Saudi Red Sea - Thetis (Yanbu)
Part of the regional summaries: Saudi Arabia - Yanbu, Saudi Arabia, and Red Sea
2024-08-09
More info
Al Madinah and Makkah (NOAA CRW virtual station)Thetis is a deep Yanbu wall dive whose two visits tell very different stories: in October 2023, during peak thermal stress (19.9 DHW), the upper wall was heavily built with Porites, Acropora, and pocilloporids, much of it bleached white, and cover measured 33% with 62% of coral tissue bleached; by August 2024 the same wall had gone darker and sparser, dominated by soft corals and sponges with hard coral reduced to patches, cover down to 13% and bleaching to 25%. Cover more than halved between the two visits.
ثيتيس غطسة جدار عميقة في ينبع تحكي زيارتاها قصتين مختلفتين جدًا: في أكتوبر 2023، خلال ذروة الإجهاد الحراري (19.9 DHW)، كان الجدار العلوي كثيف البناء من Porites وAcropora والبوسيلوبوريات، كثير منها مبيضّ، وبلغت التغطية 33% مع ابيضاض 62% من نسيج المرجان؛ وبحلول أغسطس 2024 صار الجدار نفسه أدكن وأقل كثافة، تهيمن عليه المرجانيات اللينة والإسفنج وقد تقلص المرجان الصلب إلى رقع، فانخفضت التغطية إلى 13% والابيضاض إلى 25%. وانخفضت التغطية إلى أقل من النصف بين الزيارتين.
ترجمة هذا الملخص بالذكاء الاصطناعي؛ النص الإنجليزي هو المعتمد.
8 of the 21 photographs on this page are shown beside the text that discusses them under "Site description" below.
Site description
Setting
A deep offshore wall in the Yanbu area, dived twice with Umluj Dive Center — October 2023 to 34 m and August 2024 — and one of the sites with a per-visit series in the site record. The site is a wall dive: a sheer face dropping into deep blue with overhangs and buttresses (e.g., Image 1, Image 4), divers hanging off it in the distance (e.g., Image 8), bubble trails rising against the rock.
General benthic features
In October 2023 the upper wall was heavily built — a continuous carpet of knobby and columnar Porites, corymbose Acropora, lavender pocilloporids, and bladed plates, with red anthias in clouds over it — and much of it was white (e.g., Image 6, Image 7).
In August 2024 the same wall photographs as darker, sparser, and more soft-coral- and sponge-dominated: purple and gray soft corals, scarlet and crimson encrusting sponge, orange bladed plates, feather-star-studded rock, with hard coral reduced to patches (e.g., Image 2, Image 9).
A gray-blue bubble coral (Anderson's identification) with fully inflated vesicles sits at the wall base, and a table Acropora with red sponge growing through it lies on the shoulder above.
Coral cover
- Reference: Mayfield (2025)
- Data file: (Mayfield, 2025-appendix A)
- CoralNet submission package (accession #5546)
- 16% ± 15% SD across 170 photographs.
- The two visits are recorded separately and they are strikingly different: 33% cover with 62% of coral tissue bleached in October 2023 at 19.9°C-weeks DHW on the survey date (6 October 2023; NOAA Coral Reef Watch, Madinah-Makkah (Saudi Arabia) virtual station) and 30.5°C, against 13% cover with 25% bleached in August 2024 at 14.2°C-weeks and 33.5°C.
- Cover more than halved between the two.
Stressors
The August 2024 frames still show partly bleached colonies on the wall (photos not shown), consistent with the published 25%.
Fish and other fauna
Anderson's 6 October 2023 log records a dolphin nearby, and the August 2024 frames show a clownfish in its anemone (photo not shown) and a pufferfish (e.g., Image 21).
Comparisons with other data
Apart from our own published survey shown here, we found no other surveys at this site as of 2026. If you would like to share yours, please email us.
Survey summary
Thetis: surveyed 2 times, Oct 2023–Aug 2024; bleaching ranged from 25.0% to 62.1% of coral tissue across visits (CoralNet point counts; Mayfield 2025, Environments 12:248). Surveyed in partnership with Umluj Dive Center.
CV model comparison
More info
The reference is this dive site's own published CoralNet point count by the author (Mayfield 2025, Environments 12:248, Appendix A; most frames annotated by CoralNet's AI, trained and checked by the author). The computer-vision rows are this site's own values: Claude is a blind point count of all of the site's edited photographs, DINOv3 ran on the site's benchmark frames (the photographs the author annotated point by point), and CoralSCOP is the site's finished survey of its edited photographs. They are on different frame sets from the published count, a row under 30 frames is indicative only, and no computer-vision method yet estimates coral cover accurately on new images.| Method | Hard coral cover (%) | Photos (n) ± SE |
|---|---|---|
| Published point count (CoralNet-assisted) | 16.1 | |
| In-house DINOv3 | too few photos (n=5) | |
| CoralSCOP edited | 20.5 | n = 171 |
| Claude (edited photos) | 12.4 | n = 169 · SE 1.1 |
The two computer-vision methods differ by 8.1 pp (12.4–20.5%) and read within 4.4 pp of the published point count (16.1%); the headline is the published point count because it is the site's published value (CoralNet-assisted: most frames were annotated by CoralNet's AI, trained and checked by the author).
In-house DINOv3: this method was run on 5 of 171 photos, under the 10 needed for a value, so it is not shown and is left out of any mean; every other row rests on its own photo count.
See a more detailed, stand-alone version of this table.
Dive history
Coral bleached: computer-vision (CoralNet) estimate from the survey images, per dive, as published in Mayfield (2025, Environments 12:248), Supplementary File S1.
| Date | Time in | Max depth (m) | Bottom time (min) | Coral bleached (%) |
|---|---|---|---|---|
| 2023-10-06 | 08:00 | 34.0 | 45 | 62.1 |
| 2024-08-09 | — | — | — | 25.0 |
Coral cover by year
Coral Reef Diagnostics surveyData & links
- CoralNet image repository for this region ↗
- Photo portfolio ↗
- Photo portfolio ↗
- Coral Reef Diagnostics country page ↗
All images from Coral Reef Diagnostics. Data are from Coral Reef Diagnostics and NOAA Coral Reef Watch.
© Coral Reef Diagnostics.