Reducing false alarms in urban flood detection: an enhanced NDWI (ENDWI) with Hybrid Max Fusion on Sentinel-2 Data

Almoadi, Abdulrhman M.

doi:10.5194/eo-1-43-2026

Articles | Volume 1, issue 1

https://doi.org/10.5194/eo-1-43-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/eo-1-43-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 1, issue 1

Research article

|

06 Aug 2026

Research article |

| 06 Aug 2026

Reducing false alarms in urban flood detection: an enhanced NDWI (ENDWI) with Hybrid Max Fusion on Sentinel-2 Data

Abdulrhman M. Almoadi

Download

Final revised paper (published on 06 Aug 2026)
Preprint (discussion started on 09 Feb 2026)

Interactive discussion

Status: closed

AC1: 'Comment on egusphere-2026-672', Abdulrhman Almoadi, 09 Feb 2026

Dear Editor and referees,
Thank you very much for your interest in our manuscript.
During a post-preprint check of the performance metrics, the False Alarm Ratio (FAR) has been clarified and is now reported using the standard definition (FP / (TP + FP)), where TP denotes true-positive pixels, and FP denotes false-positive pixels. The previously reported FAR values were calculated based on the total ground-truth samples, while the revised values are computed directly from predicted positive (water) pixels (TP + FP).
The revised FAR values are as follows:
Hybrid_Max: 5.50%
ENDWI_otsu: 20.59%
AWEIsh_otsu: 23.90%
NDWI_otsu: 44.33%
Importantly, the ranking of the indices in terms of false alarm reduction remains unchanged, and Hybrid_Max continues to show the lowest FAR (5.50%), which remains highly competitive and supports the core conclusions of the study.
We are committed to accuracy and will incorporate updated values in the revised manuscript submitted after the discussion phase.
Thank you again for your understanding and for the valuable interactive review process.

Best regards,
Abdulrhman Almoadi
Corresponding author

Citation: https://doi.org/10.5194/egusphere-2026-672-AC1
CC1:
'Comment on egusphere-2026-672', Chandra Mohan Bhatt, 30 Mar 2026

The study presents an attempt to improve flood detection in urban areas using the Enhanced Normalized Difference Water Index (ENDWI). However, the abstract needs to more clearly specify which spectral bands, beyond those used in conventional indices, were incorporated in the ENDWI formulation. Additionally, the authors should briefly explain the spectral rationale behind the observed reduction in false alarms, as this would enhance the general understanding and acceptance of the proposed approach.
At present, the abstract places greater emphasis on statistical results rather than adequately highlighting the methodological innovation and underlying approach. A better balance between methodological description and results would significantly strengthen the abstract.

Citation: https://doi.org/10.5194/egusphere-2026-672-CC1
- AC2:
  'Reply on CC1', Abdulrhman Almoadi, 31 Mar 2026
  
  Dear Chandra Mohan Bhatt,
  Thank you for your valuable comments,
  We will consider mentioning all indices used in the abstract; the details of the spectral bands for each index have been included in the methodology section of this version due to the limited of words normally in the abstract. Instead of identifying the term for the eight indices in the abstract, the top three indices' performance was reported.
  In addition, the ENDWI performance in reducing false alarms will be expanded across the study in the revised manuscript.
  Thank you again for your feedback.
  
  Abdulrhman Almoadi (corresponding author)
  
  Citation: https://doi.org/10.5194/egusphere-2026-672-AC2
  - CC2: 'Reply on AC2', Chandra Mohan Bhatt, 01 Apr 2026
    
    Dear Authors,
    Thank you for your response. It would be helpful if the abstract clearly specifies whether the ENDWI is being proposed as a novel index or if it is an existing method being demonstrated through a case study. This clarification will improve the positioning and contribution of the study.
    Additionally, the opening sentence of the abstract—“Although optical satellite-derived water indices have significantly advanced urban flood detection, accurately distinguishing flooded from non-flooded pixels while minimizing false positives caused by spectral confusion in built-up areas remains a considerable challenge”—could be made more concise to accommodate on ENDWI. A more focused introduction that directly highlights the ENDWI and its application would help in clearly conveying the purpose and novelty of the work.
    Regards,
    
    C. M. Bhatt
    
    Citation: https://doi.org/10.5194/egusphere-2026-672-CC2
    
    AC3: 'Reply on CC2', Abdulrhman Almoadi, 02 Apr 2026
    
    Dear Chandra Mohan Bhatt,
    Thank you again for your communication and your comment,
    The ENDWI, as an enhanced index, is completely novel; it's developed through experiments and derived from the traditional NDWI. It aims to reduce false alarms by focusing on really flooded areas. All these will pop up later in the abstract in the revised manuscript.
    Abdulrhman Almoadi (Corresponding author)
    
    Citation: https://doi.org/10.5194/egusphere-2026-672-AC3
    
    CC3: 'Reply on AC3', Chandra Mohan Bhatt, 02 Apr 2026
    
    Dear Editors,
    The study presents a valuable contribution by proposing a novel index for urban flood detection.
    The authors have agreed to incorporate the comments and suggestions provided into the revised version of the manuscript.
    Regards,
    
    C. M. Bhatt
    
    Citation: https://doi.org/10.5194/egusphere-2026-672-CC3
RC1:
'Comment on egusphere-2026-672', Anonymous Referee #1, 19 Apr 2026

Publisher’s note: this comment was edited on 20 April 2026. The following text is not identical to the original comment, but the adjustments were minor without effect on the scientific meaning.

Thank you for submitting this well-prepared and valuable manuscript. The manuscript presents a meaningful contribution by proposing an enhanced water index (ENDWI) and demonstrating its effectiveness in reducing false alarms in urban flood detection. The methodology is sound, and the results are convincing.

Only minor revisions are suggested to improve clarity:

1. The abstract could more clearly state whether ENDWI is a novel index or a modification of existing indices (e.g., NDWI), and briefly indicate its key characteristics.

2. A short explanation of the spectral rationale behind the reduction in false alarms would improve the interpretability of the results.

Overall, the manuscript is well prepared and suitable for publication after these minor clarifications.

Citation: https://doi.org/10.5194/egusphere-2026-672-RC1
- AC4: 'Reply on RC1', Abdulrhman Almoadi, 23 Apr 2026
  
  Dear Reviewer,
  Thank you for your comments. The response was uploaded as a supplementary.
  
  Citation: https://doi.org/10.5194/egusphere-2026-672-AC4
- AC5: 'Reply on RC1(Part 2)', Abdulrhman Almoadi, 23 Apr 2026
  
  Dear Reviewer,
  
  Thank you for your valuable comments. A detailed response has been uploaded as a supplementary file.
  
  Citation: https://doi.org/10.5194/egusphere-2026-672-AC5
RC2:
'Comment on egusphere-2026-672', Guy J.-P. Schumann, 19 Apr 2026

This paper presents a method to process S2 flood images using an enhanced NSWI method.
The paper is well written and I have only a few comments but some of those deserve moderate to major attention.
- It is appreciated that the authors demonstrated the innovation presented relative to other published methods, but I feel they need to better justify why their method is preferred. Is it better accuracy, more robustness/consistency, better scaling. I feel the accuracy gained does not necessarily justify the implementation.
- the authors should also perform a simple cost-benefit analysis against other methods using some of the performance measures but also computational efficiency and robustness, obtained by computing a series of different flood events.

Citation: https://doi.org/10.5194/egusphere-2026-672-RC2
- AC6: 'Reply on RC2', Abdulrhman Almoadi, 24 Apr 2026
  
  Dear Guy J.-P. Schumann,
  
  Thank you for your valuable comments. A detailed response has been uploaded as a supplementary file.
  
  Citation: https://doi.org/10.5194/egusphere-2026-672-AC6

Peer review completion

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

ED: Publish subject to revisions (further review by editor and referees) (23 May 2026) by Luca Brocca

AR by Abdulrhman Almoadi on behalf of the Authors (09 Jun 2026) Author's response Author's tracked changes

EF by Natascha Töpfer (10 Jun 2026) Manuscript

ED: Referee Nomination & Report Request started (11 Jun 2026) by Luca Brocca

RR by Anonymous Referee #1 (11 Jun 2026)

ED: Publish as is (30 Jul 2026) by Luca Brocca

AR by Abdulrhman Almoadi on behalf of the Authors (01 Aug 2026) Author's response Manuscript

Short summary

Urban floods are difficult to detect accurately in satellite imagery because non-water surfaces often resemble floodwater, causing false alarms. This study introduces an enhanced normalized difference water index (ENDWI), derived from the normalized difference water index (NDWI), combined with a second hybrid index to reduce these false alarms using freely available Sentinel-2 imagery. False alarms were reduced from 38 % to under 3 %, offering a more reliable, low-cost approach to urban flood mapping.

Reducing false alarms in urban flood detection: an enhanced NDWI (ENDWI) with Hybrid Max Fusion on Sentinel-2 Data

Download

Interactive discussion

Peer review completion

Suggestions for revision or reasons for rejection