This manuscript presents an innovative methodology and a timely study, leveraging the capabilities of the Tara schooner to monitor CO₂ fugacity (fCO₂) in a key region for the global carbon cycle. The work demonstrates the potential of such expeditions for regional carbon dynamics studies and contributes valuable observations to the limited dataset from this area. Moreover, this represents an excellent initiative to expand the scope of the Tara schooner's activities—traditionally centered on marine biology and ecology—toward geoscience observations, as exemplified by the dataset presented in this manuscript. The manuscript is well-written, well-structured, and includes high-quality figures. The article is worthy of publication and appropriate to support the publication of a data set, but the dataset itself exhibits significant shortcomings that must be first addressed. Important modifications are required to ensure the dataset meets the standards of openness, metadata completeness, and long-term usability. I recommend a major revision, primarily to address issues with the data product, as outlined below. Additionally, there are inconsistencies in the framing and presentation that detract from the overall clarity and impact of the work, but they should be easily ameliorated.

The current form of the data product significantly weakens the manuscript's suitability for publication in ESSD, a journal known for its rigorous standards in data quality and accessibility. The dataset lacks metadata, the data file itself it has an untraceable naming (“CO2Tara.xlsx” -sic-), fails to adhere to standardized variable names, and does not appropriately specify units. The dataset header lacks key information, such as details about the analytical methods used and the complete list of contributors. It is true that they are included in the Zenodo version, but the file should stand by itself. Additionally, quality control flags for variables should be assigned to enhance the dataset's reliability and usability. Metadata completeness is crucial for long-term usability and re-usability. The file format should be open source (rather than an Excel file) and it would be much better distributed as a NetCDF file as it is interoperable with other platforms. The use of proprietary or non-open-source formats is a critical limitation, so a conversion to a format compatible with community standards is necessary. Since the authors intend to submit this dataset to SOCAT, which I consider an excellent decision as it is the current reference database for fCO2 measurements, it would be beneficial if the data product presented in ESSD could offer some added value or differentiation.

Specific comments (section and/or lines):

Reframe the title to reflect the true geographic scope, as the analysis is primarily focused on the Amazon River area rather than the entire east coast of South America.

Abstract 26-29: the order of the description should be homogeneous. For example, from river to ocean.

Abstract: North Brazil Current and Brazil Current are concepts not explained in the abstract. Not easy to follow.

Intro. 40. agree with These regions present much higher temporal and spatial variability. Therefore, this highly valuable snapshot only informs about spatial variability of a single season.

Intro 45. true sentence. The low number of observations in coastal waters is somewhat unexpected. Considering that coastal zones are, by definition, more accessible than the open ocean and offshore areas, it raises the question: why are these regions underrepresented in observational datasets? While addressing this issue is beyond the scope of this study, including a brief discussion or hypothesis to guide the reader would add valuable context.

Figure 1 is highly effective and well-designed, offering a clear overview of the study area. However, the inclusion of Pacific data skews the color bar and detracts from the manuscript's focus on equatorial South America in the Atlantic Ocean. It would be more appropriate to exclude the Pacific data and revise the figure to better align with the study's regional scope.

1. It represents one of the greatest environmental gradients on “the interface between” land and ocean in the world.

64 as the “Amazon” rainforest sequesters…

1. The references to the ANACONDAS (Mu et al., 2021) and Camadas Finas III (Araujo et al., 2017) campaigns lack both date signatures and spatial context. As a result, these terms may be largely unfamiliar to readers who are not specialized in regional studies, potentially hindering the manuscript's accessibility. Providing additional information about the campaigns' timing and geographical scope would help contextualize these references.
2. The manuscript mentions the link between these two systems, but it is unclear how this connection is made, as the schooner does not sail as far as Óbidos. If the link is indeed established, further clarification is needed, as it is not apparent from the current description. Could you please clarify whether the link is made, or revise the statement to reflect the actual coverage of the study?

71 I would delete “extensively”

75 in advance: The statement in paragraph 75 claiming that the Argo program could address the scarcity of fCO2 measurements in the medium term is not accurate. Currently, this is technically unfeasible, as the Argo program is designed primarily for interior ocean monitoring. To measure fCO2, a surface-intensified approach would be required. The entire paragraph should be reassessed and revised to reflect the current limitations and the specific needs for fCO2 measurements.

1. As this dataset is based on underway measurements, you use kilometers to express the magnitude of the data, which is a valid option. However, it would be helpful to include additional information regarding the timing of the observations. For example, how many days of data were collected? Furthermore, how many different biomes were crossed during the survey? Providing this contextual information would enhance the understanding of the dataset's temporal and spatial coverage.

108-114. The port-to-port description could be better represented in a table format. This information, while useful, does not add significant value to the narrative and would be more concise and accessible in a tabular form.

In line 103, you mention 14,000 km, which is a considerable distance. However, the schooner sailed a total of 70,000 km. This discrepancy raises the question: why are the remaining 56,000 km not included in the analysis? While there may be valid reasons for excluding these data, offering an explanation would strength the manuscript's transparency.

Figure 2 does not accurately represent the full circuit, as it omits two branches. Please revise the figure to include these missing branches.

However, the dataset lacks integration with existing efforts (e.g., SOCAT) and does not demonstrate sufficient added value over what is already available. It would strengthen the manuscript to articulate why this dataset is unique and necessary in the context of global fCO₂ monitoring.

Section 2.2: You mention an important flow rate, which is often a bottleneck in underway systems on unconventional vessels. Could you please provide the model and specifications of the pump used? This information would be useful for understanding the system’s limitations. Does the pump include a filter? Additionally, what is the maximum speed of the schooner at which the pump remains functional? Is the schooner's speed included in the dataset? If not, it should be, as this could be an important variable to consider.

Line 132: The term "accurately" could be removed.

Line 138: It is unclear to me why you do not have this data. In line 140, you mention a temperature difference, are you not talking about that in line 138?

Line 150: The sentences need to be reordered for clarity. It would be more effective to first describe the atmospheric air, then the reference gases, and finally the seawater. This would improve the logical flow of the section.

Line 160: When stating that the system is "cleaned regularly," it would be helpful to include the periodicity of the cleaning process. Additionally, does the intake circuit feature any physical barriers or filters to prevent the introduction of large particles? This should be clarified for completeness.

Line 185: You mention "yellow" in Figure 3, but no yellow is visible (at least to me) in the figure. Please revise the description to match the actual content of the figure.

Figure 3: The straight line in the "Raw" data does not convey meaningful information and appears to be an artifact caused by the connection of data points in the time series. This should be corrected to ensure the figure accurately represents the data.

Figure 4: Please include a scatter subplot that illustrates the 1:1 relationship between the discrete samples and the underway measurements. Use the timing of the bottle closure from the CTD surface samples, applying a time window of 2-3 minutes for the underway. Also, include the uncertainty associated with the discrete samples (5.7 µatm) and the range of underway measurements. The current format of this Figure 4, which combines all values into one plot, makes it difficult to discern patterns due to the large range of values. It would be clearer if the figure were split into three subplots: one for high values, one for the central values around 400, and another for low values.

Figure 5: While I appreciated the broad context provided by Figure 1, I would recommend a closer zoom for Figure 5 to focus specifically on the underway track. The current version leaves a significant amount of blank space that could be better utilized. The figure should be revised to eliminate this excess space and highlight the relevant data more clearly. Additionally, the arrangement of the subplots is unclear, and the "pyramid" layout may not be the most effective. I suggest reconsidering the layout for better clarity and visual coherence before publication.

1. Overview section: Ensure that the variables are always presented in the same order: temperature, salinity, and then fCO2.

Figure 7: Please add the locations of Obidos, Belem, and Macapa to the figure. Additionally, label some isobaths of bathymetry for reference.

Figures 8 & 9: It would be more efficient to include the area shown in Table 1 directly in the figures, thereby eliminating the need for the Table 1 itself. In my opinion, this will enhance the visual presentation and reduce redundancy.

Bathymetry: The bathymetry data is sourced from ETOPO2v2 and is colocalized along the ship track. However, including bathymetry values at 0 m or depths of 1-2 m raises some questions, as this would imply the ship was at or near the seabed (which, of course, we hope did not occur!). It may be more appropriate to group these shallow measurements or explore alternative methods for representing this data in a way that better reflects the actual conditions. Another comment about bathymetry: I would strongly recommend adding the bathymetry data to the dataset. This would enable others to fully reproduce the analyses presented in the manuscript.

The manuscript presents CO₂ fugacity data measured during the Tara expedition in the Tropical and Southwest Subtropical Atlantic Ocean, with a focus on the Amazonas River and the Brazilian coast. However, the manuscript contains several inconsistencies between figures and text, along with errors in names and terminology, which need to be corrected throughout. Key regional references are lacking, and the authors show limited understanding of the area. Additionally, the authors rely on outdated literature, some over 30 years old, which may no longer reflect the current local status and climatology.

Although the dataset covers the Caribbean to Argentina, the authors primarily focus the results and discussion on the Amazonas River plume and the North Brazil Current in the Equatorial Atlantic, with minimal discussion of the rest of the Southwest Atlantic Ocean. This narrow focus, combined with the lack of relevant references, leads to an over-interpretation of some parts of dataset and findings. Many of the findings and conclusions have already been published by other researchers in the same region, but this is not addressed clearly. Furthermore, there are several unsupported assumptions throughout the manuscript that need to be addressed.

Regarding data quality, the ESSD aims to ensure that high-quality and reproducible research data sets are included in permanent repositories. The authors mention issues with standard (STD) gases used for calibration but have not provided key data (e.g., pCO2air for STD calibration, in situ DIC and alkalinity data, or cruise metadata) for review. A major concern is the limited calibration with only two gases, one of which (502.3 ppm) is much lower than the values found in the river-ocean continuum (~2000 uatm), which falls outside the range, recommended in the SOCAT CookBook (2018). This raises doubts about the uncertainty of the data beyond the STD calibration range. Additionally, the authors report using only 13 discrete samples to calibrate measurements over 14,000 km, which is statistically insufficient. This raises concerns, especially since only six in situ samples were taken during the longest part of the cruise. No statistical metrics are provided to support the calibration's accuracy and validation.

The authors claim there are no comparable datasets, focusing solely on the SOCAT database which excludes river measurements. I suggest they explore additional references to find relevant studies that might help:

• For the River-ocean continuum see: Valerio et al., 2018 (doi: https://doi.org/10.1364/OE.26.00A657) and the pre-print of Less et al., 2018 (https://doi.org/10.5194/bg-2018-465) - although the manuscript was withdrawn there is valuable information that authors can use.
• For Amazonas River Plume and North Brasil Current see: Monteiro et al., 2022 (doi: https://doi.org/10.1029/2022GB007385), Valério et al., 2021 (doi: https://doi.org/10.1016/j.csr.2021.104348), and others that authors should consult, for example, Ibánhez et al., 2016; Lefèvre et al., 2014, 2017 and 2020.
• Other references about CO₂ fluxes, ocean pCO₂ and fCO₂ in the Brazilian coast can be found in the review paper Oliveira et al., 2022 (doi: http://dx.doi.org/10.21577/0100-4042.20170970).
• For the Vitória-Trindade Seamounts: Dynamics see: Napolitano et al., 2020 (doi: https://doi.org/10.1029/2020JC016731) and Silveira et al., 2020 (doi: https://doi.org/10.1007/978-3-030-53222-2_2). pCO2 data: request to Marinha do Brasil (Brazilian Navy).

In conclusion, while this manuscript has the potential to contribute to understanding CO₂ fluxes in the Tropical and Subtropical South Atlantic, the issues with data quality, limited analysis, and insufficient calibration prevent it from being ready for publication. I recommend a major revision, changing the manuscript focus specifically on the Amazonas River Plume and North Brazil Current. Due to the concerns raised, I do not recommend submitting the whole dataset to SOCAT v.2025, especially without addressing these issues.

Please, find attached the specific comments.