Review of “Advances in Surface Water and Ocean Topography for Fine-Scale Eddy Identification from Altimeter Sea Surface Height Merging Maps” by Zhang et al submitted to Ocean Science

The authors propose a method for validating altimetry-based maps through an eddy identification framework, focusing on the South China Sea and on fine scale eddies. The approach is applied to Level-3 SWOT products, operational gridded products distributed by the Copernicus Marine Service (CMEMS), and a regional 2DVAR product specifically designed for the region. Their findings indicate that 2DVAR regional mapping identifies significantly more accurate fine-scale and mesoscale eddies compared to CMEMS products.

The scientific questions posed by the authors are relevant to the ocean science community: how can gridded altimetry products be validated using independent datasets and metrics? How to assess quality of gridded on fine scale eddies/dynamic?

While I really like the approach proposed by the author in using SWOT as an independent measure for the validation of the gridded fields, I however have the feeling that the manuscript requires substantial revision before it can be considered for publication in Ocean Science. Specifically, the Discussion section needs significant improvement to better contextualize and interpret the findings. Additionally, the current results are, from my point of view, insufficient to fully support the interpretations and conclusions presented, and I would encourage you to conduct further analyses to strengthen the study. I provide hereafter detailed comments and recommendations.

The experimental design also requires a more thorough description, and the title of the manuscript should be reconsidered to better reflect the specific focus of the study, which is conducted solely over the South China Sea.

While the manuscript is relatively well-structured, several sections require clarification, and I provide below suggestions for improvement in terms of quality and clarity.

Material and method section

To enhance the Materials and Methods section, I would recommend adding several clarifications. First, it would be beneficial to include specific references, such as a DOI, for the gridded sea surface height datasets used in the study. While the datasets are listed in the references, explicitly citing them in this section would improve clarity and accessibility for readers. Additionally, please specify the time period used, as this information is essential for contextualizing the scope of the analysis. I was a bit lost when I read/interpreted the results since I was not sure which CMEMS (DT, NRT) dataset was used. Furthermore, a more detailed explanation of the differences between Near Real-Time (NRT) and Delayed Time (DT) products you are using would be helpful. Clarifying which processing mode (NRT, DT, or both) was used for the 2DVAR mapping would also improve transparency in your methodology and intercomparison.

Eddy identification section:

Since most of the results depend on the eddy identification algorithm, it would be helpful to provide additional details about the methodology employed in your study. Specifically, could you elaborate on the method proposed by Ni (2014)? It seems challenging to find this reference—would you be able to provide a DOI or further citation details? Additionally, I would appreciate it if you could clarify the specific time period during which the eddy identification and tracking were conducted.

Eddy validation section:

Since the eddy detection is specifically applied to the SWOT swath, it might be beneficial to undertake further validation to ensure its robustness. Conducting an intercomparison with independent datasets, such as SST or chlorophyll data, could provide valuable insights and help validate the eddy detection within the SWOT track. Additionally, the eddy detection performed along the SWOT track is inherently constrained by the SWOT swath extension, which is limited to 120 km. So, the eddy contours identified within the SWOT track are, by design, confined to this 120 km width. This raises an important question: what if the detected eddy contour corresponds to an isocontour of a larger eddy observed in broader maps (let’s say the upper part of a large eddy)? In other words, how reliable is the eddy detection within the SWOT track? Is the SWOT eddy contour really a true eddy boundary? Complementary investigations and discussions on this aspect should be carried out to further validate the eddy detection in SWOT products.

Results section:

It would be nice to have a figure with the difference of ADT between SWOT and the AVISO & 2DVAR products to illustrate the amplitude difference between observed SWOT measurement and reconstructed NRT & DT AVISO and 2DVAR products.

Since the results rely on eddy matching between the SWOT track and the altimetry gridded product, it might be helpful to include an illustration of eddy detection for both the SWOT track and the altimetry gridded products, as shown in Figure 3. Or superimpose contour of detected eddies between SWOT products and altimetry gridded products

Additionally, incorporating classical statistical metrics (in addition to eddy detection metrics) to evaluate the gridded product at a finer scale would enhance the analysis. For instance, presenting results based on RMSE comparisons between the gridded product and L3 SWOT could provide valuable insight on the error level in the products. This would also help to illustrate and intercompare the performance between the AVISO maps and the 2DVAR maps in the region of interest.

In the abstract you mentioned that: “The findings indicate that SWOT provides an enhanced capability in resolving fine-scale and mesoscale eddies in the South China Sea compared with conventional in-situ data, such as drifting buoys.” Could you clarify which result supports this finding?

Discussion section:

To improve the Discussion section, it would be valuable to address several key points.

- Could you elaborate on the limitations of AVISO maps in terms of their resolution within the region of interest? Additionally, it would be helpful to discuss the differences between Near Real-Time (NRT) and Delayed Time (DT) products, as well as their implications for the 2DVAR approach.

- Could you explore potential reasons why only a limited number of eddies are detected in AVISO maps? This analysis would provide an important context for understanding the performance of different methodologies. Emphasizing the significance of developing regional configurations for merged sea surface height (SSH) products would also be a valuable addition, highlighting the potential for more accurate regional analyses.

- It would also be useful to consider the broader implications of your study. Are the conclusions drawn from your research applicable on a global scale? Could the proposed method be effectively applied at a global level? Do the findings regarding AVISO maps and the 2DVAR approach remain consistent at a global scale? Addressing these questions would provide a more comprehensive perspective. I would suggest for example performing similar analysis at global scale using CMEMS NRT & DT product and L3 SWOT data to illustrate the limitation of NRT products and test the method at global scale.

- In Figure 6, the eddies appear to remain relatively static over the 2–3 months of tracking analysis. It would be helpful if the authors could discuss whether these eddy structures might also be detectable or sampled by classical nadir altimeters and elaborate on the added value that SWOT brings compared to traditional nadir altimeters. What are (on average) observable length scale of nadir altimeter in this region? Maybe an illustration of all nadir track over specific region may highlight the difference or similarities between swot 2d product and nadir 1d products.

- Including a section that describes the importance of the South China Sea (SCS) could also further strengthen the motivation for this study