Robustness of the Link between Precipitation in North Africa and Standard Modes of Atmospheric Variability during the Last Millennium
Round 1
Reviewer 1 Report
It's a paper, with limited scientific support. The research is a compilation of databases to which a statistical correlation procedure was applied.
The authors only describe the results, but they are not discussed. The paper mixes results and discussion. Also, methodologies have not been described. For example, statistical tests to detect series rupture.
The conclusions presented by the authors are not supported by the results of the research. The conclusions are comments that can easily be deduced from a basic statistical analysis of the time series used.
Figure 1 should be placed after it is named in the text, i.e. it should be placed within section 2.
Table 1, is not cited in the manuscript.
Why is in figure 2 the precipitation values for the period 1979-2005 are shown if inline 101 its is mentioned that the first period of the analysis is from 1979 to 2013. Figure 2 has many inconsistencies; it is not clear if these are observations or simulations. The title of the figure mentions simulated and observed data, but it is not clear what they are.
Remove the header within the image in figures 1 and 6.
Figures 4, 5, 6 and 7 are presented before they are mentioned in the text.
It is very difficult to read the document and follow the steps taken by the authors to arrive at their conclusions. The results are mixed with the discussion which makes it difficult to understand the scientific contribution of this work.
Table 3 is very limited in the information it provides. It makes no sense to present the means and variances of the correlation coefficients. It is more important to know the values of the correlation coefficients.
Line 330 mentions Pettitt's test, however, it only mentions that it was applied, but this test detected a "change point". This result is very important and the authors do not address any discussion about it.
Figure 8 would seem to be of great impact on the results, however, it is very confusing and should, in its case, focus the discussion on the results represented by the red line within this figure.
There is no section in the document to follow the procedure the authors propose to correlate their results with the PDSI index.
The authors are encouraged to reorganize the manuscript and present detailed results for each time period analyzed. Discuss the results in each period and submit a clear presentation of how the drought index was used. This part of the PDSI is very confusing and limited in the text.
Author Response
Dear reviewer,
Please find attached my responses to comments
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments on “Robustness of the link between precipitation in North Africa and standard modes of atmospheric variability during last millennium” by Arab Djebbar, Hugues Goosse and François Klein.
In this research, the authors studied the winter variability and trend of precipitation over the north Africa region during the last millennium. For that the authors assessed the link between the winter precipitation and the atmospheric circulation patterns, using a variety of datasets. The Palmer Index was calculated confirming the dry trend. Although the idea between precipitation and the atmospheric patterns is clear the authors do not clearly explain the reason why they want to do it, i.e., the goal or the relevance of such study.
Overall, I think this manuscript has potential within the scope of the journal, with relevance for the stakeholders interested in north Africa precipitation. However, I think this manuscript needs a deep revision before publication. I have several suggestions I think would improve the paper, specifically in the introduction, method, results, and conclusion sections. One primary concern is the method and results, parts of the methods are missing, and results are presented without any logical threading. One of the missing parts is the logical explanation how the link between precipitation and the atmospheric patterns is assessed.
My concerns are as follows.
Abstract
The abstract focusses the main points but the purpose behind this study is missing and reason to reason to seek establish a link between precipitation and the atmospheric modes of variability for the past 1000 years This section does not report all the major findings from this investigation. It’s not clear why the CMPI5 models are used and the relevance of this study for the region.
Introduction
In general, ideas are somehow not well explained, sometime unconnected or very incomplete even missing some parts.
The authors provide a lengthy and confuse review between the Mediterranean region, its variability, climatic characterization, and its association with the major modes of variability, yet the region of interest is the western part of north Africa. I acknowledged that the region in consideration is part of the Mediterranean region but also have an Atlantic coast. This gives climatic features that not fit under the umbrella of “Mediterranean region”. I would suggest rewrite the first six paragraphs focusing only in the north Africa with a division like the one presented in Figure 1 - Atlantic and Mediterranean coasts. I also advice to choose one name for the region and use the same terminology throughout the manuscript (e.g., north Africa (title) Mediterranean (L26), “north Africa” (L34), “southern part of Mediterranean” (L42), “western part of “Mediterranean” (L47), “North Africa region” (L98)). I would avoid the word “Mediterranean” because it’s a broad region and this study is focuses on a very particular area.
Literature review for some topics like the use of the CMIP5 models, the PMIP model, the Palmer index, the datasets, and the state of the art for the region are missing.
The motivation of this study (l97) needs to be better link to previous paragraphs.
Introdution
Data and Methods
The Methods section should: describe how an experiment was done, give a rationale for why specific experimental procedures were chosen, describe what was done to answer the research question and how it was done and explain how results were analysed. This section is very incomplete. Be thorough. Add every step of calculations.
Here are some examples:
It is not clear the need to divide the region in 3 distinct precipitation regions and just one for the wind. Also, it is not explained why the wind box is geographically different from the ones with precipitation. These details matter. Also, one of the boxes covers more ocean than land and this can be problematic when averaging over the regions (the number of points over land vary).
In Table 1 there are 3 datasets for precipitation. Why use three? These datasets differ in number of points (resolution), e.g., the 20 century Reanalysis, and one of them is explicitly over land (GPCC v7). For the remains that information (over land/ocean) is missing. Why the data was not mapped to a common grid, with the same resolution and number of points? This is important because further ahead precipitation is averaged over domains with different resolutions.
The period for the variables other than precipitation is missing.
Explain the need to use the multi-model ensemble from the CMIP 5 simulations. Separate the CMIP5 and the PMIP information. As it is, is very confuse.
In section 2.3 please clarify if the indices are time series of normalized SLP or time series of the leading EOF of SLP anomalies.
The methods used are missing. We need some information regarding what was done with these data; details such as averaging, correlations etc, although the basic statistical techniques are known they must be described. In the same order that information is being presented in the results section.
Use the same terminology throughout the manuscript. ” Past1000 and historical period” (Line 140; line 147) but “past millennium”, “past century” in table 3. Please verify headers and caption of figures or tables.
Why did the authors not present the climatological characteristics of the region? This would provide more details about the precipitation regime and differences among datasets. How were calculated the precipitation series for each region?
Results and discussion
Please rewrite this section. Some parts are not completely clear (see the methods) and some of the results lack an interpretation of their implications and relevance. Do not put all the tables/figures before the text. This section must start with text and the tables and figures must go along after the text.
L 291-295- This statement is not entirely true. If I understood correctly, the correlation between precipitation and the mode of variability (NAO, …etc) was calculate. Next, all the values were averaged. I see here some caveats. First, only one figure is shown (precipitation from the Dalware dataset, Figure 4). There is no information regarding the remaining datasets. I am wondering how the other datasets correlated with the indices. Strong correlation, weak correlation? This is an interesting result and a very useful conclusion. I fail to see why aggregated the correlation results. Thus, there is no added value in the analysis. Let's imagine that one has a high correlation and the other has a low correlation, the average value is going to be the middle. In terms of physical processes, it tells very little about the degree of association among the winter precipitation with the wind at 850hPa or the indices.
Tables
Table 1 missing information about the period for the 20-century reanalysis.
Table 3 Identifiy the periods in the hearder. Caption. Identify the red, bold etc used.
Figures
The figures have poor resolution except Figure 4.
Figure 1. Above each box print the name to identify the west, centre and east box. Poor resolution the dashed and the dotted lines are indistinguishable.
Figure 2 - Poor resolution. Lines are indistinguishable. Identify the lines using markers. Fits better in results and discussion. Caption. Identify the colour of the observations.
References
Some references do not follow the journal indications.
Minor Comments
L 16. We-> results
L.28. [2]I would change this reference for a more recent study.
L 30. [4] Won't there be a more recent study?
L.38. They(?)-> Those studies
- 77. Holocene(?) do we need to go back that much?
L 117. Table 2 -> perhaps Table 1.
L 154. Mediterranean-> be specific. Focus on region of interest and not in a broad area as the Mediterranean.
- 158.160. In my opinion this fits better in Results and Discussion section. I don’t see any good agreement.
L 264. No figure or table to support this claim. The authors must define what is a weak autocorrelation. This can be done in the methods section.
Author Response
Dear Sir,
Please find attached my responses to comments on my transcript.
Best regards.
Author Response File: Author Response.pdf
Reviewer 3 Report
See attached file.
Comments for author File: Comments.pdf
Author Response
Dear Sir,
Please find attached my responses to comments on my transcript.
Best regards.
Author Response File: Author Response.pdf
Reviewer 4 Report
Dear Editor of the International Journal of Climate
I am very excited to have been given the opportunity to revise the manuscript entitled: “Robustness of the link between precipitation in North Africa and standard modes of atmospheric variability during last millennium”.
The manuscript by Djebbar and co-workers describes the author’s investigations into the relationship between the precipitation variability and some atmospheric patterns across the northern parts of Africa during last millennium.
The manuscript is generally good, but it needs to be revised, organized better and rewritten again.
Some points need reformulating in addition to that, I suggest expanding the analysis of the other teleconnection patterns affecting precipitation variability in area of study to take more accurate and logical results.
The introduction and methods of work have been well formulated, but the results could be reproduced
Major Comments:
I think it will be difficult to judge the validity of the main result, since the rest of the teleconnection patterns that influence precipitation variability are not analysed yet. The Upper-Level Mediterranean Oscillation index (ULMO), East Atlantic (EA) pattern, East Atlantic/West Russia (EATL/WRUS) pattern, Scandinavia (SCAND) pattern have demonstrated high relationship with the precipitation variability ad precipitation concentration over the Mediterranean. Look at Krichak et al, 2013, Redolat et al, 2019 and Mathbout et al, 2020.
I would suggest including the effect of these other teleconnection patterns in this manuscript to judge the end result of a more effective form.
Minor Comments:
L16. Please delete the space between two words “link” and “between”.
L16. I suggest replacing the wind velocity term into “wind speed” in in the whole manuscript.
L16. I suggest replacing the “link” word into “relationship”.
L29. Please delete the space between two words “characterized” and “by”.
L29. Please remove the space between two words “and” and “dry”.
L30 -l3.I suggest completely reformulating the sentence.
L35. I suggest replacing the “perturbations” word into “disturbance”.
L150-L151.I would suggest giving a brief idea within the context of the text and not just referring to the reference.
L 153.Do you mean their?
L157. Please remove the space before the word “underestimate”.
L67-L70.I would suggest reconsidering the section between the two lines and looking at some following reference studies.
“González-Hidalgo et al 2009 indicated that the strong precipitation events in the east of Iberian Peninsula are related to negative phases of the NAO, MO and WeMO. And, López Moreno et al., 2011; Fernández-González et al., 2012 have demonstrated that the increased precipitation along the Iberian Peninsula is mainly related to prevailing negative conditions of NAO and MO “.
L77. Please remove the space before the word “A reconstruction”.
L104. Please remove the space between two words “the” and “analysis”.
L170. Developed by [59].It is not clear for the reader. There is a clear break in the explanation and thus becomes incomprehensible to the reader. Please brief the explanation and not just mention to the reference.
L266/L267.This result is contrary to what was mentioned in the previous paragraph (between lines 67 and 70) which I mentioned to in a previous comment. Please explain the idea well.
L227- L334.I would suggest moving parts of this section to methods.
L360. Please remove the space between “of” and [30].
L372-L377.I would suggest moving this section to the methods.
L416. Please remove the space before “additionally”.
Tables and Figures:
Figure 1.I would suggest reformulating the figure caption, as it is not understood. I would also suggest redrawing the map in a more clear way
Figure 2.The figure legend is not clear.
Figure5. Axes labels are not readable.
Finally, I propose to reformulate some sections of this manuscript in accordance with the title and objectives of the article first before giving the decision in its publication.
Author Response
Dear Sir,
Please find attached my responses to comments on my transcript.
Best regards.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors made all corrections and changes requested in an appropriate and exhaustive form. I agree with the final version of the manuscript.
Author Response
We would like to thank the reviewer for the positive feedback addressed on our revised manuscript.
Best regards.
Reviewer 2 Report
2nd Review
Comments to the Author
The authors have undertaken a comprehensive review which has much improved this manuscript. The new version of the manuscript includes substantial changes and results are well presented with much interpretation which mostly fulfill my previous concerns. I appreciate the efforts of the authors in producing this new version. I think that now it is better written, the new abstract establishes more clearly the objectives and methods of the paper, the figures need some minor improvements, but the major points have been addressed.
Overall, I would recommend the authors to undertake only a minor revision.
Author Response
Dear Sir,
Please find attached the responses to the comments.
Best regards.
Author Response File: Author Response.pdf
Reviewer 3 Report
The language of the manuscript has been significantly improved, and it now reads well.
Figure 1: The map has no scale or spatial reference. It would be good if this map included all of the Mediterranean and the Iberian Peninsula, since these regions are referred to in the text, but might not be known to the international reader.
Figure 4: Change the x-axis label to Month
Line 97: Define the Common Era by years. Suggested edit “…Common Era (CE, year 0-present)”
Line 101-103: This region could also be shown in your map in Fig 1.
Author Response
Dear Sir,
Please find attached the responses to the comments.
Best regards.
Author Response File: Author Response.pdf
