Enhancing the performance of 1D–2D flood models using  satellite laser altimetry and multi-mission surface water  extent maps from Earth observation (EO) data

Charoensuk, Theerapol; Lorentzen, Claudia Katrine Corvenius; Bak, Anne Beukel; Luchner, Jakob; Tøttrup, Christian; Bauer-Gottwein, Peter

doi:https://doi.org/10.5194/hess-29-5065-2025

Articles | Volume 29, issue 19

https://doi.org/10.5194/hess-29-5065-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/hess-29-5065-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 29, issue 19

Research article

|

09 Oct 2025

Research article |

| 09 Oct 2025

Enhancing the performance of 1D–2D flood models using satellite laser altimetry and multi-mission surface water extent maps from Earth observation (EO) data

Theerapol Charoensuk, Claudia Katrine Corvenius Lorentzen, Anne Beukel Bak, Jakob Luchner, Christian Tøttrup, and Peter Bauer-Gottwein

Download

Final revised paper (published on 09 Oct 2025)
Preprint (discussion started on 25 Jul 2024)

Interactive discussion

Status: closed

RC1:
'Comment on hess-2024-175', Anonymous Referee #1, 08 Sep 2024
The author must constructively change the abstract in terms of adding error analysis values in terms of PBIAS to the result. The Author needs to write consistently.

The author must mention where they got data and the frequency of data. Statistical analysis of data must be given in Tabular format (Like Table no 1, 10.1061/(ASCE)IR.1943-4774.0001689).

Fig 2: it should be clearly described in terms of scientific manner.

For better understanding, Please add more recent literature (2024) regarding the bidirectional flood models using satellite laser altimetry.

Please modify the objective section for a clear understanding, i.e., the novelty part should be mentioned.

There are so many techniques in the recent world for the flood model; why does the author use a specified Method for research purposes? Is there any specific reason for this?

The author must add statistical components/parameters of collected data in the case study section.

Eq 3-6; please add a recent citation for reference purposes. [Read this paper: 10.1016/j.gsd.2024.101178, 10.1016/j.clwat.2024.100003, 10.1016/j.hydres.2024.04.006, 10.1007/978-981-15-5397-4_75, 10.1038/s41598-024-63490-1, 10.2166/wcc.2021.221]

A comparison statement (compare with other research articles) must be added in the result and discussion section to visualize the proposed research better.

The author must add future scope in the last portion of the manuscript.

The advantages and limitations of the proposed model must be added.

For better analysis of the result, the author must add a Box plot, Taylor diagram, and ROC Curve

The author must provide a flow chart and parameter table of proposed individual models.

The author considered different input constraints; is there any scientific reason for the same
Citation: https://doi.org/10.5194/hess-2024-175-RC1
- AC1:
  'Reply on RC1', Theerapol Charoensuk, 24 Sep 2024
  
  The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2024-175/hess-2024-175-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/hess-2024-175-AC1
  - RC2: 'Reply on AC1', Anonymous Referee #1, 26 Sep 2024
    
    Thank you for the revision; now the article may proceed to the next stage.
    
    Citation: https://doi.org/10.5194/hess-2024-175-RC2
    
    AC3: 'Reply on RC2', Theerapol Charoensuk, 28 Oct 2024
    
    Thank yoi
    
    Citation: https://doi.org/10.5194/hess-2024-175-AC3
RC3:
'Comment on hess-2024-175', Anonymous Referee #2, 14 Oct 2024
general comments
Interesting study on the use of different DEM inputs into a specific 1D-2D flood model. In the first part, the most accurate global and airborne DEMs are determined. Using these two DEMs as input, the resulting flood model maps are compared with 2 reference flood maps. Surprisingly, the global flood-optimized FABDEM derived from TanDEM-X achieves only slightly worse flood modelling quality statistics than the higher-resolution airborne DEM version LDD-JICA DEM.
The paper is generally well structured and balanced, but the context and objective of the two parts are not consistently clear.
The claimed objective of your study to present two new workflows for updating 1D-2D flood models is not plausible. It seems like your starting point for this study was "what can we do with EO data", but what you describe is how to test the input DEMs and validate your results. I can't see any real improvement in the 1D-2D flood model itself:An extensive DEM evaluation does not make sense for every new 1D-2D flood model, as the number of input DEMs is limited and you have shown extensive evaluation here. Similarily, the validation of flood modeling results with existing flood maps is not an integral update of a model.
In my opinion, there is a simple way out: move the “workflows” to the discussion/ conclusion and stick to terms like “evaluation of DEMs for” … and “validation of flood model results” … .
However, please clarify this throughout the paper, even in the title!!!

specific comments
Please improve the abstract (and title) with regard to the readability and research focus of your study. Main point: The paper gives a kind of performance test. So, your description given in 4.2. (“ .. to evaluate the performance of simulated flood maps … using various DEM products”) comprises the content more appropriate than an “upgrade by two workflows”.

In that sense, unclear in the abstract: are you evaluating the 1D-2D model results with the surface water extent maps or has this any relation to the DEM analysis part? Scientifically using SWE maps are for validation.

The same applies to the title. Please use a more precise title (laser altimetry was solely used for performance assessment, same applies for the water extent maps (validation, not for an software/model udgrade itself, …) Something like Influence of DEM quality /Performance assessment using global DEM / …

Abstract/Intro: The first part “DEM analysis” evaluates 10 DEMs compared to ICESat-2. Please explain your motivation. -> advantage of EO DEMs. The DEM choise is rather heterogenious -> Please categorize the used 10 DEMs e.g. from satellite to airborne DEMs.

In General: There might exist some specific/logical requirements for DEMs to test or mentioning in advance if they are suited for flood modeling (e. g. like in Gesch, Front. Earth Sci., 2018, Best Practices for Elevation-Based Assessments of Sea-Level Rise and Coastal Flooding Exposure, https://doi.org/10.3389/feart.2018.00230). Against this background, please justify why you start analyzing for so many DEMs the quality from scratch!

Methodology: Chapter 4, Your goal is to find out the best DEM : better in terms of the most accurate DEM compared to ground control points

4.1.1 Not described: how do you use/prepare ICESat-2 ATL03 data? If you don’t explain it, omit it.

Comment: Apart of the point comparison I like the additional value of grid and track-wise comparison.

Results: Sec. 5.2.1 To what table do you refer with the statement given in l.444 (global DEM products was +1.62 m)? Same for statistics in line 449 (no common global DEM statistic in table given).

please omit the wording “over- and underestimation” when describing small biases of 1-2 m without regarding the RMSE. (i.e. having an RMSE of 2m: an ME of 1 m is within the noise level! No real over- or underestimation) Please scan the whole document! Better use neutral terms like small positive/negative bias.

Please re-work the text in Section 5.3. to make it more readable and comprehensive.

Section 6.2.: Message is unclear, as the different maps for validation seems to have its deficits.

Please just list the used data sets / days in Appendix Tables A1 or omit table completely, Table A2 can be omitted completely. It is a service with regular, almost daily acquisitions.

The visualization of the geoid models Fig A2 should be omitted.

Minor comments:
Abstract:” Given the current uncertainties stemming from changes in weather patterns affecting flooding, reducing inaccuracies in flood models is imperative”: Please be more precise. Are the uncertainties improving?
Include in abstract: Which DEMs are finally used for your flood map analysis and why (motivation to chose one global and one local DEM)
l23-l25: Abstract: Think about your message! EO data for validation of DEM and Flood model maps were used. Conclusion/Result?
Line 480 … which can be attributed to the fact that vegetation and buildings are eliminated in this DEM …
514: we implemented: please re-word to e.g. “performed test with two DEMs,…”
Figure caption: difference of Fig. caption A7 and A9 not clear; RMSE?
Citation: https://doi.org/10.5194/hess-2024-175-RC3
- AC2: 'Reply on RC3', Theerapol Charoensuk, 21 Oct 2024
  
  The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2024-175/hess-2024-175-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/hess-2024-175-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

ED: Reconsider after major revisions (further review by editor and referees) (15 Nov 2024) by Yue-Ping Xu

AR by Theerapol Charoensuk on behalf of the Authors (02 Dec 2024) Author's response Author's tracked changes

EF by Daria Karpachova (10 Dec 2024)

ED: Referee Nomination & Report Request started (17 Dec 2024) by Yue-Ping Xu

RR by Anonymous Referee #3 (20 Mar 2025)

Suggestions for revision or reasons for rejection

General Comments
This paper, entitled “Enhancing the Performance of 1D-2D Flood Models Using Satellite Laser Altimetry and Multi-Mission Surface Water Extent Maps from Earth Observation (EO) Data” by Theerapol Charoensuk, discusses how to improve 1D-2D flood modeling accuracy by validating DEMs with ICESat-2 data and comparing model-predicted inundations against satellite-derived flood extent maps. The paper has clear methodological steps and demonstrates practical value by comparing modelled flood extents with satellite-derived inundation data. The attempt to incorporate ICESat-2 for DEM validation and to benchmark flood maps is timely and relevant. The manuscript would benefit from further discussion on the surprising result that a globally derived FABDEM (with coarser nominal resolution) seems to outperform the locally acquired LiDAR DEM product in certain respects. In addition, the choice of the acronym “SWE” for surface water extent may cause confusion with the more commonly known usage, “snow water equivalent.” With a more detailed discussion of how and why the global FABDEM can outperform an apparently high-resolution local DEM, plus a reconsideration of the acronym used for flood inundation layers, the manuscript will be substantially clearer. The two major issues are as follows.

(1) Discussion of Why a Global DEM (FABDEM) Outperforms a More Detailed Local DEM
A core finding in your study is that the coarser-resolution FABDEM outperforms—or at least equals—the merged LDD-JICA DEM (derived in part from airborne LiDAR) for certain evaluation metrics. This is somewhat counterintuitive, given that airborne LiDAR typically achieves very high resolution and accuracy. Readers will want a deeper explanation as to why this may have occurred. Potential factors include:
• ICESat-2 Footprint and Sampling: If the ICESat-2 ATL08 data used for validation has footprints or sampling intervals that favor broader-scale features, a coarser DEM may align better with the reference data.
• Local DEM Data Quality: Airborne LiDAR can theoretically be very precise, but if data collection or processing was suboptimal, it could introduce errors. Factors like incomplete vegetation filtering, outdated surveys, or vertical datum mismatches might degrade its performance.

It would be beneficial to elaborate on each of these points to clarify whether you attribute FABDEM’s advantage mainly to differences in how building and vegetation heights were removed, to possible systematic bias in the local DEM, or to the nature of the ICESat-2 validation approach itself. Additionally, highlight any limitations in your current processing steps that may have contributed to this unexpected outcome.

(2) “SWE” Acronym for Surface Water Extent
In many hydrologic and cryosphere studies, SWE commonly stands for “Snow Water Equivalent.” Using the same acronym to denote “Surface Water Extent” could confuse readers. It might be preferable to adopt an alternative abbreviation such as IE (Inundation Extent), WSE (Water Surface Extent), or another unambiguous term.

Although you clearly define “SWE” in the text as surface water extent, using a more standard term for inundation areas would improve clarity. Please consider replacing the acronym throughout the manuscript and figures with a term that would not conflict with “snow water equivalent.”

Hide

RR by Anonymous Referee #2 (15 May 2025)

EF by Vitaly Muravyev (18 Mar 2025) Manuscript

ED: Publish subject to minor revisions (review by editor) (23 May 2025) by Yue-Ping Xu

AR by Theerapol Charoensuk on behalf of the Authors (02 Jun 2025) Author's response Author's tracked changes Manuscript

ED: Publish subject to technical corrections (21 Jun 2025) by Yue-Ping Xu

AR by Theerapol Charoensuk on behalf of the Authors (29 Jun 2025) Manuscript

Short summary

The objective of this study is to enhance the performance of 1D–2D flood models using satellite Earth observation data. The main factor influencing a 1D–2D flood model is the accuracy of the digital elevation model (DEM). Two workflows are introduced to improve the 1D–2D flood model: (1) in the DEM analysis workflow, 10 DEM products are evaluated using the ICESat-2 ATL08 benchmark, while (2) in the flood map analysis workflow, flood extent maps derived from multi-mission satellite datasets are compared with simulated flood maps.