This manuscript proposes a second phase of RCEMIP, focused on “mock-Walker” simulations. These simulations include hot/cold patches, which force convection to aggregate over part of the domain, alleviating some of the issues convective aggregation adds to uniform-SST RCE simulations. As well as describing the proposed set-ups, the text also presents some initial results from mock-Walker simulations with a cloud-resolving model (CRM) and a global model (GCM).

The first phase of RCEMIP was very successful, and phase 2 is certainly warranted. I appreciate the authors’ attempt to get this going; however, I have some concerns about the proposed set-ups given here:

• the study of mock-Walker simulations is less “mature” than that of RCE simulations were going into RCEMIP-!, which leaves some nagging questions about the set-up.  The authors note that for computational reasons the SST gradient will vary slightly across the models for the range of widths proposed here. While the variation is small across models (3% or less for most models), one could equally prescribe a set-up in which \Delta T is adjusted to keep the same gradient across the models. The issue is it is unclear whether it is the gradient or the absolute SSTs which matter most for the climate of mock-Walker simulations. My guess is it’s the warm pool SSTs, which would support the current proposal, but to my knowledge this has not been systematically investigated. There is some effort here to explore sensitivity to domain size etc., and it might be a good idea to do a more thorough investigation. Without this, it may be difficult to do an “intermodal comparison” as differences in model climate could be due to model physics or to differences in domain size.
• I was very surprised by the GCM set-up, which consists of zonal bands of hot and cold SSTs. I haven’t seen this set-up before, and don’t have a good feel for it. I would have expected the authors to propose uniform SSTs with hot and cold patches on the equator. Mock-Walker simulations confine the convection in x and y; here it is only confined in y. There could be similar issues with convective aggregation within the warm latitude bands as are seen in RCE simulations, and I would also expect large internal variability, which could interfere with the interpretation of model results.

For some inspiration, the authors might want to check out the following papers from Dennis Hartmann’s group:

Hartmann, D. L., and B. D. Dygert (2022), Global Radiative Convective Equilibrium With a Slab Ocean: SST Contrast, Sensitivity and Circulation, Journal of Geophysical Research: Atmospheres, 127(12), e2021JD036,400.

Larson, K., and D. L. Hartmann (2003), Interactions among cloud, water vapor, radiation, and large-scale circulation in the tropical climate. part ii: Sensitivity to spatial gradients of sea surface temperature., Journal of Climate, 16(10), 1441–1455.

I mostly focused on the model set-ups, but also have a number of small comments/typos:

L13: SST’s -> SSTs. I also think this should be “the same three mean SSTs”. The sentence is a bit confusing because it mentions 5 simulations covering 3 mean SSTs and 3 SST gradients. I understand it after reading the paper, but for the abstract the authors might want to word things more carefully.

L14: Sentence starting “Under weak SST gradients” -> maybe a personal preference, but the commas feel misplaced in this sentence to me. I would put the first one after “domain” and delete the third one.

L20: Suggest: “two; however, we also”

L26: “Operational models” to me signals forecast models, not GCMs. Also should be “such as”

L28: Should this be “uncertainty”? This sentence also makes it seem like the tropics are the only source of bias/uncertainty in climate projections.

L30: Not sure how Held, 2005 is relevant here. I also don’t know what it means to”effectively use” an Earth System model.

L45 and afterwards: I encourage the authors to provide a bit more information about these robust responses. E.g., what is the response of deep convective clouds to warming? How is the existence of self-aggregation robust? Etc. These results are important motivations for the present paper and should be discussed in detail.

L53: is insolation the only forcing? Surely greenhouse gases also? Also, in this paragraph “sensitive”.“sensitivities” is used a lot. Suggest re-writing.

L57: “proposal” -> “propose”

L58: delete “now”

L71: I found it odd the authors wrote “compared to RCE”, implying that mock-Walker simulations are not in RCE even though they are the basis of RCEMIP-II. Maybe the authors could clarify how to think about the set-ups in relation to RCE?

L84: Suggest replacing “provided” with “described”

L90: Sentence starting “In considering”, should it be “(1) how strikingly large”…and “(2) how strong”?

L101: what does “its” refer to here? In general, I found this paragraph a little muddled with frequent use of “coupling”, “clouds” and “circulation”. Suggest streamlining.

L108: Another personal preference, but I feel writing should stand on its own, and discourage the authors’ use of bold text.

L138: I would encourage the use of LES if anyone has the resources!

L163: I would capitalize “Cartesian”

L165: Suggest: “between the maximum and minimum SST”.

L180: Why did you switch from 2\pi to 360 degrees here? (Also equation 3)

Table 1: I encourage the authors to have 5 “core” experiments (listed here) and 4 additional experiments (295/0.625,295/2.5, 305/0.625,305/2.5). It would be great if some groups ran more experiments

L198: Should this be: “there is to be no diurnal…”, to keep the tense consistent?

L228: The authors should refer to the (large) literature on the pattern effect in this paragraph, which might also provide some insight into what to expect in terms of cloud feedbacks.

L252: Have the simulations been updated, or has the model code?

Figure 3: This doesn’t look much like a Walker Circulation to me…It might be interesting to plot the overturning stream function to get a better feel for what the flow is doing.

L315: Suggest re-writing sentence starting “This suggests” (remove “doesn’t “care””).

L329: This is a good example of my concern that mock-Walker simulations are not ready for this kind of comparison. This is a required simulation, but it takes longer than 200 days (the prescribed simulation length) to reach equilibrium. Maybe in other models all simulations need longer to equilibrate.

L433: “the behavior of \Delta SST”?

L461: Should this be “SSTs”?

My review is provided as a supplemental pdf.