Comment on paper ESSD-2025-212

General comment

This is one of many papers that researchers clustered around BEC submit to journals and in particular to ESSD. This is a list (for sure not complete one)

• Improved BEC SMOS Arctic Sea Surface Salinity product v3.1 - https://doi.org/10.5194/essd-14-307-2022
• First SMOS Sea Surface Salinity dedicated products over the Baltic Sea - https://doi.org/10.5194/essd-14-2343-2022
• New SMOS SSS maps in the framework of the Earth Observation data For Science and Innovation in the Black Sea - https://doi.org/10.5194/essd-2021-364
• Nine years of SMOS Sea Surface Salinity global maps at the Barcelona Expert Center - https://doi.org/10.5194/essd-13-857-2021

It is normal that ‘general’ algorithms need to be adapted to the particular marine environments, but the dispersion of applications in different papers is not useful to the researchers interested in using SMOS products for several small seas and analyze the differences and possibly make further adjustments. The potential proliferation of articles for each single basin risks being of little use to research.

At the extreme north and south of the global ocean, important common problems are (1)Low sensitivity of brightness temperatures (TB) to salinity in cold waters, (2) Land–sea contamination (LSC) and ice–sea contamination (ISC), (3) Lack of sufficient in situ measurements. It would have been appropriate to discuss in a single publication the different peculiarities of the waters in these environments and what the appropriate solutions are. Same concept for the Baltic and Black Seas.

It is understood that in practice articles are written on the basis of collaborations that are built over time and therefore the practice of the authors is acceptable, but each article should highlight and point out the differences much more accurately. In particular, I would have expected a discussion after line 52 of the introduction on what has been done in the Arctic Sea and what needs to be done differently in the Southern Ocean.

In the remaining part of the article, the work for the geophysical corrections and for the analyses of the various SST products is appreciable. The validation and intercomparison are also very appreciable.

The paper itself is well written and practically complete and can be published after minimal corrections.

Specific comments

Introduction lines 14-15 - understanding of processes influenced by upper-ocean salinity, including ice formation – explain how it is possible to study the formation when SMOS resolution is large enough and the distance from the ice edge is more than 150 km, while within 150km there is high noise?

Figures 5, 7, 18 show very noisy areas near the continent. What does this mean for studies on the potential applications of the product: understanding of processes influenced by upper-ocean salinity, including ice formation and melt, the reduction of Antarctic sea ice extent, and the opening of offshore polynyas. It is not necessary to go into depth on the oceanographic issues, but to give a general indication of the limits of the product for the listed applications.

General comment

This manuscript describes a new sea surface salinity (SSS) product for the Southern Ocean derived from SMOS observations, incorporating algorithmic improvements aimed at enhancing retrieval accuracy near sea ice margins. The dataset spans over a decade, and the validation suggests reasonable agreement with independent in situ observations in open waters. The topic is relevant to the ESSD readership, and the dataset could be a useful resource for studies related to polar oceanography and sea-ice interactions.

There are several important aspects that require further clarification and improvement. First, the manuscript does not sufficiently distinguish the presented dataset from previously published SSS products, including those already available in ESSD. A more detailed comparison is needed to highlight what has been improved in terms of retrieval methodology, spatial and temporal resolution, coverage, or accuracy. It would be helpful to clarify which known limitations in prior products this dataset addresses.

The description of the data processing chain, particularly the modified retrieval algorithms, should be expanded. Although some methods have been previously published, ESSD readers should be able to understand the core processing logic and improvements without needing to refer to external literature. Key steps—especially those tailored to address the challenges of SSS retrieval in cold, ice-influenced waters—should be described in sufficient detail to ensure transparency and reproducibility.

The link provided for accessing the dataset is currently not functioning. The authors should ensure that the data are fully accessible via a persistent and reliable repository, in compliance with ESSD’s data availability requirements. The dataset should be accompanied by appropriate documentation and metadata.

In addition, the manuscript references other datasets used in the validation and development process, but no access links or citation details are provided. All datasets mentioned or used should be properly cited and made accessible to ensure full transparency and reproducibility.

Specific comments

1. Multiple algorithm steps labeled as “Main algorithm changes” are mentioned in Figure 1, but the manuscript does not provide detailed explanations of these changes individually in the main text.
2. All figures in the manuscript lack sufficient clarity, making it difficult to interpret the visualized information. Additionally, some figures (e.g., Figures 5–7, 10-15, and 18-19) and Tables 1 and 2 do not include units, particularly in color bars and value columns. Please revise to improve figure resolution and ensure all numerical data are properly labeled with units.
3. In line 144, the term "land-sea-sky mask" is mentioned, but its definition and implementation are unclear. Later, Figure 7 introduces a "land-sea-ice mask", yet the relationship between these two masks is not explained. Please provide a clearer and more detailed description of how the land-sea-sky mask is constructed and clarify whether it differs from the land-sea-ice mask.
4. The manuscript briefly mentions the "DNB-lite" approach as the preferred method for generating the BEC SO SSS product, but it lacks a clear description of how it differs from the standard DNB and DNB-SST methods. Please consider expanding this explanation.
5. While the section 3.3.2 compares satellite SSS with both marine mammal and TSG data, it does not discuss potential sampling differences or biases between the two in situ data sources, which could affect the interpretation of near-coast and near-ice performance. Please consider addressing this aspect.