General Comments

Marine carbon data that are of high and known quality, have appropriate metadata and are FAIR are vital in assessing the role of the ocean in the changing carbon cycle.  The SNAPO-CO2- v2 alkalinity and DIC dataset described in this manuscript is an important contribution as the data have wide spatio and temporal coverage and are of very high quality.  The v2 described in this manuscript builds on the initial dataset published earlier this year.  The manuscript details the data sources, the QC processing, and then the data assemblage and synthesis.  Five use-cases are then presented, with the spatial distribution and trend analysis of the five cases assessed.

The data are available via the SEANOE website, and via the doi number presented in the manuscript.

The paper is very well written, the tables and diagrams are clear, except where noted below, and the supplementary material is useful and suitable. 

This is a very good manuscript, I'm sure the dataset will be well used by the marine carbon community and that the manuscript will be well cited.  I recommend publication and have only a few minor comments outlined below.

Specific Comments

Line 433        missing word:  “…precision was based on triplicate analyses was estimated.…”

Figure 7, shows the surface distribution of alkalinity and DIC in the Mediterranean Sea over the decade 2014 – 2023.  Given that the trend in DIC over that decade was 0.72 umol kg-1 yr-1 , the changes in the graphed value could be up to 7.2 umol kg-1 for the decade.  I suggest either normalizing the data to a particular year, or to describe the effect in the text.

Several of the Case studies in Section 5 involve time series analysis, however the methodology for doing this is not described or referenced.  Sutton et al (2022) is a useful reference for this:

Sutton, A.J., Battisti, R., Carter, B., Evans, W., Newton, J., Alin, S., Bates, N.R., Cai, W.-J., Currie, K., Feely, R.A., Sabine, C., Tanhua, T., Tilbrook, B., Wanninkhof, R. (2022) Advancing best practices for assessing trends of ocean acidification time series. Frontiers in Marine Science, 9: 1045667. doi: 10.3389/fmars.2022.1045667

Line 930        This dataset is indeed a useful complement to other data compilations such as GLODAP and SOCAT.  Are these SNAPO-CO2 data included in  SOCAT and / or GLODAP (as appropriate)?

Figure 14.  The colour scale is not necessary, and is not used in the Figure data.  There are only two time periods identified by the colours, and these are indicated in the caption.

Fig S3 The caption is incorrect, the colour coding in the Figure is not as described in the caption.  This needs to be corrected.

Fig S4 The concept of this Figure is good, and it relates well to the text (line 339), however the map insert is difficult to read, and it is difficult to relate the colour coding of the triangle symbols in the vertical profiles with the day scale on the map. This should be clarified, -  if the time component is important the colours or symbols should be the same, otherwise the day scale is not necessary.

The manuscript describes a data set of about 67000 observations of ocean dissolved inorganic carbon and/or total alkalinity around the world. The data set is mostly based on observations by the French vessels and scientists, analysed in a lab in France.

This is a valuable compilation of data in a single format and with a coherent quality control. The manuscript is well written and perceived. Quality controlled data in a consistent format of ocean carbon variables are valuable, so the manuscript deserves to be published.

The ms refers to both the GLODAP and SOCAT data products. There is some overlap, but also differences. In particular for GLODAP there is a potential overlap since that is also dealing with interior ocean DIC and TA data. It would be good if the ms could state how large a fraction of the SNAPO data are already in GLODAP, and an estimate on how large a fraction will be submitted to GLODAP for future versions.

Very often when using ocean carbon data, there is a need and s strong correlation to other variables. I can imagine that often (but probably far from always) other variables were being measured during these campaigns, fgor instance, oxygen and nutrients (variables often needed to calculate the anthropogenic component of the DIC). However, by looking at the individual data sets for those cruises where that is available, I was in most cases not able to locate the other variables, or even find a list of other variables that could be available. I realize that amassing other variables as well, just as is done in GLODAP or SPOTS, for instance, is probably outside the scope of this work. However, it would be useful to have that information about additional variable available in a concise format, for instance in tables S1. At least for the “most important auxiliary variables”, possibly guided by variables available in GLODAP.