Impacts of the Tropical Cyclone Idai in Mozambique: A Multi-Temporal Landsat Satellite Imagery Analysis

Round 1

Reviewer 1 Report

The major weakness of the paper was the English grammar that made it slow to read and understand. I did edit some of the most obvious sentences that made understanding the paper more difficult. I hope that I can submit the edited pdf file. 

I did find a very confusing discussion in section 2.7 Vegetation Indices...

The presentation of the channels used for Red and NIR was not correct. 

They assigned the red wavelength region to the NIR variable and vice versa. 

Although, I was skeptical of the results based on these errors, it was not obvious that the results were wrong (the NDVI values). I hope the authors can clarify

this confusion for me. 

Comments for author File: Comments.pdf

Author Response

REVIEWER #1 COMMENTS:

R1/Q1. The major weakness of the paper was the English grammar that made it slow to read and understand. I did edit some of the most obvious sentences that made understanding the paper more difficult. I hope that I can submit the edited pdf file.

Response: We recognize that this was an important remark, and we would like to clarify that the language in the manuscript has been extensively checked and revised, particularly by P. Havik who is a native speaker and senior researcher in Topical Science, so as to improve syntax and the text’s general fluidity. The suggestions put forward, particularly by the Reviewer#1 were all taken into account in the review process.

 

R2/Q2.  I did find a very confusing discussion in section 2.7 Vegetation Indices... The presentation of the channels used for Red and NIR was not correct. They assigned the red wavelength region to the NIR variable and vice versa. Although, I was skeptical of the results based on these errors, it was not obvious that the results were wrong (the NDVI values). I hope the authors can clarify this confusion for me.

Response: Thanks for this important remark, it was already revised, and did not affected the results. The revised sentence is: “NIR and RED are the spectral reflectances corresponding to the fourth (0.77-0.90 μm) and third (0.63 – 0.69 μm) Landsat ETM+ bands, respectively. For Landsat 8 OLI, NIR is the fifth band (0.85 – 0.88 μm) and RED is the fourth band (0.64-0.67 μm)” (Line 231-233).

 

The authors would like to thank the Reviewers for their detailed and insightful comments, and we hope that these improvements have adequately addressed all the concerns.

We look forward to hearing from you at your earliest convenience.

Maria Romeiras, on behalf of all the authors

Reviewer 2 Report

Overall there is some merit to the document. Some of the conclusions are weak, however, and the English can be improved. The supplementary material should either be included in the document or removed from the discussion, and the link provided for the supplementary material was broken/incorrect, so it could not be reviewed. This issue alone merits a “Major Revisions Required”. There are several run-on sentences in the article. A separate reviewer should be required to read the article and improve the English before it is resubmitted.

Throughout document: When referring to Cyclone Idai, cyclone should be capitalized. When referring to Sofala Province, province should be capitalized. When referring to a cardinal direction as a place (i.e. the Northeast), it should be capitalized (also for Central) because you are designating a definite region. If you are just referring to a direction, it should be in lower case.

Specific edits

Line 22-25: Run on sentence. Break it into two sentences.

Line 24: Replace “landed” with “made landfall”.

Line 31: Spell out NDVI (as you did with LULC).

Line 57: Make “cyclone” plural. This is in other places in the article also.

Line 56-62: Run on sentence. Too many “and”s.

Line 60: Move Idai before “in March”.

Line 61: Combine all values into one. No need to list all individually.

Line 62: Remove sentence “As a cyclone… on 25 April 2019.” It is not needed and doesn’t provide any information.

Line 65: Change “Idai hit the most Sofala Province, affected surrounding…” to “Idai primarily affected Sofala Province and also impacted the surrounding…”. This is an example of poor English that needs to be corrected.

Line 70: “land use and land cover” need to be capitalized, as it is in the Abstract, since it is a formal product being measured.

Line 74: add “the” before USA and add reference to “Gulf of Mexico” if one exists (as you have with all other regions) or remove it.

Line 77-84: First sentence is another run-on sentence.   Remove “Therefore” on line 81 and move “Sofala Province is an understudied area.” to start of paragraph, remove “To our knowledge, except” and start new sentence with “Two studies,…”. Start new sentence with “Our” on line 82.

Line 95: NDVI is “Normalized Difference Vegetation Index”. Capitalize and add Difference.

Line 98: What do you mean by “consistent records”? Consistent spatially? Temporally? Or both? Or is this something completely different?

Line 104: Remove “and decision makers”. Language is too informal for a peer-reviewed article.

Line 111-112: The directions should be in lower case here since they are just referring to directions. “To the east… while to the north and to the south…”. If you said “In the East… while in the North and South” they would be capitalized.

Line 118: Remove “they”

Line 124-125: You can replace “occurring mainly… and April” with “the rainy season” since you have defined it as such previously.

Line 125”: Replace “can” with “primarily”.

Line 126-127: You cite land percentages but they don’t add up to 100%. What are the other ~45% LULC (you can group them together into one if there are many other LULC classes in the remainder)? Also, remove the space between the number and “%” sign. Finally, move “including mangroves” to before “(48.81%).

Line 132: The depression formed OVER the Mozambique Channel or OFF the Mozambique Coast.

Line 133-134: “… depression intensified into Moderate Tropical Storm Idai.” “condition” should be plural since more than one element played a part, or replace with “state”.

Line 135: Replace “into around” with “to”. Replace “Then Idai began to weaken” with “Idai then weakened”.

Line 136: Replace “starts to re-intensify” with “re-intensified”. When did it re-intensify?

Line 139: Replace “caused flood in” with “flooded”.

Line 141-142: Remove “intensification of the” (you cannot say intensification reduced). The sentence “Idai hit the most… as well.” is poor.   Reword to “Idai affected Sofala Province as well as the surrounding Provinces (…) before moving into Zimbabwe.”

Line 148: “plus” should be capitalized.

Line 150: Replace “with” with “possesses” and “were” with “and are”.

Line 154-155: What do you mean by “selected anniversary data”? Please clarify.

Line 163: There should be a colon after “following”.

Line 166: What do you mean by “cleanest”?

Line 167: Where is “distance map” in Figure 2? Distance Buffer analysis?

Line 175: Clarify/reword “to the matching process to a reference image”.   Are you remapping all of the images to a standard projection?

Line 181: Reword/remove “-off data gaps”. You are removing scan line errors using some process that replaces the missing/bad lines with some derived data. What is this derived data? An average? A duplication process?

Line 183-184: Remove “The…”.   Why is this process needed… how does it store and portray the imagery “efficiently”, and why is this necessary?

Line 188: What do you mean by “… contributing in this way to get the really amount… target.”? What is “the really amount”?

Line 199: Replace “built up areas“ with either “built-up areas” or “urban areas”, preferably the latter.

Line 218-227: Are more than one GEE image used for a reference image, or is it an average? How do you compare the LULC images to more than one GEE image (how is the process done)?   What do you mean by “extracted those values” (poor English)? Are you extracting 75 random points from the GEE image to compare with the LULC maps at the same locations to determine the ground cover in the LULC images? Why are you doing this? Why are you also using a 2016 image? Which one are you using as “ground-truth”?   What about values .2-.4 and .6-.8… what are those labeled as? “almost perfect” should probably be “near perfect”… almost is not scientific.

Line 230: “normalized difference vegetation index” should be capitalized.   Remove “measure vegetation to”.

Line 236: What is a “spectral reflectance stand”?

Line 237: Are the Landsat 8 channels the same wavelengths?

Line 256: “Worth to mention” should be replace with “It is worth mentioning”. Poor English. Also add “equate to” before “higher damage” (and remove the).

Line 261-263: “multiple ring buffer” is not a good description of what you are measuring. They are not rings. Swath, strip, tract, area, or band are better words. Perhaps “distance bands”.

Line 276: Again, not sure what the reference data is here… the GEE image(s) or the 2016 high-resolution image. What are “producer’s” and “user’s” accuracy? Please explain briefly so we know what you are doing (don’t assume the reader knows what you mean).

Like 316: Change “before” to “pre” or “post” to “after” to be consistent.

Line 319: Reword to “Significant NDVI change was observed over the entire Province, with the highest LULC changes observed… Sofala Province.”

Figure 4: Remove “Where:” or change to “… April 2019), where …”

Figure 5: Values on y-axis are actually negative values, and the orange dots don’t correlate to any values in Table 4 (for example, for Dense Vegetation, the delta NDVI value is -58.9 in Table 4, but plotted as ~65 in Figure 5). Is Figure 5 even needed?

Figure 6: Remove 0 values at start (at NDVI label on x-axis). Change x-axis labels to years since they are all April values. You should label the values in the Key, and why are Low values encompassing two separate ranges (0.1-0.2 and 0.2-0.3… why not 0.1 to 0.3; same for 0.3-0.5)? Why not combine them in Figure 6 and Table 5? If there is a definite reason for 0.1 bins for separate Low and Moderate Productivity, but a 0.3 range for High, then state why.

Table 5: Why do you use commas here but periods everywhere else in the article? Also, why use tenths decimal places for some column values and hundredths in others (and why carry out to six decimal places in the % column? Definitely not needed)? Be consistent.

Line 341: Supplementary materials should either be included in the article or removed entirely. The last thing I want to do if I am reading a paper is have to go hunt down additional information.

Line 379: Change “shown” to “show” or “exhibit”. Also change “to” to “from”.

Figure 7: Should NDVI be NDVI% (as mentioned in Line 377)?

Figure 8: The values on the y-axis are actually negative values of NDVI%. Values in the Key should also be NDVI%.

Line 431: Should NDVI be deltaNDVI?

Line 450-454. Run-on sentence. Break up into two if you can.

Line 457: What is the “costal line”?

Line 458: Remove “and not only”.

Line 462-464: Not sure what point you are trying to make with this concluding sentence. Seems out of place.

Line 473: What are the “three high and larger crowns”? Explain what they are and why they are more prone to damage?

Line 479: “Grow” should be “growth”.

Line 482: “occupy a greater area” than what? Occupy the largest % of land in Sofala? Than all other LULC combined?

Line 485: Remove “somehow”.

Line 489-492: Remove first sentence since any discussion/results regarding elevation was not presented. The next sentence can probably just be rewritten as “our results show a strong negative correlation between distance and damage”.

Section 4.3: The entire section seems to be poorly worded. Basically you are stating that the damage not entirely linear compared to distance and that other factors are related to damage, like LULC type. I don’t think areal coverage of (certain) LULC is exactly a strong conclusion, especially without some sort of hard quantification of these claims. Also, it seems like the last two sentences from Line 508-513 are sort of shoe-horned into the paper in order to try and support a conclusion that areal coverage of LULC types might be important. The last sentence makes no sense to me in this context. Overall, stating that damage is related to distance is not a new finding, and then trying to state that the damage in the different regions does not fit the simple linear relationship is fine, but some sort of relationship between LULC and distance and/or damage should be presented or it should be removed.

Line 516: Landsat should be capitalized.

Line 518: Cyclone should be cyclones.

Line 518-523: Sentences are poorly worded. This study can be used as a future reference in related studies and that more groundwork is needed to validate and quantify cyclone damage to LULC using Landsat imagery. The study noted definite changes in LULC using NDVI changes pre and post Cyclone Idai on Sofala and noted a strong correlation between damage and distance from the cyclone path, but more analysis is needed to quantify the local/regional changes that did not fit to the linear model. That is what you can conclude from this study.

Supplementary Materials: Link does not work (obviously xxx was a placeholder, but needed to be updated prior to submission).

Author Response

REVIEWERS' COMMENTS: REVIEWER #2

R2/Q1. Overall there is some merit to the document. Some of the conclusions are weak, however, and the English can be improved. The supplementary material should either be included in the document or removed from the discussion, and the link provided for the supplementary material was broken/incorrect, so it could not be reviewed. This issue alone merits a “Major Revisions Required”. There are several run-on sentences in the article. A separate reviewer should be required to read the article and improve the English before it is resubmitted.

Response: We greatly appreciated your constructive and helpful comments. We have carefully considered this comment and the whole paper has been thoroughly revised. There is no more supplementary material. The Table S1 was included in the document and on the revised version it is Table 5, whereas Figure S1 and Figure S2 were discarded. The English grammar was deeply revised.

 

R2/Q2. Throughout document: When referring to Cyclone Idai, cyclone should be capitalized. When referring to Sofala Province, province should be capitalized. When referring to a cardinal direction as a place (i.e. the Northeast), it should be capitalized (also for Central) because you are designating a definite region. If you are just referring to a direction, it should be in lower case.

Response: Thanks, it has been revised accordingly.

 

R2/Q3. Line 22-25: Run on sentence. Break it into two sentences.

Response: Thanks, it has been revised (Lines 22-25).

 

R2/Q4. Line 24: Replace “landed” with “made landfall”.

Response: Thanks, it has been revised (Lines 23-25).

 

R2/Q5. Line 31: Spell out NDVI (as you did with LULC).

Response: Thanks, it has been spelled out (Lines 31-32).

 

R2/Q6. Line 57: Make “cyclone” plural. This is in other places in the article also

Response: Thanks, it has been revised (Line 55).

 

R2/Q7. Line 56-62: Run on sentence. Too many “and”s.

Response: Thanks, it has been revised (Lines 55-59).

 

R2/Q8Line 60: Move Idai before “in March”.

Response: Thanks, it has been moved (Line 59).

 

R2/Q9. Line 61: Combine all values into one. No need to list all individually.

Response: Thanks, it has been revised. This statement was removed from the original sentence

“…as follow: 900,000, 200,000, 80,000, 650,000, 23,000, and more than 1.5 million, respectively” (Lines 57-59).

 

R2/Q10. Line 62: Remove sentence “As a cyclone… on 25 April 2019.” It is not needed and doesn’t provide any information.

Response: The following statement has been removed “As a cyclone-prone area, the country is still reeling from Cyclone Idai, a new devastating cyclone called “Kenneth”, category four, hit the Northern Provinces on 25 April 2019.” (Lines 59-60).

 

R2/Q11. Line 65: Change “Idai hit the most Sofala Province, affected surrounding…” to “Idai primarily affected Sofala Province and also impacted the surrounding…”. This is an example of poor English that needs to be corrected.

Response: Thanks, it has been revised accordingly (Lines 61-63).

 

R2/Q12. Line 70: “land use and land cover” need to be capitalized, as it is in the Abstract, since it is a formal product being measured.

Response: It has been capitalized (Lines: 66-67).

 

R2/Q13. Line 74: add “the” before USA and add reference to “Gulf of Mexico” if one exists (as you have with all other regions) or remove it.

Response: We added “the” to USA.  The Gulf of Mexico and Caribbean/Mesoamerican Region are regions analyzed in the same study done by Taillie et al [17] (Lines 71-72).

 

R2/Q14. Line 77-84: First sentence is another run-on sentence. Remove “Therefore” on line 81 and move “Sofala Province is an understudied area.” to start of paragraph, remove “To our knowledge, except” and start new sentence with “Two studies,…”. Start new sentence with “Our” on line 82.

Response: It has been revised (Lines 74-79).

 

R2/Q15. Line 95: NDVI is “Normalized Difference Vegetation Index”. Capitalize and add Difference.

Response: It has been revised (Line 89).

 

R2/Q16. Line 98: What do you mean by “consistent records”? Consistent spatially? Temporally? Or both? Or is this something completely different?

Response: The “consistent records” has been removed (Lines 91-92).

 

 R2/Q17. Line 104: Remove “and decision makers”. Language is too informal for a peer-reviewed article.

Response: It has been removed (Lines 97-101).

 

R2/Q18. Line 111-112: The directions should be in lower case here since they are just referring to directions. “To the east… while to the north and to the south…”. If you said “In the East… while in the North and South” they would be capitalized.

Response: It has been revised accordingly in whole manuscript.

 

 R2/Q19. Line 118: Remove “they”

Response: The sentence has been revised and there is no more “they” (Lines 111-113).

 

R2/Q20. Line 124-125: You can replace “occurring mainly… and April” with “the rainy season” since you have defined it as such previously.

Response: It has been revised (Line 117-118).

 

R2/21. Line 125”: Replace “can” with “primarily”.

Response: It has been revised accordingly (Line 118).

 

R2/Q22. Line 126-127: You cite land percentages but they don’t add up to 100%. What are the other ~45% LULC (you can group them together into one if there are many other LULC classes in the remainder)? Also, remove the space between the number and “%” sign. Finally, move “including mangroves” to before “(48.81%).

Response: It has been revised accordingly (Lines 118-121).

 

R2/Q23. Line 132: The depression formed OVER the Mozambique Channel or OFF the Mozambique Coast.

Response: It has been revised (Line 126).

 

R2/Q24. Line 133-134: “… depression intensified into Moderate Tropical Storm Idai.” “condition” should be plural since more than one element played a part, or replace with “state”.

Response: It has been revised (Lines 128).

 

R2/Q25. Line 135: Replace “into around” with “to”. Replace “Then Idai began to weaken” with “Idai then weakened”.

Response: It has been revised (Line 129).

 

 R2/Q26. Line 136: Replace “starts to re-intensify” with “re-intensified”. When did it re-intensify?

Response: It has been revised. The storm started to re-intensify on 13 March 2019 (Line 130).

 

R2/Q27. Line 139: Replace “caused flood in” with “flooded”.

Response: It has been revised (Line 135).

 

R2/Q28.  Line 141-142: Remove “intensification of the” (you cannot say intensification reduced). The sentence “Idai hit the most… as well.” is poor. Reword to “Idai affected Sofala Province as well as the surrounding Provinces (…) before moving into Zimbabwe.”

Response: It has been revised (Lines 136-138).

 

R2/Q29. Line 148: “plus” should be capitalized.

Response:  It has been revised accordingly (Line 151).

 

R2/Q30. Line 150: Replace “with” with “possesses” and “were” with “and are”.

Response: It has been revised (Line 153).

 

R2/Q31. Line 154-155: What do you mean by “selected anniversary data”? Please clarify.

Response: It has been removed (Line 157).

 

R2/Q32. Line 163: There should be a colon after “following”.

Response: It has been revised (Lines 162-163).

 

R2/Q33. Line 166: What do you mean by “cleanest”?

Response: It has been removed and the sentence has been revised (Lines 165-166).

 

R2/Q34. Line 167: Where is “distance map” in Figure 2? Distance Buffer analysis?

Response: Yes, it is the distance bands map obtained using buffer analysis (Line 166).

 

R2/Q35. Line 175: Clarify/reword “to the matching process to a reference image”. Are you remapping all of the images to a standard projection?

Response: Thank you. It has been reworded (Lines 173-174)”.

 

R2/Q36. Line 181: Reword/remove “-off data gaps”. You are removing scan line errors using some process that replaces the missing/bad lines with some derived data. What is this derived data? An average? A duplication process?

Response: It has been reworded. The “off data gaps” into “off gap function” suggested by USGS.  We used Band-specific gap mask files are included with every Landsat 7 Scan Line Corrector (SLC)-off Level-1 data product. These ancillary data identify the location of all pixels affected by the original data gaps in the primary SLC-off scene. The gap mask is provided as a series of individual band files, in compressed (GZIP) GeoTIFF format (Lines 177-182).

 

R2/37. Line 183-184: Remove “The…”. Why is this process needed… how does it store and portray the imagery “efficiently”, and why is this necessary?

Response:  We removed the previous statement “The radiometric calibration is always required for analysis of LULC using Landsat imageries as this correction helps to store and portray satellite imageries efficiently”. We added some detail (in blue) in the following statement: “The ERDAS… with the help of band-specific gap mask files, made with the Landsat 7 Level-1 data product. These mask files help to classify the location of every pixel affected by the original data gaps in the primary SLC-off scene” (Line 179-182).

 

R2/38. Line 188: What do you mean by “… contributing in this way to get the really amount… target.”? What is “the really amount”?

Response: We removed “… contributing in this way to get the really amount… target”. We revised the correspondent sentence (Lines 184-186).

 

R2/39. Line 199: Replace “built up areas“ with either “built-up areas” or “urban areas”, preferably the latter.

Response: It has been revised in whole document.

 

R2/Q40a. Line 218-227: Are more than one GEE image used for a reference image, or is it an average?

Response:  We used one GEE image from 2019 (it is one imagery but it’s an average of the year provided by GEE) to check the accuracy of LULC from 2019 (Lines 214-223).

 

R2/ Q40b. Line 218-227: How do you compare the LULC images to more than one GEE image (how is the process done)?

Response:  We used one image from 2019 GEE to compare with our LULC of April 2019. Seventy-five random pixels were generated. This pixels were automatically labeled using the classified LULC data and labeled through visual inspection using the 2019 GEE image. Then, the LULC values were identified for the same pixels in the referenced image (GEE image) and compared with the LULC values of classified image. We employed the kappa coefficient as the accuracy indicator (Lines 214-223).

 

R2/ Q40c. Line 218-227:  What do you mean by “extracted those values” (poor English)?

Response:  We removed “extracted those values” from the previous sentence. We produced 75 random points from the classified LULC (Lines 214-223). English has been revised in whole document.

 

R2/Q40d. Line 218-227:  Are you extracting 75 random points from the GEE image to compare with the LULC maps at the same locations to determine the ground cover in the LULC images?

Response:  Yes, we extracted 75 random points and compared the values of this points in generated LULC with the GEE image land cover (Lines 214-223).

 

R2/Q40e. Line 218-227:  Why are you doing this? Why are you also using a 2016 image? Which one are you using as “groundtruth”? What about values .2-.4 and .6-.8… what are those labeled as?

Response:  We removed this line which is not necessary (Lines 214-223). Ground truth are is the GEE image for 2019.

 

R2/Q40f. Line 218-227:  … “almost perfect” should probably be “near perfect”… almost is not scientific.

Response: Thank you. Yes, it should be “near perfect”, we changed accordingly (Line 222).

 

R2/Q41. Line 230: “normalized difference vegetation index” should be capitalized. Remove “measure vegetation to”.

Response: It has been revised accordingly (Line 226).

 

R2/Q42.  Line 236: What is a “spectral reflectance stand”?

Response: The “stand” has been removed and the sentence has been revised (Lines 231-233).

 

R2/Q43. Line 237: Are the Landsat 8 channels the same wavelengths?

Response: NIR and RED are the spectral reflectances corresponding to the fourth (0.77-0.90 μm) and third (0.63 – 0.69 μm) Landsat ETM+ bands, respectively. For Landsat 8 OLI, NIR is the fifth band (0.85 – 0.88 μm) and RED is the fourth band (0.64-0.67 μm)” (Lines 231-233).

 

R2/Q44. Line 256: “Worth to mention” should be replace with “It is worth mentioning”. Poor English. Also add “equate to” before “higher damage” (and remove the).

Response: It has been revised (Line 251).

 

R2/Q45. Line 261-263: “multiple ring buffer” is not a good description of what you are measuring. They are not rings. Swath, strip, tract, area, or band are better words. Perhaps “distance bands”.

Response: Thank you. It has been revised (Lines 255-257).

 

R2/Q46a. Line 276: Again, not sure what the reference data is here… the GEE image(s) or the 2016 high-resolution image.

Response: The GEE images. We made changes accordingly (Lines 267-269).

 

R2/Q46b. Line 276: What are “producer’s” and “user’s” accuracy? Please explain briefly so we know what you are doing (don’t assume the reader knows what you mean).

Response: Producer’s accuracy expresses how often are real features on the ground correctly shown on the LULC map and user’s accuracy show how often the class on the LULC map will actually be present on the ground (Lines 269-271).

R2/Q47. Line 316: Change “before” to “pre” or “post” to “after” to be consistent.

Response: Thank you. It has been revised (Lines 311-313). Now its is consistently used “pre-cyclone” and “post-cyclone” for the scenario before and after cyclone event, respectively.

 

R2/Q48. Line 319: Reword to “Significant NDVI change was observed over the entire Province, with the highest LULC changes observed… Sofala Province.”

Response: Thank you. It has been revised (Lines 316-318).

 

R2/Q49. Figure 4: Remove “Where:” or change to “… April 2019), where …”

Response: It has been revised.

 

R2/Q50. Figure 5: Values on y-axis are actually negative values, and the orange dots don’t correlate to any values in Table 4 (for example, for Dense Vegetation, the delta NDVI value is -58.9 in Table 4, but plotted as ~65 in Figure 5). Is Figure 5 even needed?

Response: Thank you. The Figure 5 has been removed.

 

R2/Q51. Figure 6: Remove 0 values at start (at NDVI label on x-axis). Change x-axis labels to years since they are all April values. You should label the values in the Key, and why are Low values encompassing two separate ranges (0.1-0.2 and 0.2-0.3… why not 0.1 to 0.3; same for 0.3-0.5)? Why not combine them in Figure 6 and Table 5? If there is a definite reason for 0.1 bins for separate Low and Moderate Productivity, but a 0.3 range for High, then state why-

Response: The 0 value was removed from the previous figure 6 (at NDVI label on x-axis) and we changed the x-axis labels to years, accordingly. We differentiated the values in detail (more NDVI range classification) to make easy readers understanding, since we classified NDVI range as:

• Low productivity: 1 – 0.2 and 0.2 – 0.3,
• Moderate Productivity: 0.3 – 0.4 and 0.4 – 0.5, and
• High Productivity: 0.5 -0.8.

 

In the NDVI general classification some of the vegetation exists in different range, for example shrubs identified from 0.2 to 0.4 but sometimes 0.2 indicates barren land too. In order to make it clear we differentiate the NDVI value ranges in detail.

0.1 – 0.2  NDVI indicates  empty area, rocks and sands;

0.2 – 0.3  NDVI indicates shrubs, grassland and meadows;

0.3 – 0.4   NDVI indicates shrubs and senescing crops;

0.4 – 0.5  NDVI indicates senescing crops and dense vegetation;

0.5 – 0.8  NDVI indicates dense vegetation and forest.

The previous Figure 6 and Table 5 were combined, accordingly. It is now Figure 5.

 

R2/Q52. Table 5: Why do you use commas here but periods everywhere else in the article? Also, why use tenths decimal places for some column values and hundredths in others (and why carry out to six decimal places in the % column? Definitely not needed)? Be consistent.

Response: Thank you. It has been revised accordingly. The previous Table 5 is now an inset table of Figure 5.

 

R2/Q53. Line 341: Supplementary materials should either be included in the article or removed entirely. The last thing I want to do if I am reading a paper is have to go hunt down additional information.

Response: Thank you. The supplementary materials have been removed.  The previous Table S1 was included in the document as Table 5 whereas figure S1 and figure S2 were discarded.

 

R2/Q54. Line 379: Change “shown” to “show” or “exhibit”. Also change “to” to “from”.

Response: Thank you. It has been revised (Lines 373-375). 

 

R2/Q55. Figure 7: Should NDVI be NDVI% (as mentioned in Line 377)?

Response: Thank you. It has been revised (Lines 373-375).

 

 R2/Q56. Figure 8: The values on the y-axis are actually negative values of NDVI%. Values in the Key should also be NDVI%.

Response: Thank you. It has been revised accordingly. It is now figure 7 and the linear correlation between NDVI% and distance to Idai path became positive.    

 

R2/Q57. Line 431: Should NDVI be deltaNDVI?

Response: No. It has been revised (Line 417).

 

R2/Q58. Line 450-454. Run-on sentence. Break up into two if you can.

Response: Thank you. It has been revised (Lines 434-439).  

 

R2/Q59. Line 457: What is the “costal line”?

Response: Typing error. It should be coastline (Line 443).  

 

R2/Q60. Line 458: Remove “and not only”.

Response: Thank you. It has been removed (Line 443).  

 

R2/Q51. Line 462-464: Not sure what point you are trying to make with this concluding sentence. Seems out of place.

Response: Thank you. We have removed the following statement : “Therefore, whole Sofala coastal area including highlighted regions (“NE”, “C”, and “S”) are riverine swamp and floodplain characterized by predominance of mangroves as one of the integral ecosystems of wetlands” (Line 449)

 

R2/Q52. Line 473: What are the “three high and larger crowns”? Explain what they are and why they are more prone to damage?

Response: Thank you. It should be “higher tree layer and greater tree canopy cover” (Line 458). Numerous studies have reported that forests with higher tree height and larger tree crown/canopy cover are more prone to the physical damage by a tropical cyclone than the barren areas and herbaceous or sparse vegetation (see the following researches [67-69], for further details).

 

R2/Q53.  Line 479: “Grow” should be “growth”.

Response: Thank you. It has been revised (Line 462).  

 

R2/Q54. Line 482: “occupy a greater area” than what? Occupy the largest % of land in Sofala? Than all other LULC combined?

Response: Dense vegetation (40,854.4 km²) occupy the greatest area in the whole entire Sofala Province, followed by wetland vegetation (7,294.20 km²), and shrub land (9,963.52 km²). Following these three woody vegetations, agricultural land is in the fourth position with 5,134.84 km². Therefore, after your suggestion we ignored to use areas occupied by each LULC type as an important damage influencing factor.

 

R2/Q55. Line 485: Remove “somehow”

Response: Thank you. It has been removed (Line:467).  

 

R2/Q56. Line 489-492: Remove first sentence since any discussion/results regarding elevation was not presented. The next sentence can probably just be rewritten as “our results show a strong negative correlation between distance and damage”.

Response: The first sentence has been removed and the next sentence has been revised (Line 471).   

 

R2/Q57. Section 4.3: The entire section seems to be poorly worded. Basically you are stating that the damage not entirely linear compared to distance and that other factors are related to damage, like LULC type. I don’t think areal coverage of (certain) LULC is exactly a strong conclusion, especially without some sort of hard quantification of these claims. Also, it seems like the last two sentences from Line 508-513 are sort of shoe-horned into the paper in order to try and support a conclusion that areal coverage of LULC types might be important. The last sentence makes no sense to me in this context. Overall, stating that damage is related to distance is not a new finding, and then trying to state that the damage in the different regions does not fit the simple linear relationship is fine, but some sort of relationship between LULC and distance and/or damage should be presented, or it should be removed.

Response: Thank you. The section 4.3 has been deeply revised. We discarded areal coverage by each LULC type as an important damage influencing factor (Lines 471-486).

 

R2/Q58. Line 516: Landsat should be capitalized.

Response: Thanks. It has been revised accordingly (Line 480).

 

R2/Q59. Line 518: Cyclone should be cyclones.

Response: Thanks. It has been revised accordingly (Line 482).

 

R2/Q60. Line 518-523: Sentences are poorly worded. This study can be used as a future reference in related studies and that more groundwork is needed to validate and quantify cyclone damage to LULC using Landsat imagery. The study noted definite changes in LULC using NDVI changes pre and post Cyclone Idai on Sofala and noted a strong correlation between damage and distance from the cyclone path, but more analysis is needed to quantify the local/regional changes that did not fit to the linear model. That is what you can conclude from this study.

Response: Thank you very much for your suggestion.  It has been revised (Lines 482-486).

 

R2/Q61. Supplementary Materials: Link does not work (obviously xxx was a placeholder, but needed to be updated prior to submission).

Response: Thank you. It has been revised. The Table S1 was included in the document as Table 5 whereas Figure S1 and Figure S2 were discarded.

 

The authors would like to thank the Reviewers for their detailed and insightful comments, and we hope that these improvements have adequately addressed all the concerns.

We look forward to hearing from you at your earliest convenience.

Maria Romeiras, on behalf of all the authors

Round 2

Reviewer 1 Report

The paper has been much improved with the extensive professional English editing that was performed. This improved the paper dramatically. 

 

Also the questions concerning NIR and RED channels was corrected to my satisfaction. 

 

Line 34 might benefit from a change from (NDVI 0.5-0.8)

to (NDVI 0.8 to 0.5) since the values of NDVI decreased over time.

One potential optional change is to incorporate a few more details of the NDVI results in the abstract.  

Reviewer 2 Report

Authors have sufficiently edited the document and provided sufficient modifications to rectify issues cited previously.