The manuscript presents a detailed description and validation of an ocean reanalysis, and it is very appropriate to accompany the dataset itself. The paper is well written, and explains very honestly the strengths and weaknesses of such dataset. I enjoyed reading it. As such, I recommend the paper for publication, but I ask the authors to better discuss, illustrate and explain several issues, summarized below. I think having a dedicated section "Discussion" before the Conclusions is a good option.

General points (for which I suggest including a more detailed discussion).

- Neglecting sea-ice modeling in the reanalysis seems to me a weak point, considering the effects of sea-ice induced circulation around e.g. ACC, which is certainly an area of interest for the authors. A deeper discussion about what they expect to miss, and if there are plans for including a sea-ice model will benefit the paper.

- The validation mostly focus on observation diagnostics. In my opinion a reanalysis is unique in the sense that can capture integrated diagnostics (OHC trends, transports, etc.) which are to some extent unobserved. It seems from the presentation of the work that climate applications are not the focus at all of this dataset. Also, the heterogeneous assimilated data (many datasets switching from delayed to real-time mode) may also compromise the low-frequency variability. Any thoughts about this, to include in the Summary/Discussion?

- The DA system is detailed in a companion paper, not available at the moment (in review for Ocean Modeling). Because of that, some aspects of the DA system are probably not explained, so the reader may have troubles to understand the formulation, if there is some lag between the two publications (see also below specifically).

Minor points

Line 4: it is said up to 2019. It will be useful for the interested readership to specify the plans for update (near real time, once in a while, or never?).

Line 7; "for some variables" please be specific in the abstract.

The spinup period seems short (3 months). Do you have evidence that is enough for stabilizing most low-frequency variability indexes, or was chosen more for practical reasons?

Line 72 vertical resolution seems quite low compared to most other reanalyses. It is also not clear (Line 142) how can the system assimilate both night and day time SST data, since the diurnal cycle will be for sure underestimated. Or maybe I am missing something in the explanation (lines 141-142 are not very clear to me).

Line 74: Forcing fields masked if sea-ice. Not clear how do you use them if sea-ice occurs? Which fluxes would you use instead?

Line 108: it is clear from the text below that the ensemble anomalies in the EnOI are "climatological" (i.e. not flow-dependent). Better to specify here for clarity.

The multiscale Data assimilation formulation seems sub-optimal: no proper scale separation is used, and the use of the same observational data between the two steps implies non-zero cross-covariances between observations and background (in the second step). This seems theoretically sub-optimal and should be mentioned. Another issue is that the time dimension does not change between the two steps. It would be reasonable to assume that longer (broad scale) dynamics is associated to longer time scales, while here the 3-day time scale applies to both. Any thought about this? Again, maybe this is included in the manuscript submitted to Ocean Model., but without it being published it is worth to discuss.

MDT (lines 155-157): I understand this is computed as Mean SSH from a free-running-model run. This means that long-term mean barotropic-dominated transports (e.g. in ACC) in the reanalysis should look very similar to the control experiment, by construction. Some comments about this will be beneficial, as most other reanalyses use other strategies (either an "observed MDT" or one with assimilation of in-situ data)

Table 1. TEM vs TEM2 and SAL vs SAL2: the difference is not explained in the text.

Line 180: Doesn't sound better/Isn't more common to say "super-obbing" with double "b"?

Table 2,3,4. I really like the efforts in quantifying the error growth. Perhaps reporting all those values also in the figure 7,8 (in the profile panel) will help to better see the error growth and the differences with BRAN2016

Also, better to say immediately which observations are used to validate and form the statistics: are those assimilated in BRAN2016 only or also those supplemented in BRAN2020 (like sea mammals, etc.)? Both are possible choices in my opinion, but the interpretation of the results will be different.

Linked to this: lines 265-269 are not very clear to me. I don't understand why differences in skill scores improvements between surface and sub-surface data should lead to the authors' conclusions? It is because of the same atmospheric forcing/ingested surface data? Better to state clearly.

Line 323 typo in "reanalysed"

Line 331: SPINUP-EI is a misleading name. Perhaps change to CTRL or similar

Review of “Next generation of Bluelink ocean reanalysis with multiscale data assimilation: BRAN2020” manuscript.

The Next generation of Bluelink ocean reanalysis with multiscale data assimilation: BRAN2020” manuscript presents a new attempt to better estimate ocean dynamics in the multi-decadal global ocean arena. The structure of the manuscript is well organized and scientific ideas are correctly exposed.

General comment:

In the manuscript, the authors applied the EnKF-C method to propagate observations information inside the ocean model. In my opinion, omitting sea ice in the system is a major problem, it is an important phenomenon influencing dynamics in the Southern Ocean. Using analysis every 3 days seems too frequent for the global ocean model having a spatial resolution of 1/10 degree. For example, GLORY NEMO experiment is using a 14-days assimilation window. Some discussion along the lines would be necessary; how authors decided for 3-days as an appropriate assimilation window. Seems to me that it’s quite short (or is quite often – every 3 days) time span between analysis, and is a way of imposing stiff control over the ocean system i.e. suppressing model physics to fully develop. Initialization of the temperature fields by using the daily averaged values seems strange, the model vertical resolution in the surface layers is 5m which is prohibiting diurnal oscillations to fully develop. In that sense, the model is not resolving diurnal SST dynamics (not sure about the temporal frequency of atmosphere forcing), and this shouldn’t be the reason for using daily fields (which are dynamically unbalanced). As assimilation is done on the two different scales, could the time step for applying analysis be different (longer for large scale and shorter for mesoscale)?

Specific comments:

Line 54: Are authors referring to common term residuals of data assimilation when they talk about the difference between the analysis and observation?

Line 180: Not sure if this is a typo mistake: “ super-obing”. It sounds a bit strange, usually, we refer to “super-obs” or “super-observations”.

Line 197: Analysis innovations are sensitive to the observation errors, and in that sense are the observation errors constant in space/time or they are varying (for specific observation type)? If not do authors think it would improve the assimilation system?

The manuscript is well written and presents relevant resutls and discussions. Thank you for dedicating time to publish a detailed discussion and analysis of this dataset. However, a more detalied discussion on a few topics would greatly benefit the manuscript.

Main comments:

How does the multiscale data assimilation implemented here compare to previous work, such as Li et al (2015) and Tissier et al (2019)? I understand more details should be available in Chamberlain et al (2021a). But since this is not yet publish, it would be good to have some more details in the present manuscript.

The model has the Southern Ocean as one of its regions of interest. The absence of a coupled ice model can have significant impacts on the circulation and water column structure in this part of the world, with potential repercussions to the global deep ocean. Could you please explain the impacts on the model results and how potential problems are minimized?

I am a bit confused about the definition of innovation. To my best knowledge it is defined as the difference between the observations and the model maped to the obs locations. However, all the values presented in Fig. 2 are positive. Does this mean there are model BIASES for all the fields? Or are the innovation and increment defined in a different way?

A few minor comments:

Lines 64-65. It looks there is a typo. Please review it.

Line 97. Typo: “of” should be “on”.

Lines 190-191: Comparing the daily average fields to observations excludes a lot of high frequency processes that will be smoothed out. I imagine the observations were also averaged for the comparisons.

Line 215. Could you please explaing how the multiscale assimilation eliminated the biases?

Line 216. How to avoid the over-fitting in the EnOI scheme?

General comment: The technical contents and descriptions are good enough to understand about BRAN2020 reanalysis system. Main conclusion of the paper is about the contribution of multi-scale data assimilation (DA) approach to resolve mesoscale features of ocean conditions. Overall quality of the paper is fairly acceptable. But acceptance decision can be made once a few requests are followed up and discussed.

Technical comments:

1. BRAN2020 combines both coarse and fine resolutions in its multi-scale DA approach. One of the main issues is that sea ice model is not included. This naturally leads to what is the benefit of increasing ocean model resolution without considering coupled sea ice ocean modeling system. Concern is that the absence of realistic sea ice condition may deteriorate the analysis result in high latitude areas. Not enough information is provided about the issue along the high latitude areas.

2. As relatively fine scale ocean model is used in the multi-scale approach, another natural question is about its benefit on circulation dynamics. Climate index comparisons are described but questions still remain about dynamics: currents, transports, etc.

3. Super-observation scheme is used in the reanalysis run from 1993 to 2019. It will be great  if authors can provide information about overall computational cost and quantification of observation data quality improvement of the super-obbing in the BRAN2020 system.

4. In page 10 (line #217), authors used the term "we think" to talk about observation error specification issue. They consider that better analysis result might be obtained if larger error is used for avhrr sst data. To make a conclusive opinion, they have to provide a direct evidence. A small set of analysis experiment might be possible. Without such a direct evidence, it will end up to a simple guessing.

5. The study claims that multi-scale DA approach is beneficial even for non-argo time period (especially before 2000). Most of the comparisons of the study is based on data sets applied in the DA system. OISST and climate index comparisons are provided but additional comparison can be made against another third party reanalysis products.

6. Additional minor editorial comments can be provided once enough technical discussion and feedback is provided.