Atmospheric Rivers in Africa Observed with GNSS-RO and Reanalysis Data

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is an ambitious study on an important topic.  Please see attached comments.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

 

This study investigates atmospheric rivers (AR) over Africa, using ERA5 reanalysis moisture data compared with GNSS-RO data. The findings showcase consistent and promising results, significantly contributing to the understanding of AR behavior over Africa. The technical foundation and methodology are well-detailed, effectively demonstrating the characteristics of AR and the tracking method. The validation with GNSS-RO profiles presents consistent and exciting results.

For this reasons, my suggestion is to approve it for publication.

Techinical Sugestion

To enhance the visualization of the statistical comparison between GNSS-RO and ERA5 profiles for selected events, it is recommended to present the results in a table similar to Table 1.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

see pdf

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

Summary:

This study analyzed atmospheric river (AR) events over Africa over the 2009-2019 period, using water vapor retrieved by Global Navigation Satellite System (GNSS) radio occultation (RO) profiles. These GNSS RO water vapor retrievals were compared to water vapor analyzed by the ECMWF Reanalysis version 5 (ERA5), and it was found that although the RO and ERA5 results showed reasonably good agreement, the RO profiles tended to show a drier troposphere in moist regions in/around the AR transport jet, which was consistent with some prior studies. This study also applied a water vapor pattern-tracking algorithm (IPART) to the ERA5 fields, which yielded some interesting new findings on the climatology of AR events over Africa. For example, this study found that the December-February period, which is meteorological  summer in the Southern Hemisphere, accounts for peak AR frequency in Southern Africa. However, Northern Africa experienced peak AR activity during the Northern Hemisphere spring months (March-May). These results differ somewhat from previous studies, and the authors related these findings to the large-scale weather regimes driving these seasonal AR patterns over Africa. This paper also presented in interesting case study showing how the distant Amazon region of South America can provide a water vapor source for AR events in Southern Africa. Overall, this study is scientifically sound, with clearly presented results, and it covers a topic highly relevant to better understanding the African hydroclimate that has received relatively little attention in previous research studies. I am recommending Major Revision because the manuscript needs to have some additional information explaining how GNSS RO water vapor retrievals are obtained. Otherwise, the manuscript is in good shape and it will make a good addition to our scientific understanding of how AR events strongly influence Africa's precipitation climatology .

General Comments:

1) It is important that you add some more information to Section 2.1.2 about how the GNSS RO water vapor retrievals are derived. At the minimum, you should mention that they are retrieved from RO refractivity via a 1DVAR technique that uses gridded numerical weather prediction datasets as a background. Also indicate what background datasets were used. If the ERA5 was used as a background for any of these retrievals, this could cast some doubt on the significance of the ERA5-RO data comparisons reported here. Readers unfamiliar with GNSS RO technology need to know that GNSS RO water vapor profiles are not "pure observations".  Additionally, this manuscript would be further strengthened if you could provided some quantitative estimate of the relative contributions of the background and the RO data to the water vapor retrieval, if that would be feasible. 

2) Since both RO-retrieved water vapor profiles and ERA5 fields are not "pure observations", and both have numerous sources of uncertainty, the authors should consider their wording carefully in Section 3 and Section 4 when describing instances where the ERA5 forecasts higher IWV than a collocated RO profile. Unless you can provide some additional information (such as a collocated radiosonde or other reliable observation), we really don't know the extent to which the ERA5 - RO IWV discrepancy comes from the ERA5 having a moist bias or from the RO profile having a dry bias.

Minor comments:

Line 62: add “low” before pressure

Line 63: Please clarify whether this is vertical or horizontal vertical wind shear.

Line 82: It is interesting, and surprising for this reviewer, to learn that the polar jet stream can move over Africa. Would this be far northern or far Southern Africa? Some more specifics would be helpful.

Line 91: Please clarify whether this is boreal or austral winter.

Line 179: It might be worth mentioning that it is the high vertical resolution of GNSS RO profiles that makes them a valuable complement to reanalyses and other observation types when sampling the atmosphere.

Lines 384-386: In my manuscript copy, these lines have a strikethrough over them. I’m not sure why because they explain the meaning of variables shown in Eq. 1. 

Eqs. 2 and 3: It’s not clear from reading lines 406-412 what the set of “m” points with subscript “j” are supposed to be. Also is there a maximum distance tolerance from point “s” to the “I” points with known values? Please clarify in the manuscript text.

Line 616: Here you say that the average number of African ARs per year is 158, but on line 511 you say that this number is 159. Please clarify.

Line 646: Please clarify what you are referring to when you describe RO data’s “reduced sensitivity” in the lower troposphere

Lines 662-663: Recommend re-wording the phrase “remote sources like South America” to “a remote South American source”

Line 665: Maybe it would be better to add the phrase “surrounded by a dry environment” or something similar after
“strong moisture transport”. This would explain the large range of IWV values sampled by RO/ERA5 over the regional domain, and it is supported by Fig 8b.

Line 668: add “along” before “the regression line” and also remove “in” before “an RMSE”

Lines 681-685: The two sentences beginning with “The filled circles represent…” and “The IWV values derived from..” Are redundant and one of them should be removed.

Lines 685-687: How do we know whether the disagreement between the ERA5 and RO IWV measurements in high-moisture regions is caused by a positive bias in ERA5? If there are other studies that support this you should add some references.

Line 684: “illustrate” —> “illustrated”

Lines 721-738 and also 771-775: Recommend revising some of the wording in this discussion regarding ERA5’s accuracy. Just because the IWV RMSE computed between ERA5 and the RO data is larger for certain cases, can we really assume that the ERA5 data has the larger errors, and not the RO data? 

Line 778: “RO misses a part of the water vapor” Are you referring to the lowest penetration height of RO profiles tending to be some distance above the ground/ocean surface? Recommend clarifying this point.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 5 Report

Comments and Suggestions for Authors

This manuscript discussed 10-year Atmospheric River (AR) climatology for Africa.  It also has a small but very important component of inter-comparison of integrated water vapour derived from GNSS-RO  and ERA5  during selected AR events. To the best of my knowledge such comparison has not been performed so far and it makes the paper of high interest and concussions very relevant to other ERA5 only AR studies. The authors discuss appropriately the weakness of GNSS-RO and one suggestion can be to use ground based GNSS from Africa to further confirm their findings.  International GNSS Service (IGS) may have appropriate stations and time series for such attentional comparison. Overall, the results in this manuscript are interesting but the structure of the manuscript will require further revisions.  Some suggestions are given below.

Title:

The title can be improved by specifying the satellite observations namely by replacing satellite which is too general with GNSS-RO.  

Introduction:

Please consider improving introduction content by reviewing GNSS-RO state of the art as well as AR tracking methods.  Please motivate why you selected IPART for AR detection.  As such introduction is AR focused with unnecessary repetition and extensive motivation of socio-economic impact.  Which gap is your paper going to address?  Please state clearly the aim of your manuscript as this is critical to evaluate if your work reached this aim.

Materials and Methods:

Please chance the section name to "Data and method". There are no materials as such in this manuscript.

Page 5, line 225-228: Please consider moving this paragraph in the Introduction section of your manuscript "Vertical profiles derived from these measurements have been widely used for monitoring atmospheric temperature and water vapor [27, 28, 29, 30]. GNSS-RO data are increasingly employed to observe ARs and improve AR forecasts, as they provide detailed vertical moisture profiles, global coverage,  and all-weather capability [ 31, 32, 33, 34, 35, 36]."  A manuscript structure is compromised please pay attention to the required elements for each section.  Method section only gives specifics for the implementation of the technic not introducing it. Introduction of GNSS-RO is to be done in the  "Introduction" section.

Page 5, line 241-243: Please consider moving this paragraph in the Introduction section of your manuscript "While GNSS-RO data tend to underestimate IWV compared to satellite-based  instruments like SSMI/S, they align closely (~95%) with ECMWF background data. This systematic difference can be attributed to GNSS-RO's limited sensitivity in the lowest few hundred meters of the atmosphere [21] ." 

Subsections in the section "Materials and Methods" are unacceptable short.  Please consider removing subsections shorter than 5-10 sentances. Expected are structured paragraphs not 2-3 senates cut by subs sub section. Structuring a manuscript is a vital part of communicating work.  It is advisable that authors check online course for manuscript structure and content. 

Results:

Avoid using "boreal winter/spring" in the manuscript. Once you define winter (DJF) spring (MAM) use the abbreviations DJF and MAM instead.

Other suggestions:

Page 3 line 115-118

The following text is best suited for the discussion section "However, our findings  reveal that AR activity in Northern Africa also exhibits a significant peak during boreal  spring (March to May), suggesting a more prolonged period of elevated AR activity than previously reported. Boreal fall (September to November) appears to serve as a secondary period of AR activity, especially in October, rather than the primary peak. " Please consider moving it there. In the introduction section review of the state-of-the-art is given without the finding in the present manuscript.  

Page 3 line 122-124

The following text is best suited for the discussion section "Our analysis, however, indicates that the primary peak in AR activity occurs during austral summer  and autumn (December to May), aligning with the region’s rainy season. Austral winter represents a secondary peak, consistent with earlier observations." Please consider moving it there.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

To the Authors,

Thank you for your revisions.  Aside from now missing a 'Figure 9' in the body of your paper, I see no persistent issues with this manuscript. 

The Reviewer

Author Response

Comment 1: Thank you for your revisions.  Aside from now missing a 'Figure 9' in the body of your paper, I see no persistent issues with this manuscript.

Response 1: Thank you for this observation! The mistake has been corrected, the previous Figure 10 has been renamed to Figure 9 to and the text was improve accordingly. (p.18, lines 652-653)

The former Figure 9 in the Appendix was therefore switched to have the figure caption: Figure 10. (p. 21, line 773)

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have turned this paper around very quickly and did not read the paper critically as they should have done. Although improved, the paper is still too long.

1. Section 1.2 needs to be redrafted to take into account the results of this paper, which contradict what has been published previously. 

2. Section 2 is too long, particularly 2.3.2 which laboriously defines terms like inverse distance weighting, integrated water vapour and specific humidity. There are whole sentences in section 2 which are redundant, e.g. l.317-322, 334-336, 342-343.

THe discussion of ENSO is still not satisfactory. Why is there no figure showing the ENSO index together with the AR statistics, so that readers can see for themselves whether there is a credible link? The authors have simply added a caveat at the end of this section, which does not address the question of whether it should be there at all. Of the three highest numbers of ARs in fig. 2, two are la Nina years and one an el Nino year. The text is just waffle and either needs rewriting in a rigourous manner, or removed.

 

Author Response

Please see the attachement.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

The authors have done a good job in responding to my comments and revising the text. This paper should provide a good addition to scientific literature on the African hydroclimate.

I recommend acceptance after the authors give the text a final proofread. For example, there are still typos, such as "tropspgere" on line 131 or "missin" on line 135.

Author Response

Thank you.
We have corrected the two typos identified, and checked for others.

All references are relevant to the content of the manuscript - and we have now put

them in the right order, reflecting the changed structure of the manuscript.

Reviewer 5 Report

Comments and Suggestions for Authors

Thank you for addressing in detail the suggestions!  

Author Response

Thank you.