber, N., Carter, B., Feely, R., Ishii, M., Lange, N., et al. (2023). Decadal trends in the oceanic storage of anthropogenic carbon from 1994 to 2014. AGU Advances, 4, e2023AV000875. https://doi.org/10.1029/2023AV000875

Naëck, K., Boutin, J., Swart, S., du Plessis, M., Merlivat, L., Beaumont, L., Lourenco, A., d'Ovidio, F., Rousselet, L., Ward, B., and Sallée, J.-B.: Anomalous summertime CO2 sink in the subpolar Southern Ocean promoted by early 2021 sea ice retreat, Biogeosciences, 22, 1947–1968, https://doi.org/10.5194/bg-22-1947-2025, 2025.

Takahashi, T., S C. Sutherland, R.Wanninkhof, C. Sweeney, R.A. Feely, D. Chipman, B. Hales, G. Friederich, F. Chavez, A. Watson, D. Bakker, U. Schuster, N.Metzl, H.Y. Inoue, M. Ishii, T. Midorikawa, C.Sabine, M. Hoppema, J.Olafsson, T. Amarson, B.Tilbrook, T. Johannessen, A. Olsen, R. Bellerby, Y. Nojiri, C.S. Wong, B. Delille, N. Bates and H. De Baar, 2009. Climatological Mean and Decadal Change in Surface Ocean pCO2, and Net Sea-air CO2 Flux over the Global Oceans. Deep-Sea Res II, doi:10.1016/j.dsr2.2008.12.009

;;;;;;;;;;;;;;;;;; end review

Review of the manuscript essd-2025-360 « Atmospheric and Ocean CO2 Measurements in the South Indian Ocean Made by Two Uncrewed Surface Vehicles in 2022 and 2023» by Chambers et al. submitted for discussion to ESSD Biogeochemistry

General comment :

This manuscript describes data of atmospheric and oceanic CO₂ recorded by two uncrewed surface vehicles (Saildrones) deployed in the southern ocean in 2022 and 2023. The Saildrones were equipped with the ASVCO2 instrument package that allows high quality measurement of CO₂ with regular calibration. The so-called “chemistry data” (which are focusing on CO2 measurements) are archived at NOAA -NCEI following the data format that has been used for other deployments of the ASVCO2 instrument. The manuscript also describes a so-called “Full mission data” data set which is archived on the University of Southern Florida data repository. This data set merges all measurements made by the different instruments on the saildrones (including the ADCP). It also includes col-located wind estimates from satellites data products. Finally, it also includes a flag that informs on the presence of eddies along the track of the saildrone. This full mission dataset is available with a one-minute sampling resolution and hourly averaged time resolution.

In my opinion, this manuscript is in line with ESSD editorial policy. In addition to the description of the data, it gives some useful information on the history and the decisions taken during the deployments which are important for potential users of the data. The authors also give a preliminary analysis of the pCO₂ data that are important to fully understand the potential of the data.

This manuscript is a detailed and useful description of a significant and original dataset. The manuscript is well written. The figures and tables are of good quality but could be improved. I would be glad to support the publication of this manuscript after some revisions (which I believe to be minor). My concerns are detailed in the following sections.

Specific comments :

L177-178: Data from the DuraFET pH sensor were not reported in the dataset (all values for pH are -999 in the files) because the sensor was uncalibrated. However, the manuscript mentions that this data was used for the flagging of CO2 measurements. To my knowledge the procedure to assess quality flags for CO2 measurements from the pH data are not described in Sutton et al. 2024 and Sabine et al. 2020. Some information needs to be given on how the pH values were used to do quality checking.

L210 : Please give a reference to access the “ Mesoscale Eddy Trajectory Atlas”

L216 to 227: The methodology used to assess if the Saildrone is outside or inside an eddy is described in this section. I believe that it would be useful to the reader to also indicate that the variable EDDY_DIRECTION has only three values : 1 for Anticyclonic Eddy, 0 for no Eddy and -1 for Cyclonic Eddy.

L260-265: The authors mention a bias between the pCO2 values from the saildrone and the mean monthly climatology. It should be noticed that the mean monthly value of this climatology is based on the mean for a given month of the values spanning from 1988 to 2020 whereas the measured values are from 2022-2023. The temporal increase of the atmospheric CO2 fraction could (at least partially explain this bias).

L251 : I would suggest renaming this section. The current title for this section “Analysis of the Chemistry data” is maybe confusing. I believe that the authors only make a preliminary illustration of the potential of the data rather than a real analysis of the data.

Figure 1 : This figure could certainly be improved. It is sometimes difficult to distinguish the trajectory of the saildrone due to the choice of the colors. The legend of the figure should also mention white more detail what type of data have been used for the standard deviation of the surface height variability.

Figure 2 and figure 3 : The color of the dots used to describe the center of the eddies is easy to distinguish. Maybe, it is not useful to give this information to these figures.

Technical corrections

L46 : For nitrate concentrations, the “3” of NO3 should be in subscript rather than superscript. Please note also that nitrate is an ion and should be noted NO3-.

L48 and L70 : The “2” of pCO2 should be subscript

L141 : “Ocean color” should not be used to describe the Chla-A concentration estimated by fluorescence.

L249 : Replace “/” by “.”