Diagnosing ISO Forecast from GloSea5 Using Dynamic-Oriented ISO Theory
Round 1
Reviewer 1 Report
This study evaluates the tropical intraseasonal variability simulated by the GloSea5 model. The analysis is focused on MJO and BSISO and applies some of the conventional diagnostics established by previous studies. The manuscript topic is of interest to the readership of the journal. However, the manuscript has a number of methodological limitations (listed in the following) that need to be addressed for this manuscript to be reconsidered for publication.
There is scope to tighten up and refine the language to reflect professional and scientific content. Some examples include:
L51: The Kelvin wave easterly wind and Rossby wave westerlies
L53: A few studies suggested the slow change of shallow (or congestus) cloud to deep convection, which may be an important factor for MJO eastward propagation.
L162-163: two to six phases among a horizontal pattern of winds at 850 hPa and precipitation associated with the MJO are shown
L245-246: It was found from previous studies that the boundary layer is moistened, which develops slowly from the surface before the deep convection among the MJO phases
L357-358: In dynamic and thermodynamics structures, the weak BLMC anomalies in the Maritime Continent and western Pacific may result from poor convective heating by shallow or congestus clouds.
Several citations are missing from references or some references are not cited in the text.
L115: Wang et al. (2018) is not in the list of references.
L121: Wang et al. (2019) is not in the list of references
Fu and Wang (2009) not in the list of references
Cao et al. 2018 listed in references but not cited in the manuscript.
Goswami et al. (2003) and Goswami and Shukla (1984) listed in references but not cited in the manuscript.
The authors need to proofread the manuscript. The manuscript still has instructions provided by the journal’s template.
L128-130: The paragraph needs to be removed.
Section 2 needs to be Data and Methods and include a description of how and why regression on precipitation is used to compute moisture convergence, vorticity, etc. The methodology section needs attention to detail so that the approach is clearly defined and reproducible by an independent party. There are a number of significant issues that require attention including:
L132: SST what dataset is used for observations? Figure 1 shows 2-m temperature, which is not the same as SST. SST is not defined over land.
Figure 1 shows precipitation and zonal wind, however these fields are not mentioned in any discussions.
L219: Explain how diabatic heating is computed and which figure shows diabatic heating.
The discussion should omit content not specifically related to results of this study.
L175-177: No analysis shown here is related to moist parameterization or land processes. The authors cannot conclude that “failure of the GloSea5 in the western Pacific and Maritime continent is attributable to relatively poor parameterization related to MJO eastward movement (e.g., moist parameterization or land processes)”.
L339: GloSea5 simulations are not compared to CMIP5 historical simulations.
Figure 6: Why day-1 appears after 925 hPa? What are the units of the field shown here?
Figure 9: Wind vectors shown in black do not distinguish from the purple background. It has the same caption as Fig. 8.
Figure 11 Units described in the caption do not agree with units shown on the panels.
Summary and discussion address only the MJO simulation whereas the results also include the BSISO.
Author Response
Thank you for your valuable comments and suggestions. In the revised manuscript, 1) we added the result from the GloSea5 (hereafter ‘GS5’) climatological run to compare with GS5 hindcast data, 2) we revised all figure and captions with improved visibility and the detailed explanation, 3) we modified the model description, hindcast, and climatological run data, how to produce hindcast including ensemble perturbed methods, 4) we refine language to show more precise scientific expression, and 5) check the reference list as the reviewer suggested. We believe that the revised manuscript provides better figures and explanations by addressing your constructive comments.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This work provided diagnostics, mostly on dynamics and thermal dynamics aspects associated with the MJO and BSISO propagation using GloSea5 hindcast during 1991-2016. As compared to the observation, the authors indicated that the propagation properties of MJO and BSISO were reasonably reproduced in the model but with some weakness in the Maritime Continent and West Pacific. The authors attributed the weakness in MJO propagation mainly to the weaker mean-state SST and boundary layer moisture convergence anomalies and provided a mechanism for their results.
This is an interesting work that documented the MJO and BSISO related dynamics and thermal dynamics from different points of view. The manuscript is of scientific interest and significance. But the current form of the manuscript has some weakness on scientific soundness. One of my major concerns is the speculation. For example, the authors stated a causal chain in BLMC —convective heating by shallow or congestus clouds—implementation of convective trigger, deep convection entrainment rate and shallow convection scheme—easterly anomaly to the east of MJO precipitation—horizontal structure of MJO circulation—BLMC at the east of MJO center etc. I know their results could be consistent with other modeling work, but unless the authors provide direct evidence to support for the mechanism they described here, they can’t draw conclusion based on their speculation. As such, the presentation/description need to be improved. Along with other concerns, the manuscript is subjected to major revision before consideration of the publication.
Major comments:
1). The authors used “may” in many places, which sounds their statements are not certain. So the tone needs to be revised. If the statement is for sure, use affirmative sentence. If not, state like “based on.. we speculate …”.
2). Introduction
Paragraph 1: if the authors can extend a little on the prediction limitation especially on MJO propagation, it will make readers target the topics of this manuscript even more easily.
3). Please provide detailed information on the hindcast, data and methods instead of citing a previous manuscript so that it is more convenient for the readers to follow. More specifically, what are the basic physics schemes used in the convection, radiation, microphysics in the model? Since the authors mentioned the role of the shallow convective heating on the propagation of the MJO, it is necessary to indicate the corresponding schemes used in this model.
4). The hindcast used in this work is during 1991-2016 and the diagnostics is based on Nov.1-Apr.30 that initialized four times each month (1st, 9th, 17th and 25th) right? When compared with the observation, how is the observation matched with the hindcast? What’s the total sample size for the observation and hindcast?
5). One figure is missing (e.g. Fig.3c) and one figure is mis-captioned (e.g. Fig.9).
Other specific comments:
1). Ln 102: “The GloSea5 includes atmosphere, land, surface…”, which surface model is included in GloSea5?
2). Ln 105-108: “with three ensemble members”. So the diagnostics in this manuscript used ensemble mean? If so, please indicate it. Also, how is the ensemble generated? As I mentioned earlier, more detailed information concerning the hindcast needs to be provided here instead of citing a previous work. In that way readers can connect the data and the results more easily.
3). Ln 109-112: Both GPCP and ERA interim data are during 1997-2017?
4). Ln 112: the 20-70 days band pass was applied on what? Is it on seven month of integration? If so, please explicitly indicate it.
5). Ln 132-133: What’s the source of the observation SST? Does the boreal winter season in GloSea5 mean initial time or validation time? Since it indicates in Fig.1 that the temperature is 2m, what temperature did this work use exactly? Surface or 2m?
6). Fig.1: please add more detailed information in caption, such as how to get this maps (I know some information was provided in the text but it is more convenient for the readers to follow if author can provide more detailed information here). Please add like “Obs” on the top of the left panel and “GloSea5” on the right and they can add a column to show the difference between model and obs.
7). Fig.2: Same suggestion on adding “Obs” and “GloSea5” on the top and another column to show the difference (or model bias). Does the “filtered variance” mean variance of the 20-70 day time scale and “total variance” for all time scales?
8). Ln 175-176: Without showing the information on model test in this work, how do they know that the failure of the GloSea5 in WP and MC is due to relatively poor parameterization or land processes?
9). Fig.3: where does the MJO phase data come from? Is it from model or obs? Or a third party data? Based on my understanding, the hindcast data is 26 years (1991-2016), 6 months (Nov-Apr) per year and 4 times (1st, 9th, 17th and 25th) per month. Then the total sample size is 26x6x4=624 cases. How are the numbers of days obtained in the figures and why the number of days in model is much larger than obs?
10). Fig.4: Are the spectrum for model calculated based on seven-month integration and then averaged over all cases? Also, please use same color scale for the two plots.
11). Ln 207: Where is Fig.3c?
12). Ln 232-236: Like I mentioned before, the statement that moisture physical processes or land model may be poor is based on the speculation instead of direct evidence from same model in this work. If the results are consistent with other works, they can just state the consistency but can’t apply the conclusion from other work to the phenomena shown in this work. The author used “may be due to…”, suggest to reword it to indicate the statement is based on speculation but needs to be proved by further modeling test.
13). Ln 241-242: “black line”, do they mean the box?
14). Ln 285-290: What is the evidence for this causal chain?
15). I believe caption for Fig.9 is wrong.
16). Fig.11: How to explain that over 15N-20N,model shows stronger convective instability but weaker vorticity and BLMC than the obs? How to explain the positive convective instability near EQ in the model and how do these meridional structures affect the northward propagation of BSISO in the model?
17). Ln 356-364: Is the mechanism proved in this work or other work? If it is based on other work, the author needs to indicate that.
18). Summary and Discussion only included the MJO although both MJO and BSISO are investigated in this work.
Author Response
Thank you for your valuable comments and suggestions. In the revised manuscript, 1) we added the result from the GloSea5 (hereafter ‘GS5’) climatological run to compare with GS5 hindcast data, 2) we revised the model description, explanation of hindcast including ensemble perturbed methods, and climatological run data (this is used to check whether the weakness of the GS5 results from the model deficiency or not), 3) we improved figure quality and captions, and 4) revised the discussions as the reviewer suggested. We believe that the revised manuscript provides better figures and explanations by addressing your constructive comments.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors have addressed satisfactorily my major concerns.
Reviewer 2 Report
The authors made a thorough and substantial revision on the manuscript. The manuscript has been greatly improved. There are only a couple of minor edits needed before the acceptance of this manuscript.
1). Ln 71 "A study that the reforecast from e ECMWF.....".
Please revise this sentence.
2). Ln 128 "The GS5 includes atmosphere, land, surface....".
"surface" is still there although the authors indicated that it was misprinted in the response.
Author Response
We appreciate you for your valuable comment. We believe that revised manuscript has been improved.
Author Response File: 
Author Response.docx
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
