Dear Authors
Thank-you for your revisions. Eventually I may send this back to reviewer 1 (who recommended rejection) but I think some more improvements are needed first.
Yours sincerely
John Huthnance (editor)

Specific comments.

Lines 127-128. I think you should include here some of your response to referee 2 comment 1 about removing discontinuity between AVISO and FSMSLA. Maybe refer to Taqi et al (2017) for more detail?

Lines 131, 135. Two equations (2)!

Lines 137-138. The choice of level of no motion is still arbitrary. You have shown only pragmatism and not logic in choosing 500 m. There might be flow at all depths or indeed significant flow only in the top 200 m. Evidence should be used in the choice and Bower and Farrar (2015) refer to ship-borne ADCP measurements usually down to 600 m which would be good evidence. Does the choice between 500m and 700m make a significant difference to the near-surface geostrophic currents?

Figure 2. You have not answered the Referee 2 question about how you show correlation with AVISO data near the coast where there are no AVISO data.

Figure 4. Referee 2 asked for justification for these specific areas and periods. Your revised text explains the times but not the areas; Referee 1 found that the cruise in Bower and Farrar (2015) covered a larger area. This is another case where the referee suggests more quantitative analysis.

Table 3 for March 2010. I am wondering if the values are all correct in that the RMSE and Stdv are both worse for FSM-SLA than for AVISO but the correlation coefficient is better. I can understand that the bias might affect RMSE but it should not affect the Stdv?
There is a question as to whether Table 3 is informative in that agreement with CTD-derived geostrophic current is not necessarily agreement with the true current.

Line 196. Referee 1 questioned the 6-year averaging. You refer to 6 years but the SLA data only span part of 2009 and 2014. Then figure 8 for “2014” will contain some bias opposite to any October-December seasonal signal?

Page 13, figure 5. Where is the arrow showing the “reference current length =0.5m/s”?

Lines 274-275. “amplitude of eddies in the Red Sea is about 4 cm which is an agreement with global values”. Is there any reason to hope for or expect “agreement”. Lower latitudes and relatively narrow Red Sea might lead one to expect smaller eddy amplitudes.

Figure 8. Earlier Reviewer 1 commented “The comparison and error estimation is very qualitative (comparing figures) and in figure 4a&b (the largest area covered) it is impossible to visualize the results.” I think the same now applies to the figures in response to Referee 1 comment [4] and to new figure 8. Perhaps the revision should include some statistics about the difference between 2010 and climatology. Or preferably, what is the standard deviation of SLA in the panels of figure 8 compared with the standard deviation of the “6-year mean”?
Figure 8 arrows. Do these show direction only (they seem to all have the same length)?

Referee 2 comment “it would be important in my opinion to add informations on where it provides new informations, particularly along the coasts.” You have many statements about geostrophic flows, including along the coast, and about cyclonic and anticyclonic eddies. However, you also have many statements that what you find agrees with previous findings. I think the referee wants you to say what you have found that was NOT previously found. I want this too as an editor of Ocean Science which is for publishing new knowledge and understanding.

Thank-you again for your revised manuscript. In the following, I copy “Referee 2 comments” (which you may have seen) but have added in some of my own and some which attempt to represent the original concerns of Referee 1 which still need more attention in my opinion. (Unfortunately I do not have a re-review from Referee 1).

“In this article, the authors use altimetric data extended toward the coast to describe the monthly climatological circulation in the Red Sea. This study provides a spatial configuration of the mesoscale eddies in the basin and their evolution along a climatological year. The authors also discuss the interannual variability of this surface circulation.
After taking into account the different reviews, the overall quality of the manuscript has improved. Particularly the data and method and the validation sections, where the authors provided more informations on their calculations as well as comparisons with in-situ observations.
The structure of the paper and its aim is clear, but in my opinion some figures remain unclear and the text still contains some grammatical errors. I recommend publication after minor revisions.

Specific comments
FIGURES
Figure 4 : Change the last colorbar to have matching fonts. I also suggest changing the maximal and minimal value of the colobar as figures 4c to 4f only exhibit one color.
Figure 5 and 6 : I still think these figures are not readable which is a shame as they show the main result of the paper. For each month, only a color is exhibited, arrows are quite small, and for the months of June through September, the « AE » and « CE » fonts are not readable over a yellow color. As the SLA can be centered at 0, a « balance colorbar » (as the one found here :
https://matplotlib.org/cmocean/) could help improve greatly the figure.” I think March and April are the wrong way around (either in layout or labelling).
“Figure 7 : Same remark as the previous figure for the axis of the colorbars.” This could be outside the panel border if that helps.

“I provide hereafter some grammatical corrections for the english. But as non-native English speaker I do not think it covers everything in the paper. I suggest having a thorough review of the english in the text.
ABSTRACT”
Line 10: delete initial “The”.
Lines 10-11. “Due to this,” -> “Hence”
Line 13. “. . data set . .”
Line 16. “patterns”
“Line 19 : I suggest changing the end of the sentence by « except along the eastern coast where they flow in the opposite direction».
Line 21 : « except for a small area near the eastern coast ».
Line 21 : « This flow is modified by the presence of ».”
Line 26. Omit “while in winter both cyclonic and anticyclonic eddies are present” (unnecessary repetition)

“INTRODUCTION
Line 31 : « between the continents ».
Line 32 : the words « latitude » and « longitude » are not necessary.
Line 52 : « that are often present ».
Line 54 : « 18°N and 24°N »
Line 70 : « from the coastline ».
Line 84: I suggest « advect » instead of « circulate ».”
Line 90. Delete “coastal region where in the”
Section 2.
Line 132. “. . sea: either . .”
Line 142. “very” -> “vary”
Line 145. Delete “is”
Line 150. “negligibly small”. To answer referee 1, I think you should give a value (of the RMS difference, and compare that with the RMS value of the surface current).

“RESULTS AND DISCUSSION
Line 159 : « on the other hands » should not be used. Change to something like « In contrast,
near the coasts, weak correlation is found between the two datasets, the correlation coefficient being 0.45 to 0.7 ».”
Line 165. Better “. . observed coastal tide gauge data . .”
Lines 174-176. This sentence repeats what is in previous text and Table 2.
“Line 175 : « shows ».”
Line 183. “central” – why this restriction?
Line 189. “in agreement with Bower and Farrar (2015) findings” Is this now for all their data (referee 1 comment)?
“Line 199 : « three cruises ».”
Table 3 and Table 4. What is the meaning of Standard Deviation as distinct from RMSE here? Why of all the values given is Stdv > RMSE only for the top row of Table 3 – is there a mistake here?
“Line 224 : « starts ».”

Section 3.2. Figures 5, 6 show monthly patterns from averaging the 5 or 6 years month by month (as I understand it). Reviewer 1 questioned this. Differences between these monthly figures are only significant if they are greater than the differences between all the Januarys, all the Februarys etc. So I think you also need to show maps of the standard deviations of the January maps, February maps etc., i.e. 12 maps corresponding to Figures 5, 6 which show a measure of the uncertainty of Figures 5, 6.
Line 264. Better “. . October but an anticyclonic eddy forms between . .”
Lines 305-307 is very unclear. Why should the annual and 6-year mean values be the same? How can all the biases be negative if the mean is of all the contributing years? “less” than what?
Lines 316-317. “. . along the middle from 20°N to the north.” seems to partly repeat and partly contradict “The mean flow is toward the north mostly along the western coast and center up to 23°N . . .”
Lines 320-321. Better “. . southern part the flow occupies most of the width of the Red Sea with a narrow northward flow near the eastern coast.”

“CONCLUSION
Line 329 to 331 : This sentence should be removed as it is repeated in the last paragraph.
Line 332 : « shows ».”
Line 333. “. . parts show negligible . .”
“Line 334 : « geostrophic current ».
Line 335-336 : « except along the eastern coast where they flow in the opposite direction ».
Line 337 : « except a small part […] and of the western coast ».
Lines 341-2 : «. . northern Red Sea more than in the southern part. The mean amplitude . .».
Line 341-343 : Please rewrite this sentence.”
Line 347. Omit “than winter”
Line 348. “. . paper is considered . .”

Dear Authors
Thank-you for your revisions. I am still asking you to consider many “Technical Corrections” (see below). Three of these marked * are more serious in that they concern the scientific basis of your findings; you should remember that readers will be able to see all the comments and whether you have properly responded. After this you should enter the manuscript to the Copernicus / Ocean Science production system. It will be copy-edited and you should check that your intended meaning is kept. Thank-you for publishing in Ocean Science.
Yours sincerely
John Huthnance

Technical corrections.
In many places you express a latitude range using “between”. This should always be “between m°N and n°N”, i.e. always “between . . and . .” This applies in lines 116, 118, 194, 213 (twice), 214, 215, 218, 219, 224, 231, 237, 238, 239, 246, 249, 250, 262, 263, 344 and perhaps others, please check.

Line 22. Delete “the”
Line 47. Delete “S.”
Line 48. “. . on a modelling approach”
Line 61. “sea level” (lower case “s”)
Lines 65-67. Better “. . satellite altimetry which offers large coverage . . SSH (hence sea-level anomaly – SLA), wave height . .”
Line 68. Delete “the”
Line 78. “. . (1) echo interference . . as inland water” (delete two “the”)
Line 89. “. . important in densely . .”
Line 90. “. . regions, . .”

Line 98. “. . Jason-2 altimetry along . .”
*Line 98. “SLA” Anomaly relative to what? For example, is it relative to the mean of the whole 6-year period, or are July anomalies relative to the average of the 6 Julys, or . . ?
Line 100. “To cover all the period”; what period?
Lines 102-103. “. . removal from SLA of outliers . . from the mean. . .”
Line 127. “calculating” -> “calculate”
Line 129. “. . for comparison . .”
Lines 131-132. Better “. . using interpolation (kriging) to smooth the dataset. The merged data are hereafter called FSM-SLA.”?
Line 137. “is using” -> “uses”
Line 141. “. . exceed 500”
Line 142. “. . as the level . .”
Line 144. “shown” -> “showed”
Line 145. “is” -> “are”
Line 148. “level” –> “levels”
Lines 149. “. . with root-mean-square error (RMSE) around . .”
Line 152. Omit “show”

Line 156. Delete first “The”
Lines 161-162. Can now reduce to “. . where the RMSE is less . .”
Line 164. “. . data comparison with observed . .”
Table 2 Note. “. . comparisons is . .”
Line 177. “. . SLA compared with . .”
Figure 4 fonts. “October 2011” should match “March 2010” and “April 2011”. April longitudes should match March and October longitudes. October colour scale values should match March and April colour scale values. This is for good appearance!
Line 180. Better “. . FSM-SLA. Green . .”
Line 184. “. . timing of the three cruises . .”
Line 187. “. . FSM-SLA with geostrophic currents from CTD data . .”
Line 189. “. . currents near the coast estimated from FSM-SLA match”
Line 191. “. . AVISO do not always match CTD-derived currents, especially . .”
Line 195. “. . while the AVISO . .”
Lines 200-201. “. . months March 2010, April 2011 and October 2011 shows . .”
Lines 208-209. Better “. . current along the eastern coast of the Red Sea is northward while along the western coast it is southward. . .”
Line 212. “2002).” Here you start descriptions for each month. Like the others, January should start with a new paragraph “Fig.5 . .” (omit “The”).
*Line 212. You need to say something here related to your response about the small standard deviations between the Januarys, between the Februarys etc. as shown in the figure you put in your response. Perhaps an overall standard deviation (average over all months and locations) and locations and magnitude of the largest standard deviations.
Line 231. “. . 24° – 25°N and 20° – 22°N, . .” (delete first “,”)
Line 232. Delete first “,”.
Line 232. “. . seen that several . .”
Line 233. “. . eddies are distributed . .”
Line 235. Delete “the flow of”
Line 235. “. . reversed their direction”
Line 236. “The accompanies formation of a large . .”
Lines 238-239. “. . strength as in May. . .”
Lines 240-241. “. . following the normal . .”

Line 245. Delete “the flow of”
Line 249. “. . Red Sea. Fig. 6 . .”
Line 254. “. . flow. Consequently . .”
Line 256. “. . with a cyclonic eddy, . .”
Line 257. “. . season. Summer . .”
Line 264. “. . towards the south . .”
Line 267. “. . currents is similar . .”
Line 271. “. . Red Sea, with anticyclonic”
Lines 272-273. “. . wind and thermohaline forces (Neumann . .”
Line 275. “. . modified by the presence . .”
Line 276. “. . eddies; identification of eddies in the study area was conducted”
Lines 277-279. Better “Figure 7 shows latitudinal variability . . . centre of the eddies for 6 years. . .”
Lines 280-281. Better “. . stronger than at other latitudes. The amplitude . . defined as the difference between . .”
Lines 283-285. “interior. The mean amplitude for an anticyclonic eddy is between 1.3 cm in the southern Red Sea and 5.3 cm in the northern Red Sea and for a cyclonic eddy is between 1.6 cm in the southern Red Sea and 4.2 cm in the northern Red Sea. . .”
Line 289. “. . about 5-10 cm/s, but reaches three-times greater in the 25°-26°N . .”
Line 290. “. . anticyclonic eddies. These results match those observed in a previous study by Zhan . .”
Line 292. “. . anticyclonic eddies (left”

*Lines 294, 295, 307. “6-yr mean”. You do not have 6 whole years. You must define exactly what mean you are using.
Line 306. “The bias is very small.”
Line 307. “. . compared with deviation . .” But deviation of what? You must define the variable whose deviation is given in the table.
Line 315. “Figure 9 shows a general schematic . .”
Line 316. “. . north over most of the width of the Red”
Line 317-318. “. . addition, southward geostrophic currents . .”
Line 321. “while in the southern . .”
Line 331. Delete “flow”
Line 334. “. . along a small part . .”
Line 335. “. . study, a northward . .”
Line 337. “Cyclonic eddies were relatively larger than anticyclonic eddies . .”
Lines 338-339. “. . southern part. Anticyclonic . . small mean amplitude and”
Line 340. “. . larger mean amplitude. . .”
Line 341. “it is” -> “they are”
Line 342. “transferring”
Lines 344-345. “. . 27.5°N. During summer cyclonic . .”
Line 346. “. . polarities were observed . .”
Line 347. “besides”

Line 350. “. . providers: JPL . .”
Line 353. “. . Red Sea. The authors thank the . .”