Assessing land elevation in the Ayeyarwady Delta (Myanmar) and its relevance for studying sea level rise and delta flooding

Seeger, Katharina; Minderhoud, Philip Simon Johannes; Peffeköver, Andreas; Vogel, Anissa; Brückner, Helmut; Kraas, Frauke; Nay Win Oo,; Brill, Dominik

doi:https://doi.org/10.5194/egusphere-2022-1425

Preprints

https://doi.org/10.5194/egusphere-2022-1425

Preprints

04 Jan 2023

| 04 Jan 2023

Assessing land elevation in the Ayeyarwady Delta (Myanmar) and its relevance for studying sea level rise and delta flooding

Katharina Seeger, Philip Simon Johannes Minderhoud, Andreas Peffeköver, Anissa Vogel, Helmut Brückner, Frauke Kraas, Nay Win Oo, and Dominik Brill

Abstract. With their low lying, flat topography, river deltas and coastal plains are extremely prone to relative sea level rise and other water related hazards. This calls for accurate elevation data for flood risk assessments, especially in the densely populated Southeast Asian deltas. However, in data-poor countries such as Myanmar, where high accuracy elevation data is not accessible, often only global satellite based digital elevation models (DEMs), suffering from low vertical accuracy and remote sensing artefacts, can be used by the public and scientific community. As the lack of accurate elevation data hampers the assessment of flood risk, studying available information on land elevation and its reliability is essential, particularly in the context of sea level rise impact. Here, we assess the performance of ten global DEMs in the Ayeyarwady Delta (Myanmar) against the new, local, so called AD-DEM, which was generated based on topographical map elevation data. To enable comparison, all DEMs were converted to a common vertical datum tied to local sea level. While both CoastalDEM v2.1 and FABDEM, perform comparably well, showing the highest correspondence in comparison with AD-DEM and low elevation spot heights, the FABDEM outperforms the CoastalDEM v2.1 by the absence of remote sensing artefacts. The AD-DEM provides a high accuracy, open source and freely available, independent elevation dataset suitable for evaluating land elevation data in the Ayeyarwady Delta and studying topography and flood risk at large scale, while small scale investigations may benefit from a FABDEM locally improved with data from the AD-DEM.

Based on latest IPCC projections of sea level rise, the consequences of DEM selection for assessing the impact of sea level rise in the Ayeyarwady Delta are shown. We highlight the need for addressing particularly low lying populated areas within the most seaward districts with risk mitigation and adaptation strategies while also more inland delta population should be made aware to face a higher risk of flooding due to relative sea level rise in the next ~100 years.

Received: 08 Dec 2022 – Discussion started: 04 Jan 2023

Download & links

Preprint (PDF, 4040 KB)

Supplement (9055 KB)

Katharina Seeger et al.

Status: closed

RC1:
'Comment on egusphere-2022-1425', Anonymous Referee #1, 08 Feb 2023

Review of egusphere-2022-1425

Assessing land elevation in the Ayeyarwady Delta (Myanmar) and its

relevance for studying sea level rise and delta flooding

by

Katharina Seeger et al.

In this paper, the authors compare a DEM build based on information from toporgaphic maps with global DEMs such as ASTER, SRTM, ALOS, etc. The study area is a large delta in Myanmar. Besides comparing the DEMs, the authors also run sea-level change simulations.
Although the manuscript felt a bit long to read, I think that the methods are sound, and all analysis were done correctly. I only have a few minor points I'd like to mention/disscus.
abstrac, line 25: I wouldn't use "open source" here, but just "open", as open source feels more appropriate for software code (the source code that is open).
usage of colors in figures. I feel that the figures have quite a lot of information, and showing plots with several categories represented only by colors, is always a challenge. In figure 5, for instance, is very difficul to identify the lines for each DEM. In the graphical abstract, we have 10 (?) lines all with the same symbols, and I have doubts if people with color-impaired vision will be able to see all the lines. I suggest the authors to re-think these stilistic choices.
In fig.5, also try to rename the sub-figures so they have the same letter as the profile, because "fig.5D, profile C", etc is a bit confusing. In my opinion, the enlaged part of profile A does not help much...
In figure 1, I would rather see a larger version of the DEM (maybe with a shaded relief as texture), as large as the page, because that is very important. The insets can be repositioned, the colorscale could be continuous (and use less space), and the map of points could be a small map of point density, since that its the message (and you have a version of that map in the supplemental material). The larger DEM should also bring the names of places mentioned in the text, as we readers are usually not so familiarized with the area.
Another choice that I questioned as I was reading was to show ACE2 results instead of, say, ALOS AW3D30. It seems to me that AW3D30 has a much larger user community than ACE2.
In figure 3, we can see some major differences between the DEMs, but not the artifacts mentioned in the text. A slope map would be much better for this.
abouth the methods:

the authors could be clearer on why they digitized point heights for the lower areas and contours for the higher areas. From fig.1, I can see that, being a very flat area, there are not much more topographical information in the maps besides spot height for these low-lying areas, but that is not very clear in the text.
in line 180 you say: "From these, points were extracted at 250 m and 2000 m intervals, respectively" what do you mean? I don't understand exactly what these two values represent. Does it mean that the distance between vertices of the digitized lines are 250m apart? why? what is the reationale behind this? I see that you created DEMs from these two datasets, but I still don't see the reason for them (I would have done only one version)
line 189 - what was the resolution? say it here and don't leave your reader wondering.
lines 300-306 - the details mentioned in the text can't be easily seen on the map of fig.1. you need a larger map for that
line 317 you say "whereas topography as represented by AW3D30 is affected by distortions that reflect the sensor swaths" - what sensor swaths? SRTM has those artifacts caused mostly by the mast oscillation, but aw3d30 is built by photogrammetry of alos images.

Citation: https://doi.org/10.5194/egusphere-2022-1425-RC1
- AC1: 'Reply on RC1', Katharina Seeger, 17 Mar 2023
  
  Dear Reviewer,
  Thank you very much for this detailed review of our submission, your positive feedback and helpful comments that will allow for improvement of our manuscript.
  We agree with the vast majority of your suggestions and will integrate them in the revisions. We agree that the text passages you mentioned should be rephrased to clarify their message. For example, this holds true especially for your comments on the method section and the so far improper wording on the artefacts of the AW3D30. We decided to provide results for the ACE2 in the main text instead of the AW3D30 to show the entire range of DEM performance: Compared to the local AD-DEM and all other global DEMs used in this study, ACE2 indicates the lowest elevation. AW3D30 has a similar tendency as SRTM or ASTER, i.e., indicating higher elevation than the local data. However, to be inclusive and serve the international userbase of the AW3D30, we will add the AW3D30 to the other DEMs in the main text and restructure the figures accordingly.
  We are also very grateful for your thorough suggestions on the figures of our submission. Though we checked the figures for their readability by readers with colour vision deficiencies, we agree the graphical abstract and Fig. 5 can be improved. For these figures, we used colours instead of changing line type (dotted, dashed, etc) for consistency as the bar charts in Figs. 7, S22, and S23 require a colour scheme. In the revision, we will address and improve these figures on the issues you have highlighted, for example by a combination of different colours and symbology.
  Further aspects you mentioned to improve the readability of our figures can be easily implemented, e.g., enlarging and adjusting subfigures of Fig. 5, or adding a slope map to provide supplementary information to Fig. 3. Fig. 1 can be improved by enlarging subfigure (a) and adding features like latitudes and longitudes to support our statements of ll. 300–306. We deliberately did not provide further information on deltaic settlements in Fig. 1a as otherwise, the most relevant information, i.e., the presentation of the new AD-DEM would not become clear. But with the general revision of Fig. 1, we will consider to add key references and improve in-text referencing to Fig. S25 which contains the location of all settlements (e.g., in ll. 511 ff.).
  Thank you again for your valuable feedback.
  Best regards,
  Katharina Seeger (on behalf of the authors)
  
  Citation: https://doi.org/10.5194/egusphere-2022-1425-AC1
RC2:
'Comment on egusphere-2022-1425', Anonymous Referee #2, 17 Feb 2023
The authors present very relevant work by assessing the quality of elevation maps for a data scarce country like Myanmar and making improved datasets available. It would be helpful if the authors could add the link to Pangea data repository as well as a data when the data will become availble.
In general the research seems sound yet at some points difficult to read. I feel that the work would benefit from a more comprehensive overview of the source data of the different elevation models. This would also allow a more systematic discussion of the artefacts.
The comparison with flood data is interesting yet it is difficult to get a quantitative understanding of the correlation between floods and depressions from Figure 5.
Few other comments/suggestions:
Line 76-79: reference needed

Figure 1c and d: improve readability

Line 213: add date of retrieval

Line 244: repeat reference to sampling technique

Line 250: Why is AW3D excluded?

Line 261: "on investigations" can you explain this more specific?

Line 264: which satellite microwaves are used?

Line 302: what imperfection?

Line 325: How where trees and buildings removed?

Line 341 and further: are the 2 numbers behind the comma a meaningful level of significance?

line 349: is a R2 of 0.85 also good given the spatial correlations/dependencies in the data?

Line 424: what is meant with this sentence?

Line 644: coastal squeeze may be mentioned explicitely
Citation: https://doi.org/10.5194/egusphere-2022-1425-RC2
- AC2: 'Reply on RC2', Katharina Seeger, 17 Mar 2023
  
  Dear Reviewer,
  Thank you very much for reviewing our submission, your positive feedback and helpful comments that will allow for improvement of our manuscript.
  We agree with the vast majority of your suggestions and will integrate them in the revisions. We agree that the text passages you mentioned should be revised to clarify their message and strengthen the argumentation. In case of revision, this will be done by adding references (e.g., in ll. 76–79, l. 244, etc), rephrasing (e.g., in the method and results section) and providing further details, for example on the relation between flooding and topography (section 4.2.4) and the DEM artefacts. We deliberately put the overview of the source data of the global DEMs used in the supplementary material instead of the main text in order not to further extend the length of the manuscript further with general and common DEM information (reviewer 1 commented on length). However, apart from referencing to the supplement, we may integrate some aspects on the DEM source data if they are relevant when discussing the artefacts. We also decided to provide the AW3D30 results in the supplementary material as they show a similar tendency as SRTM or ASTER, i.e., indicating higher elevation than the local data. Instead, we included the ACE2 in the main text as it indicates the lowest elevations. Thus, we can show the entire range of DEM performance without generating an overload of information. However, to be inclusive and serve the international userbase of the AW3D30, we will add the AW3D30 to the other DEMs in the main text and restructure the figures accordingly. Your questions on satellite microwaves (l. 264) and tree and building removal (l. 325) address the processing of published datasets we used; we made reference to their publications. We will check the manuscript to add any additional references when needed.
  We are also very grateful for your thorough suggestions on the figures of our submission. In case of revision, we will adjust Fig. 1, which may also clarify your question on the imperfection of the AD-DEM (l. 302) then.
  Thank you again for your valuable feedback. We will upload the AD-DEM in the Pangaea repository and put the link in the manuscript once we are invited to revise the manuscript.
  Best regards,
  Katharina Seeger (on behalf of the authors)
  
  Citation: https://doi.org/10.5194/egusphere-2022-1425-AC2

Status: closed

RC1:
'Comment on egusphere-2022-1425', Anonymous Referee #1, 08 Feb 2023

Review of egusphere-2022-1425

Assessing land elevation in the Ayeyarwady Delta (Myanmar) and its

relevance for studying sea level rise and delta flooding

by

Katharina Seeger et al.

In this paper, the authors compare a DEM build based on information from toporgaphic maps with global DEMs such as ASTER, SRTM, ALOS, etc. The study area is a large delta in Myanmar. Besides comparing the DEMs, the authors also run sea-level change simulations.
Although the manuscript felt a bit long to read, I think that the methods are sound, and all analysis were done correctly. I only have a few minor points I'd like to mention/disscus.
abstrac, line 25: I wouldn't use "open source" here, but just "open", as open source feels more appropriate for software code (the source code that is open).
usage of colors in figures. I feel that the figures have quite a lot of information, and showing plots with several categories represented only by colors, is always a challenge. In figure 5, for instance, is very difficul to identify the lines for each DEM. In the graphical abstract, we have 10 (?) lines all with the same symbols, and I have doubts if people with color-impaired vision will be able to see all the lines. I suggest the authors to re-think these stilistic choices.
In fig.5, also try to rename the sub-figures so they have the same letter as the profile, because "fig.5D, profile C", etc is a bit confusing. In my opinion, the enlaged part of profile A does not help much...
In figure 1, I would rather see a larger version of the DEM (maybe with a shaded relief as texture), as large as the page, because that is very important. The insets can be repositioned, the colorscale could be continuous (and use less space), and the map of points could be a small map of point density, since that its the message (and you have a version of that map in the supplemental material). The larger DEM should also bring the names of places mentioned in the text, as we readers are usually not so familiarized with the area.
Another choice that I questioned as I was reading was to show ACE2 results instead of, say, ALOS AW3D30. It seems to me that AW3D30 has a much larger user community than ACE2.
In figure 3, we can see some major differences between the DEMs, but not the artifacts mentioned in the text. A slope map would be much better for this.
abouth the methods:

the authors could be clearer on why they digitized point heights for the lower areas and contours for the higher areas. From fig.1, I can see that, being a very flat area, there are not much more topographical information in the maps besides spot height for these low-lying areas, but that is not very clear in the text.
in line 180 you say: "From these, points were extracted at 250 m and 2000 m intervals, respectively" what do you mean? I don't understand exactly what these two values represent. Does it mean that the distance between vertices of the digitized lines are 250m apart? why? what is the reationale behind this? I see that you created DEMs from these two datasets, but I still don't see the reason for them (I would have done only one version)
line 189 - what was the resolution? say it here and don't leave your reader wondering.
lines 300-306 - the details mentioned in the text can't be easily seen on the map of fig.1. you need a larger map for that
line 317 you say "whereas topography as represented by AW3D30 is affected by distortions that reflect the sensor swaths" - what sensor swaths? SRTM has those artifacts caused mostly by the mast oscillation, but aw3d30 is built by photogrammetry of alos images.

Citation: https://doi.org/10.5194/egusphere-2022-1425-RC1
- AC1: 'Reply on RC1', Katharina Seeger, 17 Mar 2023
  
  Dear Reviewer,
  Thank you very much for this detailed review of our submission, your positive feedback and helpful comments that will allow for improvement of our manuscript.
  We agree with the vast majority of your suggestions and will integrate them in the revisions. We agree that the text passages you mentioned should be rephrased to clarify their message. For example, this holds true especially for your comments on the method section and the so far improper wording on the artefacts of the AW3D30. We decided to provide results for the ACE2 in the main text instead of the AW3D30 to show the entire range of DEM performance: Compared to the local AD-DEM and all other global DEMs used in this study, ACE2 indicates the lowest elevation. AW3D30 has a similar tendency as SRTM or ASTER, i.e., indicating higher elevation than the local data. However, to be inclusive and serve the international userbase of the AW3D30, we will add the AW3D30 to the other DEMs in the main text and restructure the figures accordingly.
  We are also very grateful for your thorough suggestions on the figures of our submission. Though we checked the figures for their readability by readers with colour vision deficiencies, we agree the graphical abstract and Fig. 5 can be improved. For these figures, we used colours instead of changing line type (dotted, dashed, etc) for consistency as the bar charts in Figs. 7, S22, and S23 require a colour scheme. In the revision, we will address and improve these figures on the issues you have highlighted, for example by a combination of different colours and symbology.
  Further aspects you mentioned to improve the readability of our figures can be easily implemented, e.g., enlarging and adjusting subfigures of Fig. 5, or adding a slope map to provide supplementary information to Fig. 3. Fig. 1 can be improved by enlarging subfigure (a) and adding features like latitudes and longitudes to support our statements of ll. 300–306. We deliberately did not provide further information on deltaic settlements in Fig. 1a as otherwise, the most relevant information, i.e., the presentation of the new AD-DEM would not become clear. But with the general revision of Fig. 1, we will consider to add key references and improve in-text referencing to Fig. S25 which contains the location of all settlements (e.g., in ll. 511 ff.).
  Thank you again for your valuable feedback.
  Best regards,
  Katharina Seeger (on behalf of the authors)
  
  Citation: https://doi.org/10.5194/egusphere-2022-1425-AC1
RC2:
'Comment on egusphere-2022-1425', Anonymous Referee #2, 17 Feb 2023
The authors present very relevant work by assessing the quality of elevation maps for a data scarce country like Myanmar and making improved datasets available. It would be helpful if the authors could add the link to Pangea data repository as well as a data when the data will become availble.
In general the research seems sound yet at some points difficult to read. I feel that the work would benefit from a more comprehensive overview of the source data of the different elevation models. This would also allow a more systematic discussion of the artefacts.
The comparison with flood data is interesting yet it is difficult to get a quantitative understanding of the correlation between floods and depressions from Figure 5.
Few other comments/suggestions:
Line 76-79: reference needed

Figure 1c and d: improve readability

Line 213: add date of retrieval

Line 244: repeat reference to sampling technique

Line 250: Why is AW3D excluded?

Line 261: "on investigations" can you explain this more specific?

Line 264: which satellite microwaves are used?

Line 302: what imperfection?

Line 325: How where trees and buildings removed?

Line 341 and further: are the 2 numbers behind the comma a meaningful level of significance?

line 349: is a R2 of 0.85 also good given the spatial correlations/dependencies in the data?

Line 424: what is meant with this sentence?

Line 644: coastal squeeze may be mentioned explicitely
Citation: https://doi.org/10.5194/egusphere-2022-1425-RC2
- AC2: 'Reply on RC2', Katharina Seeger, 17 Mar 2023
  
  Dear Reviewer,
  Thank you very much for reviewing our submission, your positive feedback and helpful comments that will allow for improvement of our manuscript.
  We agree with the vast majority of your suggestions and will integrate them in the revisions. We agree that the text passages you mentioned should be revised to clarify their message and strengthen the argumentation. In case of revision, this will be done by adding references (e.g., in ll. 76–79, l. 244, etc), rephrasing (e.g., in the method and results section) and providing further details, for example on the relation between flooding and topography (section 4.2.4) and the DEM artefacts. We deliberately put the overview of the source data of the global DEMs used in the supplementary material instead of the main text in order not to further extend the length of the manuscript further with general and common DEM information (reviewer 1 commented on length). However, apart from referencing to the supplement, we may integrate some aspects on the DEM source data if they are relevant when discussing the artefacts. We also decided to provide the AW3D30 results in the supplementary material as they show a similar tendency as SRTM or ASTER, i.e., indicating higher elevation than the local data. Instead, we included the ACE2 in the main text as it indicates the lowest elevations. Thus, we can show the entire range of DEM performance without generating an overload of information. However, to be inclusive and serve the international userbase of the AW3D30, we will add the AW3D30 to the other DEMs in the main text and restructure the figures accordingly. Your questions on satellite microwaves (l. 264) and tree and building removal (l. 325) address the processing of published datasets we used; we made reference to their publications. We will check the manuscript to add any additional references when needed.
  We are also very grateful for your thorough suggestions on the figures of our submission. In case of revision, we will adjust Fig. 1, which may also clarify your question on the imperfection of the AD-DEM (l. 302) then.
  Thank you again for your valuable feedback. We will upload the AD-DEM in the Pangaea repository and put the link in the manuscript once we are invited to revise the manuscript.
  Best regards,
  Katharina Seeger (on behalf of the authors)
  
  Citation: https://doi.org/10.5194/egusphere-2022-1425-AC2

Katharina Seeger et al.

Supplement

https://doi.org/10.5194/egusphere-2022-1425-supplement

Katharina Seeger et al.

Viewed

Total article views: 546 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
377	150	19	546	51	7	8

HTML: 377
PDF: 150
XML: 19
Total: 546
Supplement: 51
BibTeX: 7
EndNote: 8

Views and downloads (calculated since 04 Jan 2023)

Month	HTML	PDF	XML	Total
Jan 2023	160	45	7	212
Feb 2023	86	26	6	118
Mar 2023	77	28	6	111
Apr 2023	24	30	0	54
May 2023	29	19	0	48
Jun 2023	1	2	0	3

Cumulative views and downloads (calculated since 04 Jan 2023)

Month	HTML	PDF	XML	Total
Jan 2023	160	45	7	212
Feb 2023	86	26	6	118
Mar 2023	77	28	6	111
Apr 2023	24	30	0	54
May 2023	29	19	0	48
Jun 2023	1	2	0	3

Viewed (geographical distribution)

Total article views: 555 (including HTML, PDF, and XML) Thereof 555 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 02 Jun 2023

Short summary

Low-lying deltas are prone to relative sea level rise (SLR) and flooding, making accurate elevation data essential for flood risk assessment. We assess the land elevation of the Ayeyarwady Delta (Myanmar) in relation to local mean sea level (MSL) by generating a new, local DEM based on topographical map elevation data, and analysing the performance of 10 global DEMs, referenced to local MSL. We identify deltaic areas prone to SLR and monsoon flooding and interpret their relation to topography.


Total:	0
HTML:	0
PDF:	0
XML:	0