Correlating Extremes in Wind Divergence with Extremes in Rain over the Tropical Atlantic
Round 1
Reviewer 1 Report
Review on “CORRELATING EXTREMES IN WIND DIVERGENCE WITH EXTREMES IN RAIN OVER THE TROPICAL ATLANTIC” by King et al.
In this manuscript, the authors use C band scatterometer wind vectors to derive the wind convergence and divergence and analyze the correlation between the divergence and rain extreme. Contingency tables and Odd ratios are used to reveal the correlations. The study is novel and interesting, especially demonstrating the extended application of the C-band scatterometer winds compared with the Ku-band scatterometer data. The methodology and results are detailed, and the paper is well organized. I support its publication in “Remote Sensing”. My concern is how to deal with the different spatial resolution between the rain data and the scatterometer wind in the rain data regridding. The size of the rain data cell is 3 km, but the wind divergence is 12.5 km. Thus, the 1 wind convergence grid should include 4X4 rain cells.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In the study relationships between wind divergence and rain in the Inter-Tropical Convergence Zone (ITCZ) of the Atlantic Ocean were found using 12.5 km wind fields from the ASCAT-A and ASCAT-B tandem mission, collocated with short time series of Meteosat Second Generation 3 km rain fields. The research is interesting and valuable for better understanding of processes over the tropical part of Atlantic Ocean. Nevertheless, I have a few remarks which may improve the manuscript. Specific comments are given below.
First of all, format the ms following the instructions of MDPI Remote Sensing.
Page 1, Line 1 and many other examples: this expression is not necessary. Format the ms following the instructions of MDPI Remote Sensing.
Pages 2-5: the pages are not necessary. Format the ms following the instructions of MDPI Remote Sensing.
Page 8, Figure 2: Add scale and the unit of wind. Add the explanation of longitude (E & W) and latitude (N & S).
Pages 8-9, Lines 228-262: Although these lines broadly introduce the objectives of research, I would prefer to have a shorter, but comprehensive information on objectives and range of research.
Page 9, section 2.1: could you write a little about causes and range of errors of satellite-derived wind speed and direction.
Page 10, Line 288 and many other examples: sometimes you use Fig., sometimes – Figure. Unify according to Remote Sensing standards.
Page 10, section 2.2: could you write a little about causes and range of errors of satellite-derived rain rates.
Page 11, Figure 3: Add the explanation of longitude (E & W) and latitude (N & S). Explain the meaning of “I” & “III”
Page 12, Line 347 & Figure 4: perform a simple statistical test to confirm that there are no significant differences between winter and summer. Is it necessary to use capitals for “Winter” and “Summer”?
Page 13, section 3.2: there is no text in this subsection. Please describe the figure and the table in the subsection where they are inserted. Do not start the subsection with a figure or a table.
Page 13, Figure 6; Page 14, Line 383; Page 15, Figure 7 and others: I wonder why you use the Gaussian distribution. One more sentence on the reason of using it could be added to the methods section. Moreover in my view, the values below 10-6 are cut in Figure 6.
Page 15, table 2: the table is not cited in the ms.
Page 16, section 3.3.: add numbers of subsections
Page 16, table 3; Page 17, Figure 8 and many other examples: insert figures and tables into subsections where they are described.
Page 21, Figure 11: the figure is cited in other subsection (5.1.2).
Page 22, Table 6: the table is cited in other subsection (5.1.3).
Page 22, Equation 16: it is rather a table, not an equation
Page 23, Equation 17: it is rather a table, not an equation
Page 24, Figure 13: do not start the subchapter with a figure, please
Page 25: I can’t see any discussion of results with earlier studies.
Page 26, section 6.2.: An equation and a figure look strange in the last chapter of ms. Avoid inserting figures and equations in the last section.
Page 30, Figure A1: in my view, the values below 10-6 are cut.
Page 31, line 673 – Add the title: References
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Accept in present form
Author Response
Thank you.
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.
Round 1
Reviewer 1 Report
The authors did a interesting work, which aims to reveal the correlation between wind convergence / divergence and rain based on the ASCAT winds. I think it is publishable when the question is treated properly.
1 How do you deal with the contamination of rainfall on ASCAT wind retrievals?
Reviewer 2 Report
The paper tried to reveal the rain and wind field interaction using an empirical analysis of ASCAR derived wind fields collocated with MSG dataset. The authors claim that when there is an extreme rain, there is wind divergence or convergence in the vicinity of the are under precipitation, providing some statistics on how synchronous these events are.
Exploring the rain-wind interactions over oceans to improve the knowledge associated with MCS can be a valuable endeavor. However, the paper does not convince me for a publication. My main concerns:
- The paper does not discuss how rain might affect the scatterometer obervations. I am aware of the huge efforts conducted to characterize rain effects on ASCAT measurements. Going through the published works (e.g. [4], [5], [6]), I can not conclude that there is no non-geophysical effect of rain, in a sense of attenuation or splash effects, on the measurements. Even though the wind variability in such conditions is greatly discussed, as in the most of these publications suggest, the rain contamination needs further investigations. I cannot believe that, at least at extreme rains, the observations are unaffected and whatever seen is a geophysical signature. This might be why in [7] a rain rate threshold (<3 mm/h) is established.
- Extreme rains are highly correlated with highly variable and extreme winds. In addition, the errors (both Gaussian and systematic) are expected to increase in scatterometer measurements at high winds and in rain conditions. As a result, the probability of artificial divergence events in the captured wind field by scatterometers is expected to increase. Implementing a statistical approach, can be dangerous and I am not convinced the statistical approach, implemented in this study, presents a geophysical fact.
- The results do not support the claims and conclusions made. Figure 5, reports on a low probability of synchronous extreme rain and divergence. The authors discuss on the size of neighborhood as a reason. Figure 12 provides a histogram of shortest distance from an XR to an XCD. My question: Is this figure for all the observations and is there and XCD for any XR. If it is only for the cases that there is an XCD for the XR, it cannot justify the false alarms in Table 5. The authors could repeat Table 5 with an optimized neighborhood size. Before that, they need also need to discuss the distribution of XCD in a non-rain affected wind field (otherwise always an XCD is captured with large enough neighborhood size). Currently, the results are too confusing and I am not able to decide if the method shows a success.
- The authors might need to discuss better on the novelty of this work. What is the main message? The Scatterometers (more specific: ASCAT) ability in observing MSC surface wind divergence events, or introducing a geophysical phenomenon? Both are discussed in [7] (not discussed in the introduction). In the summary, the authors highlight that they are focused on methodology, what are the advantages of this method?
Nevertheless, I would suggest major revision, being open to be convinced. However, I would like to highlight that, in my opinion, an extensive revision including robust discussions and clarifications are needed before any publication. In addition, the authors might consider following posed questions:
- Line 1-6: The scatterometers measurements are assimilated into NWPs. If the scatterometers are able to resolve the small-scale wind field, why NWPs are not (e.g. ERA5 having ASCAT measurements)?
- Abstract: If the authours are focused on methodology (as they mention in the summary) they could be more focused on the advantages of their method, rather than ASCAT ability and so on, which are already demonstrated.
- Line 49-51: Please be more specific and discuss the cited studies.
- Line 54-56 “In line with this, recent studies show that ASCAT-derived wind fields and its spatial derivatives are indeed useful under moist convection conditions”. How is it related to the previous sentence “Rain contamination, i.e., signal attenuation, volume scattering from drops, and rain splashing effects at the surface [5]”. The authors must discuss these effects in their case and clarify how they may affect the study avoiding generalization.
- Line 67-72 “This paper therefore aims at further exploiting the information content of the ASCAT-derived wind”: Not consistent with the summary, if you are focused on methodology (as mentioned in Line 302). Please clarify the novelty and the study objectives. If you are focused on methodology, discuss previous methods and advantages of your method, otherwise please clarify what are the new derived information on MSCs in Tropical Atlantic (as commented above).
- Figure 2: The colorbar needs a unit.
- Line 86-87: “the higher Ku-band radar frequency results in about a sixfold impact of rain 87 attenuation and scattering” needs a reference.
- Figure 3 comes before the citation in the text.
- Line 142: What is the exact number of collocated observations? There must not be too many observations at extreme rains, is that enough for your analysis?
- Line 149 “Second, since rain is a spatially highly variable quantity, it is desirable to average over the smallest achievable spatial footprint”. What the authors mean by the smallest achievable spatial footprint?
- Line 161: Please define WVC before using the abbreviation.
- Line 167: What is timulus-response (SR) contingency tables? Reference?
- Line 184: Please be specific why the threshold based on airplane-scale measurements can not be a good idea?
- Generally, the figures are not optimally located in the paper. It is not easy to follow the figures and their discussions in the text. The authors should suggest the optimal location in the submitted version, facilitating the review process.
- Line 330-335: The results do not support the claims. “a 50km-by-50km spatial neigbouhood would likely achieve a significant reduction in the number of false alarms and misses.” Why Table 5 is not repeated with this neighborhood size? Currently we cannot make any conclusion.
Reviewer 3 Report
This manuscript presents studies on the correlation between the wind convergence and divergence and rain fall. The studies are novel and help to better understand air interactions in Mesoscale Convective Systems (MCSs). I recommend to accept this manuscript after some minor revisions listed as below:
1) Line 62-63: The wind convergence and divergence are not clearly to be shown in Figure 2. The authors may clarify them in Figure 2.
2) Line 100-111: It is mentioned that "the NWP forecasts from the European Centre for Medium range Weather Forecasting (ECMWF) model are interpolated to the swath grid and packaged with the ASCAT product." Is the interpolation to increase the scatterometer wind spatial resolution? If it is, what the final spatial resolution of the wind product?
3) Line 141-142: Since the study area between 25S and 25N, so it is better to add boreal winter and boreal summer here.
4) LIne 181: What is WVCs here? Please give the definition.
