The paper presents a comprehensive description of the EMIC CLIMBER-X. This description is well balanced with a main text presenting the important principles and choices at the basis of model development as well as the model performance while the equations and parameters are given in the appendices. The model is designed as the successor of a very successful model, CLIMBER-2, which has been used extensively over the past 20 years. CLIMBER-X will very likely be as successful and this model description will thus be very useful for the scientists running the model and analyzing its results.

The paper is very clear and well written. It includes all the key elements needed for the description of a new model. I thus have mainly some minor suggestions of clarification. The only point that would deserve more discussion is the model tuning or calibration. Tuning is mentioned in the appendices and once in the main text for specific components as well as in the author’s contribution but without much details or a global view of the way tuning was achieved. I think it would be nice to have a specific section, for example as a section 2.5, to explain the tuning strategy. I guess some parameters have defaults, fixed values that are not supposed to be modified. Some parameters in the various modules were tuned specifically ‘offline’ for those modules (e.g., lines 281, 820, 852) while some others may have been tuned so that the full model fits with observations. It would be interesting to describe which are the main parameters that were tuned, what were the calibration targets for this general tuning (for example global mean temperature, total amount of precipitation, sea ice extent ,temperature trend over the past century, etc.) and the method followed for the tuning.

Minor points

Lines 7-8. I would be more specific in the abstract on the performance of the model, giving some examples where the agreement with observations is good and mention the limitations, such as the poor performance in the tropics.

Line 101. I would suggest to state at this stage that the two dimensions are latitude and longitude. If I follow well, all the vertical variations are then prognostic only. If it is the case, that would be good to mention it explicitly here. It is also probably worth insisting on this point for comparisons with observations including vertical variations (like figures 3, 6 and 8).

Sulphate aerosols are mentionned first line 159 but the spatial distribution is discussed line 166. This may give the feeling that this distribution is only valid for longwave.

Line 171. It would be useful to justify in one or two sentences why an approach even simpler that the one in CLIMBER- 2 has been retained.

Line 200. What is exactly meant by ‘lack of a Gulf Stream extension propagating to high latitudes’? Does it have an impact on model biases, on the location of deep-water formation or a link with the AMOC simulated by the model?

Line 250. Are the fluxes in the open ocean computed in the same way as in the ice-free fraction of the grid cells?

Line 254. Are the latent and sensible heat fluxes computed in the same way over the ocean? More generally, the computation of the surface balance is well explained here for SISIM in section 2.3 but not explicitly for the oceanic part in section 2.2.

Line 276. This would be useful to add here a few sentences describing the main characteristics of the model without the need to read Willeit and Ganopolski (2016).

Line 378-380. Is the deep water formed in Antarctica above the continental shelf or because of open ocean convection?

Caption of Figure 28. The reference to Paul et al. 2021 is not the same as on the figure (GLOMAP).

Paper provides detail description of new version of CLIMBER model and evaluation of model performance for different historical periods, as well as model response to the changes in atmospheric CO₂ concertation. Present day climate, climate change over resent historical period (1850-2100) and response to both doubling and transient increase of the atmospheric CO₂ concentration simulated by CLIMBER agree well with available observations and results obtained with more sophisticated climate modes. At the same time, there are noticeable differences in climate state for Last Glacial Maximum between CLIMBER and CMIP5 models.

This does not necessarily mean that CLIMBER is wrong, but in my view, additional validation would be useful. As said in the paper “The atmospheric component of CLIMBER X is based on a statistical-dynamical approach, which employs a number of significant simplifications and assumptions.” The fact that “these simplifications and a set of parameterizations explicitly derived from present–day climate limit the models’ applicability to climate states fundamentally different from the present one”. Authors mention Snowball Earth case as such climate state. Possibly Last Glacial Maximum climate is also different enough from preset one.

As indicated in the paper “CLIMBER-X includes code to diagnose the strength of the different climate feedbacks”.  I would suggest calculating climate feedbacks for LGM climate and comparing them with estimates available in literature.

Specific comments.

There seems to be discrepancy between Table 1 and Table 2.

Total evaporation in Table 2 are larger than observation, while latent heat in table 1 is smaller

On page 14 (near line 355).

Following sentence: “During the winter months and at high latitudes, CLIMBER-X also captures the near-surface temperature inversions.” It is better to say, “During the winter months, CLIMBER-X also captures the near-surface temperature inversions at high latitudes”

This paper serves as an introduction to the CLIMBER-X Earth system model. I found it to be extremely well written, and therefore pleasant to read. It describes in detail all of the key components of the model and how they interact, and then follows this with an evaluation of the model performance over the period 1850-present, which most readers will easily be able to relate to due to the authors’ efforts to compare their simulation with both observations and the CMIP simulations and assessments.

CLIMBER-X itself performs well in almost every aspect shown, and the authors also make clear the model’s deficiencies and limitations, which is crucial for potential users with applications in mind. As a paper intended to present a new model to the community, I think it serves that purpose well.

Minor comment:

I think it would be of interest to the reader to know what degree of calibration has been done, and how. There are a couple of comments in the text, but I think the ‘dark art’ of how we tune our numerical models should be made more visible (see e.g. Mauritsen et al., 2012 for a good example). In this case, where it is clear that CLIMBER-X is well suited to methods of performing robust exploration of parameter space (see e.g. Williamson et al., 2013), it would be interesting to know the extent to which this has been done, or remains to be done. I suggest including a section (or possibly a subsection in 2), that describes the calibration (tuning) that has been done to the model. For example, were components, or schemes within the individual models tuned in isolation, or was the entirety of CLIMBER-X calibrated as one? To what observations or metrics was the model tuned to?

Specific comments/corrections:

Line 77/78: Replace “16 CPUs” with “2 x 8 core CPUs”. Also, if your computation is on a single node then the detail “Infiniband FDR14 Lenovo/IBM” is redundant and can be removed.

Line 96: I suggest “We made extensive use of…” instead of “We extensively made use of…”

Line 162: Seem to jump from Appendix 5 (line 153) to Appendix 7. I suggest carefully checking and renumbering/reordering as required.

Line 233: question about rigid lid pressure?

Line 269: Typo: “occupyied” > “occupied”

Line 295: Typo: “distiction” > “distinction”

Line 377: Could compare with Johns et al., (2011). Also worth noting that whilst the overturning transport itself is slightly higher than observation-based estimates suggest, the associated heat transport is lower than observation-based estimates suggest. This discrepancy suggests potential issues with the vertical structure of the transport and the ocean temperature and salinity, evidence of which can be seen in Figs. 12 and 14.

Line 433: Typo: “troposhere” > “troposphere”

Line 449: Suggest citing e.g. Held and Soden (2006) here as a reference for the relationship between the hydrological cycle and Clausius-Clapeyron. Do you see the same relationship between the amplification of the surface salinity pattern and the increasing hydrological cycle as Zika et al. (2018)?

Line 466: Confusing sentence – reword.

Line 487: Not just EMICs, but also some low resolution AOGCMs (e.g. Hawkins et al. 2011). On this topic, have you calculated the F_ov for CLIMBER-X? In contrast to observation-based estimates, many low resolution models show positive values of F_ov.

Line 491: Typo: “equilibrim” > “equilibrium”

Line 506: Typo: “models’” > “model’s”

Line 673: “anyway” is not necessary - remove

“K" is multiply defined:

• Line 604: “K" is the kinematic vertical viscosity coefficient
• Line 697: the eddy kinetic energy, “K".

Line 767: Typo “atmophere" > “atmosphere”, and “computed” > “compute”

Lines 866, 868: “Mcdougall” > “McDougall”

Lines 932, 934, 956: Suggest changing “ocean water” to “seawater”

Line 944: Suggest rewording “Melting of sea ice leads to…” to “Melting of sea ice results in…” to avoid potential confusion with sea ice leads.

Figures 11 and 29: The white text on the contours is difficult to read.

Figures 3, 11, 13, 24, 29: add (a) and (b) for consistency with other figure style

Hawkins, E., Smith, R. S., Allison, L. C., Gregory, J. M., Woollings, T. J., Pohlmann, H., and de Cuevas, B. (2011), Bistability of the Atlantic overturning circulation in a global climate model and links to ocean freshwater transport, Geophys. Res. Lett., 38, L10605, doi:10.1029/2011GL047208.

Held, I. M., & Soden, B. J. (2006). Robust Responses of the Hydrological Cycle to Global Warming, Journal of Climate, 19(21), 5686-5699, https://doi.org/10.1175/JCLI3990.1

Johns, W. E., Baringer, M. O., Beal, L. M., Cunningham, S. A., Kanzow, T., Bryden, H. L., Hirschi, J. J. M., Marotzke, J., Meinen, C. S., Shaw, B., & Curry, R. (2011). Continuous, Array-Based Estimates of Atlantic Ocean Heat Transport at 26.5°N, Journal of Climate, 24(10), 2429-2449, https://doi.org/10.1175/2010JCLI3997.1

Mauritsen, T., Stevens, B., Roeckner, E., Crueger, T., Esch, M., Giorgetta, M., Haak, H., Jungclaus, J., Klocke, D., Matei, D., Mikolajewicz, U., Notz, D., Pincus, R., Schmidt, H., and Tomassini, L.: Tuning the climate of a global model, J. Adv. Model. Earth Sys., 4, M00A01, doi:10.1029/2012MS000154, 2012.

Williamson, D., Goldstein, M., Allison, L. et al. History matching for exploring and reducing climate model parameter space using observations and a large perturbed physics ensemble. Clim Dyn 41, 1703–1729 (2013). https://doi.org/10.1007/s00382-013-1896-4

Zika, J.D., Skliris, N., Blaker, A.T., Marsh, R., Nurser, A.J.G., & Josey, S.A. (2018). Improved estimates of water cycle change from ocean salinity: the key role of ocean warming, Environmental Research Letters, 13(7), 074036