the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Modelling tides and storm surge using intertidal bathymetry derived from the waterline method applied to multispectral satellite images
Wagner Luiz Langer Costa
Karin Roisin Bryan
Giovanni Coco
Abstract. Bathymetric data are essential for accurate predictions of flooding in estuaries, because water depth is a fundamental component in the shallow-water hydrodynamic equations used in numerical models. Where LiDAR or acoustic in-situ surveys are unavailable, recent efforts have centred on the use of satellite images to estimate bathymetry (SDB). This work is aimed at (1) determining the accuracy of SDB, and (2) assessing the suitability of the SDB for surge/tidal modelling of estuaries. The SDB is created by extracting the waterline as it tracks over the bathymetry with changing tides, and is applied to 4 different estuaries in New Zealand: Whitianga, Maketū, Ōhiwa and Tauranga Harbour. Results show that the waterline method provides similar bathymetries to the LiDAR with root-mean squared error equal to 0.2 m, and it is slightly improved when two proposed correction methods are applied to the bathymetry derivations: the removing of statistical bias (by 2 cm) and hydrodynamic modelling correction (by 1 cm). Finally, the use of SDB in numerical simulations of surge levels is assessed for Tauranga Harbour with 4 different scenarios that explore the use of SDB in comparison to bathymetry data collected using non-satellite survey methods. One of these includes the well-known Stumpf-ratio method to extract the SDB of subtidal regions (so that only satellite information is used). The use of the satellite derived bathymetry in hydrodynamic models does not result in significant differences in terms of water levels, when compared with the scenario modelled using surveyed bathymetry.
- Preprint
(2331 KB) -
Supplement
(1646 KB) - BibTeX
- EndNote
Wagner Luiz Langer Costa et al.
Status: final response (author comments only)
-
RC1: 'Comment on nhess-2021-387', Anonymous Referee #1, 25 Apr 2022
The manuscript presents an interesting approach to deriving intertidal bathymetry from the waterline method through multispectral images, covering four (4) estuarine study areas on the east coast of Aotearoa New Zealand’s North Island (Tauranga, Ohiwa, Maketu and Whitianga harbour). It represents a current thematic area, and it can be particularly useful to be applied in remote or inaccessible areas or where the bathymetric or cartographic data is very outdated. The main objectives of the study are to determine if multispectral images can be used to extract accurate intertidal bathymetric area and to assess the use of the SDB for hydrodynamic modelling of estuarine.
Good English level however the manuscript is not well-structured, quite confusing and the reader easily misses the main guidelines and the aim of the study. In section 1 (Introduction) is very difficult to establish a connection between the different ideas and paragraphs. A deeper revision of the state of the art is needed to bring the reader into the SDB theme and waterline method. The flow chart in chapter 2 is useful but does not really explain the methods used by the authors. Furthermore, the Methods Chapter establish that the main method was divided into 2 steps (1-SDB estimation and 2-Hydrodynamic modelling assessment) and that step 1 is also two methods for removing the bias, but a clear explanation of the methodology is not present in this section of the manuscript. A very short discussion and a shorter conclusions section are shown, where no clear main findings can be found. Modelling Storm surge is only referenced in the title of the manuscript.
The manuscript shows that a lot of work has been made, however, a big gap throughout the presented structure is noticed and the methodology used is not well described, creating a lot of misunderstanding between the methods applied and the different steps described by the authors. I, therefore, do not recommend the publication of this manuscript as it was presented. A major revision of the structure and methodology form is recommended. My main critics are the following:
- In the Introduction section the theme is not quite explained, and only part of the aim of the study is presented in the last sentence of the last paragraph. In this section is expected that the authors explain the reasons that have motivated this study, as well as what will be presented in the different sections of the entire manuscript.
- The different figures do not follow a consistent presentation. The geographic coordinates in some cases are presented as latitude/longitude with no reference datum associated (Fig.2(a)); others as X/Y coordinates (km) WGS84/UTM60S (Figs. 2 (b), (c), 3 (a)) and even other examples as X/Y coordinates NZGD2000 (km).
- The same Figures, presented in different sections, have different SI unit references, like Fig.2 (b) and (c) are expressed in X/Z coordinate (km) and Figures S1 and S2 in X/Z coordinate (m) – show a lack of consistency.
- The areas A, B, C and D depicted in Figure7 (central figure) are not quite perceptible, and the small figures (a1, a2, b1, b2, c1, c2, d1 and d2) do not have geographic coordinates associated, neither the scale factor.
- The profile lines drawn in Figure 9 (m1) are barely noticed. Maybe the authors could choose a different colour palette.
- In the text, the figures are not correctly cited, like Fig 2A (line 85); Fig 2B and 2C (line 88). In the Figures, the panels are mentioned with small letters (a, b and c), as well as in the figure capture.
- The data access information at the reference links (lines 94-97) is missing.
- I do not understand how the intertidal area is identified, the method is not well explained. Is used the tidal level at the time of the acquisition of each image? Or is used an average tidal range (tidal amplitude?) for all the images. Is also not clear the tidal level for each image, as depicted in Figure 2(d). All images are used to generate the intertidal area presented in Figure 3(a)?
- The threshold value used, and all the contour extraction method (lines 153-159) are quite confusing. And which values of threshold and water level were used for the other study areas, regarding that Figure 4 presents the water level and threshold values for each image. A table with this information, for all the different study areas, as supplementary information could be very useful.
- What do you mean with the Stumpf-ratio method was applied for deeper areas (lines 164-165). The Stumpf ratio method (Stumpf et al., 2003) is not quite good for all different benthic areas and for very deeper areas. What was the maximum depth value which the authors have used this method?
- It was not explained by the authors all the pre-processing steps applied to the multispectral images, such as sun glint correction (for example Hedley, J.D.; Harborne, A.R.; Mumby, P.J. Simple and robust removal of sun glint for mapping shallow-water benthos. Int. J. Environ. 2005, 113, 2107–2112). If this step was considered, it should be enunciated in the manuscript. The authors described that Level 2 image was used, with BOA values corrected for the effects of the top-atmosphere (lines 103-104), but it was not explained why they used these images rather Level 1 with atmospheric correction.
- I can not understand if the evaluation of the model performance in section 2.4 is one of the results of this study. And if they are, why not present them in the results section? Lines 201-215 have a challenging interpretation.
- Why an” Average” line in table 3. Does not make sense.
- Lines 234-239 should be included in the Discussion section, not here, where the results are presented.
- In the ESA Sentinel 2A images used as background in several figures are missing the data acquisition time and the water-level information (Figures 2, 7, S1, S2, S3, S5, S6, S7).
- The authors cannot quantify as good or strongly correlated/related the R2 achievements (lines 242-245). Why R2=0.70 should be considered as strongly related? Once more the authors are discussing the presented results in the Results section, and it is a recurrent procedure throughout this section. Perhaps if the authors had previously described in section 1 the contents of each section, the reader could understand better the manuscript. The structure of each section is quite confusing.
- Lines 274-278: R2 values assumption/classification (low/higher). And R2 is referred to as the coefficient of determination (line 276) and a coefficient of correlation (line 278) in the same paragraph. Is not coherent.
- The authors do not explain why the results and the application of the methodology were only presented for one study area (Tauranga Harbour). They are free to do it, even for editorial figures or pages restrictions, however, this fact should be mentioned and explained in the manuscript and the main results for each area should be resumed (table format perhaps) in the supplementary material section.
- What is the spatial grid resolution value (line 298)? Is 20 m as assumed in line 336?
- Lines 313-316 are quite confusing. A better explanation is needed.
- The prediction of water level using the SDB is presented in section 3.4 for the 3 tide gauges (Omokoroa, Hairini and Oruamatua) (lines 323-331). The average error parameter presented is for each tide gauge and Figure 10 shows the average between all the tide gauges. Was this methodology that was used? I am confused.
- Can I assume that, for lower tide values images, the presented methodology can not be used? Or only for the Stumpf ratio method application (SDB)? Lines 338-341. The Stumpf ratio method can not be directly applied to intertidal areas, exactly due to the image reflectance of the dry pixels (low water level images).
- What represents the rectangle-shaped figure in Figure 11? Survey bathymetry data or LIDAR data?
Citation: https://doi.org/10.5194/nhess-2021-387-RC1 -
AC1: 'Reply on RC1', Wagner Costa, 26 Jul 2022
Dear Reviewer anonymous referee 1,
Thank you for your observations regarding our preprint. Your suggestions helped to improve our manuscript.
Based on your revue and those provided by the other reviewers, we understand that we need to modify the structure of the paper because the current format is confusing. We have undertaken a deeper revision of the state of the art on SDB (adding new references and text to the introduction section). We have added a much clearer aim to the introduction, and worked on linking the methods to the aim in a much more clear and logical order. In the methods section, we added further explanations about the different methods implemented to remove the bias (i.e., statistical and dynamical methods). In the discussion and conclusion sections, we built further on the context provided in the new references added in the introduction part. Please note that we did not model the storm surge, but we analyzed the maximum astronomical tide in all simulation scenarios and compared the outputs between scenarios using only surveyed bathymetry, only SDB, and mixed surveyed bathymetry combined with SDB. In terms of coastal flooding, the maximum water level is the main parameter studied and in most places in the world, the water level is dominated by the tide. In summary, we are happy to modify the paper structure as you and other reviewers recommend.
Please find in the supplement the answer to your specific questions.
Best regards,
The Authors.
-
AC1: 'Reply on RC1', Wagner Costa, 26 Jul 2022
-
CC1: 'Comment on nhess-2021-387', Matías Dinapoli, 13 May 2022
The manuscript presents a novel methodology to deriving intertidal bathymetry for four estuaries in New Zealand (Tauranga, Ohiwa, Maketu and Whitianga harbour) characterized by a complex morphology. I find this thematic interesting, as it allows to update and improve the boundary conditions of regional numerical models. However, I think the structure and writing of the manuscript require further work to reflect all the work done. The manuscript needs a better use of English, a restructuring of the chapters and above all to emphasize the purpose of the work as well as the authors’ motivation and innovations. Therefore, I do not recommend the publication of this manuscript as submitted. This review is critical, nonetheless the authors have the potential to have a great manuscript and I would like to encourage them in their progress.
I mainly concern of the reasoning and the reading flow, which is quite confusing and the reader can easily miss the guidelines of the study. Section 1 does not clearly show the developments achieved by the scientific community, the relevance of the chosen methodology and, above all, the authors’ motivations. Section 2 is very long and presents too many technical concepts, and even results, that is hard to follow how they were implemented. The study area should be extpanded with a description of the main processes describing water level dynamic, since the work’s title mentions the storm surge modelling. Results and figures in Section 3 present a lack of consistency of SI units, authors should homogenize them. I think criteria presented in Fig. 3 and 4 are unclear and need a deeper discussion. Errors should be accompanied by their percentage for better interpretation. Unfortunately, the color map chosen for Fig. 7 and 9 is not good for presenting such significant results. The conclusions are extremely short and not summarize the reasoning of the work.
Citation: https://doi.org/10.5194/nhess-2021-387-CC1 - AC4: 'Reply on CC1', Wagner Costa, 26 Jul 2022
-
RC2: 'Comment on nhess-2021-387', Anonymous Referee #2, 19 May 2022
The manuscript presents a novel methodology to deriving intertidal bathymetry for four
estuaries in New Zealand (Tauranga, Ohiwa, Maketu and Whitianga harbour) characterized
by a complex morphology. I find this thematic interesting, as it allows to update and
improve the boundary conditions of regional numerical models. However, I think the
structure and writing of the manuscript require further work to reflect all the work done.
The manuscript needs a better use of English, a restructuring of the chapters and above
all to emphasize the purpose of the work as well as the authors’ motivation and
innovations. Therefore, I do not recommend the publication of this manuscript as
submitted. This review is critical, nonetheless the authors have the potential to have a
great manuscript and I would like to encourage them in their progress.
I mainly concern of the reasoning and the reading flow, which is quite confusing and the
reader can easily miss the guidelines of the study. Section 1 does not clearly show the
developments achieved by the scientific community, the relevance of the chosen
methodology and, above all, the authors’ motivations. Section 2 is very long and presents
too many technical concepts, and even results, that is hard to follow how they were
implemented. The study area should be extpanded with a description of the main
processes describing water level dynamic, since the work’s title mentions the storm surge
modelling. Results and figures in Section 3 present a lack of consistency of SI units,
authors should homogenize them. I think criteria presented in Fig. 3 and 4 are unclear
and need a deeper discussion. Errors should be accompanied by their percentage for
better interpretation. Unfortunately, the color map chosen for Fig. 7 and 9 is not good for
presenting such significant results. The conclusions are extremely short and not
summarize the reasoning of the work.Citation: https://doi.org/10.5194/nhess-2021-387-RC2 -
AC2: 'Reply on RC2', Wagner Costa, 26 Jul 2022
Dear Reviewer anonymous referee 2,
Thank you for your observations regarding our preprint. Your suggestions helped to improve our manuscript.
We appreciate that you found our manuscript interesting, despite the problems regarding its structure. The inadequate structure is a problem highlighted by all the 3 reviewers. We intend to modify it accordingly. For instance, we plan to describe the content of each chapter at the end of the introduction section and set the paper up with more clearly presented aims.
Please find in the supplement the answers to all your specific questions.
Best regards,
The Authors.
-
AC2: 'Reply on RC2', Wagner Costa, 26 Jul 2022
-
RC3: 'Comment on nhess-2021-387', Anonymous Referee #3, 09 Jun 2022
I have read the paper entitled “Modelling tides and storm surge using intertidal bathymetry derived from the waterline method applied to multispectral satellite images” by Costa et al. The study aims to determine whether satellite imagery can be used to extract accurate intertidal bathymetric data; and assess the use of the SDB for hydrodynamic modelling of estuaries. The paper is interesting, and one can see that quite a lot of effort, based on the complex methodology, was put into it. However, the manuscript is quite confusing making it difficult to understand. Part of the problem I had was regarding the use of the term SDB to represent extracted shorelines when the term is coined to deriving bathymetry. The study seems to have its merits but needs a complete re-structure. I found results difficult to understand.
Because it presents lots of technical concepts, I’d divide this paper in two manuscripts and focus on convincing the reader that waterline extraction can be useful to derive intertidal DEM in NZ and leave the SDB and modelling approach for another opportunity. A short discussion is also presented in this submitted version for such complex topic. Therefore, a major revision or complete new submission is recommended
It is not clear to me why one would embark on a shoreline extraction method to create an intertidal model, when one could use SDB (Stumpf and others) to obtain bathymetry, especially where white water/waves are absent.
I found the introduction a quite confusing as it mixed two different uses of satellite. One to derive bathymetry and the other to derive shoreline positions. This is carried out from the Abstract to Introduction to the other parts of the text, and therefore I suggest a complete rewrite of these sections.
Some specific points below:
L9 - I’d suggest to modify text to make better use of the acronym – Satellite-derived bathymetry (SDB) which obviously differ from “use of satellite images to estimate bathymetry”
L11 four instead of 4. Same in L16
L18 The use of the satellite derived bathymetry in hydrodynamic models does not result in significant differences in terms of water levels, when compared with the scenario modelled using surveyed bathymetry. This seems a big claim to me considering that the method was only used in microtidal settings and NDWI performance in macrotidal places can be more complicate due to the larger wet areas. Sea grass bed areas appears also to be an issue.
LN23- what about meteorological tides? They seem quite important for predicting floods
LN24 no hyphen in sea level
LN25- to my knowledge atmospheric pressure is the one driving storm surges along the coast, fluvial discharge definitely adds to it, but it is the difference in pressure that elevates the water level
Ln34 the references following this sentence should focus on SDB and not shoreline “To overcome these issues, efforts have centred on using spaceborn remote sensing (RS) techniques (Bishop-Taylor et al., 2019; Bué et al., 2020; Caballero and Stumpf, 2019),” should be replaced
LN41 Again I don’t think the Bishop et al. ref is appropriate here, as their article addresses shoreline and not bathymetry
LN 42 rewrite “use a radiometric approach, which uses the property that different wavelengths are attenuated to varying degrees in the water column”
LN55 detecting the land-water boundary has nothing to do with deriving bathymetry with satellites. The shoreline is the interface not the morphology of the seafloor. Bishop et al., didn’t derive bathymetry. They derived intertidal DEM, linking terrestrial and bathy datasets
LN57-59 only here I start to get a feeling of why you are talking about SDB, but even after that I think that you are creating a DEM of a few mts-depending on the local amplitude- instead of lets say shallow water bathymetry
LN71 bathymetry. You are talking about creating a DEM based on shoreline positions. Some of these positions will be above tide datum. Does that make bathymetry or topography?
LN77 2 main steps (Fig. 1).
LN78 two instead of 2
LN88 the intertidal zone is easily distinguished by the colour of sand accentuating reflectance in the near infrared band – This sentence seems out of context or needs some further clarification as Fig 2 is not a false colour image.
LN89 Associated with tidal flats, mangrove forest can be observed in all the studied estuaries. Where can I observe mangrove and seagrass banks? Modify text
Ln92 I get confused here.” For the implementation of SDB techniques, only tidal levels and imagery are needed. We used additional in situ bathymetric data to validate the SDB.” Do you need bathy or not for implementing SDB?
LN114 t seems to me that some of the derived shoreline elevations cannot be considered bathymetry and this is part why I don’t like the use of the term for this shoreline extractions. Some of the elevations will be above MSL. Can you still call it bathymetry? Shouldn’t be topography then?
LN140 you are using NDWI to define the intertidal area. Please explain know the 9 images for Tauranga or the 7 for Ohiwa are capturing the full extent of the intertidal area. Where they acquired during the lowest-highest tidal range?
LN156 Again- colloquialism
LN157 (Fig. 4)
LN158 once the waterline for a given image is identified, a height value is assigned to it accordingly to the corresponding tide level observed at the closest tide gauge (Omokoroa for the Tauranga Harbour case study, Fig. 2D). I see 4 gauges in the estuary. What’s the rationale for choosing Omokoa? Oraumatua seems way closer to let’s say Rangataua Bay! How do you account for the tidal lag? The level at the entrance is different than at the head.
Ln202 hyphen mean-sea level
LN259 Tauranga Harbour’s waterline-derived SDB (primary SDB)- Sometimes I get really lost- What’s primary SDB and how it differs from the other SDBs. Please explain
Fig 8 font size too small
Fig 9 Why are the coordinates in this map in NZGD? This figure needs to be improved. The colour scheme does not allow differentiation btw gauges and lines. It has 2 contradicting legends showing water lines as points and lines. Where are the LiDAR and the dynamic waterlines? Are they only shown in profile? I bit confusing to understand
LN323 The simulation scenarios showed that it is possible to obtain similar, or even enhanced water level predictions, by using the SDB rather than the surveyed bathymetry – I’m a bit lost here. My understanding is that we need bathymetry to do SDB! At least I had to use a few lines in the past.
LN 383 Bathymetric data are fundamental for solving the hydrodynamic equations in shallow water – This seems obvious, isn’t?
LN396 Is that it? A 1.5 pg long discussion, for such a complex paper?
Citation: https://doi.org/10.5194/nhess-2021-387-RC3 -
AC3: 'Reply on RC3', Wagner Costa, 26 Jul 2022
Dear Reviewer anonymous referee 3,
Thank you for your observations regarding our preprint. Your suggestions helped to improve our manuscript.
We are pleased that the reviewer finds our manuscript interesting. We are aware (based on your review and others) that the structure of the paper is not ideal, and the structure makes the paper confusing to read. We intend to re-structure the paper according to the reviewers’ comments. Regarding the methods implemented, there is a misunderstanding. The identification of the waterline position is done as part of the waterline method for estimating bathymetric data in the intertidal zone. We will make this more clear.
Please find in the supplement, the response to all your questions in detail.
Best Regards,
The authors.
-
AC3: 'Reply on RC3', Wagner Costa, 26 Jul 2022
Wagner Luiz Langer Costa et al.
Wagner Luiz Langer Costa et al.
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
529 | 223 | 28 | 780 | 45 | 13 | 11 |
- HTML: 529
- PDF: 223
- XML: 28
- Total: 780
- Supplement: 45
- BibTeX: 13
- EndNote: 11
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1