Saeed, Muhammad, Alhosani, Mohamed H., Mikhaylov, Dmitry, Wahedi, Yasser F. Al · Oct 24, 2026
arXiv physics.ao-ph: Satellite-derived bathymetry at Arabian Gulf ports
Markdown · JSON · Original source
The authors, Muhammad Saeed, Mohamed H. Alhosani, Dmitry Mikhaylov and Yasser F. Al Wahedi, argue that how satellite-derived bathymetry (SDB) is tested can matter more than which model is used. Against multibeam soundings from six Abu Dhabi port areas, LightGBM’s RMSE rose from 0.57 m with a random pixel split to 1.30 m with spatial blocks and 1.66 m at an unseen site. In transfer to an unseen site, the Stumpf ratio was not beaten by LightGBM, U-Net or Lyzenga. With local control, a hybrid reached 1.11 m RMSE over 0–11 m, but the seabed could not be read beyond 11 m. Clear-water results were better. The authors describe Sentinel-2 SDB with about a hundred control depths per image as useful for reconnaissance and change monitoring to about 10 m, while saying it does not meet IHO S-44 Order 1 without local control. They also caution that random-split accuracy should not be treated as operational accuracy. Their results make validation design central to judging whether these maps can support use beyond surveyed locations. Performance remains limited by water conditions and depth.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
physics.ao-ph · Article
Source details
Original source date: Oct 24, 2026 (2026-10-24).
Date basis: Date basis not provided.
Added to Escape Meridian: 2026-11-11T00:55:46.555Z. First observed: 2026-10-27T04:07:02.465Z. Retrieved: 2026-11-02T06:46:41.270Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
The evidence is the paper's abstract, which summarizes its methods and results; detailed procedures and results beyond that account are not available here. The authors report that the seabed could not be read beyond 11 m and that summer images were five to six times more turbid than winter ones. The paper reports a 60 m reconnaissance mosaic calibrated cell by cell on ICESat-2 without vessel survey, but the abstract says held-out blocks over half of the product reached 1.42 m RMSE; it does not establish accuracy across the entire mosaic.
Open question
How well would the reported accuracy hold at port areas beyond the six tested, especially where water conditions differ?
Source-stated
The authors compare SDB methods against 285,619 multibeam soundings from twelve surveys across six Abu Dhabi port areas.
against 285,619 multibeam soundings from twelve surveys at six port areas
Source-stated
For LightGBM, the reported RMSE increased as validation moved from a random pixel split to spatial blocks and an unseen site.
the same LightGBM model gave 0.57 m RMSE under a random pixel split, 1.30 m under spatial blocks and 1.66 m at an unseen site.
Source-stated
In transfer to an unseen site, the Stumpf ratio was not beaten by the tested LightGBM, U-Net or Lyzenga models.
In transfer the Stumpf ratio (1.45 m) was not beaten by LightGBM, the U-Net or Lyzenga.
Source-stated
The authors report that a hybrid with local control reached 1.11 m RMSE over 0–11 m, but the seabed could not be read beyond 11 m.
With local control a hybrid of the tree output and the band ratio reached 1.11 m over 0-11 m
Source-stated
The authors characterize Sentinel-2 SDB with about a hundred control depths per image as useful for reconnaissance and change monitoring to about 10 m, but not sufficient for IHO S-44 Order 1 without local control.
Sentinel-2 SDB with about a hundred control depths per image is a usable reconnaissance and change-monitoring tool in these waters to about 10 m; it does not meet S-44 Order 1 without local control
Source-stated
The authors warn against treating accuracy from random splits as operational accuracy.
accuracies from random splits should not be taken as operational accuracy.
Source and contributor context
Saeed, Muhammad, Alhosani, Mohamed H., Mikhaylov, Dmitry, Wahedi, Yasser F. Al · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
No suitable media was found by the supported page extraction.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
arXiv
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Editorial provenance
Version 1. Publication ID: 75c76bce-c0cd-489a-8745-0c5ceb9c4159.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-11T00:55:46.555Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: f573ffda94c8da061880b7319580b08dc34021f1880c381dadc8bcaf86ff6ae1. Source text hash: e6c32cd79e0f500885acba3b62ce2eb5b0fb6fb4196a50942c9a706cfe3784af.
Revision and source-check history
Version 1 · 2026-11-11T00:55:46.555Z
arXiv physics.ao-ph: Satellite-derived bathymetry at Arabian Gulf ports
The authors, Muhammad Saeed, Mohamed H. Alhosani, Dmitry Mikhaylov and Yasser F. Al Wahedi, argue that how satellite-derived bathymetry (SDB) is tested can matter more than which model is used. Against multibeam soundings from six Abu Dhabi port areas, LightGBM’s RMSE rose from 0.57 m with a random pixel split to 1.30 m with spatial blocks and 1.66 m at an unseen site. In transfer to an unseen site, the Stumpf ratio was not beaten by LightGBM, U-Net or Lyzenga. With local control, a hybrid reached 1.11 m RMSE over 0–11 m, but the seabed could not be read beyond 11 m. Clear-water results were better. The authors describe Sentinel-2 SDB with about a hundred control depths per image as useful for reconnaissance and change monitoring to about 10 m, while saying it does not meet IHO S-44 Order 1 without local control. They also caution that random-split accuracy should not be treated as operational accuracy. Their results make validation design central to judging whether these maps can support use beyond surveyed locations. Performance remains limited by water conditions and depth.
Original source
physics.ao-ph · Article
Source details
Original source date: Oct 24, 2026 (2026-10-24).
Date basis: Date basis not provided.
Added to Escape Meridian: 2026-11-11T00:55:46.555Z. First observed: 2026-10-27T04:07:02.465Z. Retrieved: 2026-11-02T06:46:41.270Z.
Limits
The evidence is the paper's abstract, which summarizes its methods and results; detailed procedures and results beyond that account are not available here. The authors report that the seabed could not be read beyond 11 m and that summer images were five to six times more turbid than winter ones. The paper reports a 60 m reconnaissance mosaic calibrated cell by cell on ICESat-2 without vessel survey, but the abstract says held-out blocks over half of the product reached 1.42 m RMSE; it does not establish accuracy across the entire mosaic.
Open question
How well would the reported accuracy hold at port areas beyond the six tested, especially where water conditions differ?
Source-stated
The authors compare SDB methods against 285,619 multibeam soundings from twelve surveys across six Abu Dhabi port areas.
against 285,619 multibeam soundings from twelve surveys at six port areas
Source-stated
For LightGBM, the reported RMSE increased as validation moved from a random pixel split to spatial blocks and an unseen site.
the same LightGBM model gave 0.57 m RMSE under a random pixel split, 1.30 m under spatial blocks and 1.66 m at an unseen site.
Source-stated
In transfer to an unseen site, the Stumpf ratio was not beaten by the tested LightGBM, U-Net or Lyzenga models.
In transfer the Stumpf ratio (1.45 m) was not beaten by LightGBM, the U-Net or Lyzenga.
Source-stated
The authors report that a hybrid with local control reached 1.11 m RMSE over 0–11 m, but the seabed could not be read beyond 11 m.
With local control a hybrid of the tree output and the band ratio reached 1.11 m over 0-11 m
Source-stated
The authors characterize Sentinel-2 SDB with about a hundred control depths per image as useful for reconnaissance and change monitoring to about 10 m, but not sufficient for IHO S-44 Order 1 without local control.
Sentinel-2 SDB with about a hundred control depths per image is a usable reconnaissance and change-monitoring tool in these waters to about 10 m; it does not meet S-44 Order 1 without local control
Source-stated
The authors warn against treating accuracy from random splits as operational accuracy.
accuracies from random splits should not be taken as operational accuracy.
Source and contributor context
Saeed, Muhammad, Alhosani, Mohamed H., Mikhaylov, Dmitry, Wahedi, Yasser F. Al · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
No suitable media was found by the supported page extraction.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
arXiv
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Editorial provenance
Version 1. Publication ID: 75c76bce-c0cd-489a-8745-0c5ceb9c4159.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-11T00:55:46.555Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: f573ffda94c8da061880b7319580b08dc34021f1880c381dadc8bcaf86ff6ae1. Source text hash: e6c32cd79e0f500885acba3b62ce2eb5b0fb6fb4196a50942c9a706cfe3784af.
uncertain · 2026-11-11T00:55:46.555Z · anthropic/claude-opus-5.5
Publication: 75c76bce-c0cd-489a-8745-0c5ceb9c4159
attribution — pass: The four named authors match the current arXiv listing. The findings are framed as the authors' results and characterizations. The title's 'arXiv physics.ao-ph' label matches the publisher metadata. The page lists cs.LG as the primary category, but the paper is cross-listed in physics.ao-ph, so the label is not wrong.
Authors:Muhammad Saeed, Mohamed H. Alhosani, Dmitry Mikhaylov, Yasser F. Al Wahedi
summary — uncertain: The reading tracks the abstract faithfully, but central jargon is left unexplained for a general reader. Bathymetry (mapping water depth) is never defined, nor are multibeam soundings (ship sonar depth measurements). RMSE, LightGBM, U-Net, the Stumpf ratio, Lyzenga, 'local control' and IHO S-44 Order 1 are also unexplained. Without these, a reader cannot grasp what was compared or what failing Order 1 means. Smallest repair: add a plain-language gloss for SDB and multibeam surveys, and add term-library tooltips for RMSE, the model names and S-44.
Satellite-derived bathymetry (SDB) is often proposed as a low-cost complement to multibeam surveys of ports
date — pass: The displayed original publication date of 2026-10-24 matches the v1 submission date. It does not use the v2 revision date of 30 Oct or the feed date of 2 Nov.
[Submitted on 24 Oct 2026 (v1), last revised 30 Oct 2026 (this version, v2)]
support — pass: Every claim, figure and quoted conclusion matches the abstract. The inference that validation design is central to judging use beyond surveyed sites follows from the authors' own comparison of validation designs and their warning about random splits. The question about other port areas is a genuine open issue raised by the six-site scope and the turbidity findings. The mosaic limitation is appropriately cautious.
The validation design changed the result more than the model
media — pass: The selected media is the verified link to the arXiv abstract page itself, which is the most direct way to reach the source.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXivarXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXiv