This manuscript focuses on the development of a global coupled model on forecasting MJOs. The propagation of MJOs along the equator can significantly affect the precipitation in many regions, so the relevant model works have been devoted by many previous studies. I appreciate the authors’ efforts for continuously improving the model forecast on this multi-scale weather system. Unfortunately, one thing I am trying to find in this manuscript is their unique contributions to the broad society. According to the title, it seems like the authors feeling confident in the usage of a 1-D SIT model for predicting MJOs. At the end of Introduction, the authors barely mention their motivation is to “examine how air–sea coupling can improve MJO simulation, especially that of the eastward propagation that has been poorly simulated in many climate models”. Because many global coupled models use the 3-D ocean models, the connection between the title (1-D SIT model) and motivation (effect of air-sea coupling on MJO propagation) is unclear. Are the authors trying to convince readers the effect of 1-D model enough for the forecast? Or is there anything special inside the SIT model? The importance of air-sea coupling should have been extensively emphasized and agreed by many studies, and I do not think any ongoing research still trying to use a global model without ocean parts. Repeating the work may be meaningless. I believe their motivation needs to be rewritten.

On the other hand, because the authors introduce some models unable to simulate the MJO propagation reliably, I believe one of their expected results is to improve the motion of MJOs (also mentioned in the motivation). However, it seems like the authors do not summarize how much improvements can be seen in their results, or which factors can affect the simulation the most. Because there are some interesting experiments inside this manuscript, such as the coupling regions, I do not think it should be rejected at this moment. However, the structure and quality of the manuscript are very poor. It is very close to my standard for rejection (too many things to be fixed). I only list some problems below, not all. I recommend a major revision for this work in this review.

1. I do not think conducting an experiment for studying the difference between A-CTL and C-30NS is needed. In my point of view, we do not need another paper talking about the importance of coupling the upper ocean in the global models. In other words, please simplify the description in section 4.1. All you need is to show your coupled model sufficient for simulating the MJOs.

2. I am super uncomfortable in the description of the ERA-interim results as the “observation”. It is impossible to measure the global wind at 850 hPa directly. Besides, the precipitation data looks like a post-processed product constituted by many satellite measurements. It happens to the OISST as well.

3. I think you need to reconsider your structure in the main text. There are some unnecessary and redundant materials that can be moved to the appendix or supplemental material. For example, you do not adjust the coefficients in the 1-D TKE closure scheme. Why do you need to describe the full equations? I also don’t care about the numbers of depths from lines 207 to 212 (yes, your units are wrong).

4. You do not need section 3, because people like me already forget the details when we are reading sections since 4.2. Please reorganize the structure.

5. I do not think that section 4.2 is discussing the vertical resolution… It is more like the thickness of the first layer. A lot of information is missing here. For example, what is your surface mixed layer depth? If the surface mixed layer depth is less than 30 m or 10 m, what do you do for C-LR34m C-LR12m? Are you trying to test the effect of a slab model in your global coupled model?

6. What do you mean “ocean bottom” at line 476? Is it seafloor?

7. Rewrite section 4.6. I cannot understand which fluxes you are using.

8. I cannot understand why the runs are 30 yr? What are the initial conditions of atmosphere and ocean? Is the forcing the same as the values in the real world from 1990-2020?

This is a very well written paper providing comprehensive testing on how air-sea coupling can improve simulations of the Madden-Julian Oscillation (MJO) in the CAM5 AGCM, such as role of the vertical resolution and depth in the ocean model, coupling domain, and diurnal coupling. While findings from this study largely support previous studies, they provide important guidance in designing an optimal air-sea coupling strategy for improved MJO simulations. I would be glad to recommend this manuscript for publication after a minor revision.

Minor comments:

When describing model results, I would suggest to use “present tense” instead of “past tense” throughout the paper.

Line 37: move “in the year 2011” after “Dynamics of the MJO”?

Line 68: may delete “and climate models”

Line 109: may change to "regarding the effect of air-sea coupling on the MJO"?

Line 273-274: Are U850 anomalies not averaged over 10N-10S, instead of just on the equator?

In general, figure quality can be improved (many look blur with detals difficult to identify), and some figures can be a bit enlarged. 

Line 305: the “observed” MJO characteristics

Line 467: in the first few meters “below the surface” allows ….?

Line 556: I didn’t see faster MJO propagation when the diurnal coupling is turned off based on Fig. 9b. If compared to Fig. 5a, seems to me the MJO propagation speed is even faster in the C-30NS run with diurnal coupling. This is also related to the following comments on Fig. 10. Generally, I don’t see significant differences in MJO simulations between the no-diurnal coupling experiment and the control experiment.

Fig. 10: It would be better provide more details on how the U850 and P slopes are determined, e.g., based on which longitude bands. Also the colors for “C-30NS-nD” are not consistent between the figure and legend.