i 5-year mean max. DHW is the mean of the annual maximum DHWs over the five complete years named above, not conditions today. More infoGulf of Oman (NOAA CRW virtual station)
Photographs
314
23 shown on this page
Site summary
Read this summary in:
Novice Bay is a strongly zoned Muscat reef, its shallow flat of brain corals and Pocillopora thickets grading into leather-coral slopes and then deep coralline buttresses, with a resting green turtle, a banded sea snake, and a camouflaged scorpionfish among the highlights. Published hard coral cover is 39% (mixed manual and automated point counts), with soft coral at 22%; hard coral bleaching was a trace (0.2% of points), and Anderson photographed bleached leather corals, weeks before the region's July 2024 bleaching event began in earnest.
خليج المبتدئين شعاب في مسقط واضحة التقسيم، يتدرج مسطحها الضحل من المرجانيات الدماغية وأجمات Pocillopora إلى منحدرات المرجان الجلدي ثم إلى دعامات مرجانية عميقة، ومن أبرز ما فيها سلحفاة خضراء مستريحة وثعبان بحر مخطط وسمكة عقرب متخفية. والتغطية المنشورة للمرجان الصلب 39% (عدّ نقاط يدوي وآلي معًا)، والمرجان اللين 22%؛ وكان ابيضاض المرجان الصلب أثرًا ضئيلًا (0.2% من النقاط)، وصوّر أندرسون مرجانيات جلدية مبيضّة، قبل أسابيع من بدء حدث الابيضاض الإقليمي في يوليو 2024 فعليًا.
ترجمة هذا الملخص بالذكاء الاصطناعي؛ النص الإنجليزي هو المعتمد.
Sea surface temp. (satellite)
30.3°C
satellite sea surface temperature (SST) on the survey date (NOAA Coral Reef Watch, Gulf of Oman), not an in-water reading; it is the warmest-pixel (90th percentile) reading in the station area and can run 1–1.5°C above the reef Published estimate, not a logged temperature: Mayfield (2025, Platax) reports an extreme thermocline at the four Muscat sites, about 20-21°C 1-3 m above the bottom at 5-12 m, from Anderson's dive computer (the data were later lost).
Seven independent survey stations shown with the program’s logo — click one to see how far it is from this site.
One other CRD dive site with its own page within 5 km shown in orange — click a marker to open its page.
3 of the 23 photographs on this page are shown beside the text that discusses them under "Site description" below.
Dived 2024-05-20, Bandar Khayran / Muscat, Oman — one of the seven 2024 Oman survey sites reported in Mayfield (2025), Platax.
A strongly zoned reef below a barren Muscat headland. Two of the twenty photographs are surface shots showing the dive boat moored against a bare, tan, scrub-dotted rock headland under a cloudless sky — they confirm the setting directly and leave eighteen underwater frames.
This description is a full review of all twenty photographs from the dive.
Benthos
A shallow flat of brain corals, lobophylliids, and Pocillopora thickets grades into slopes carpeted by Sarcophyton-type leather corals, then deep coralline buttresses and a rubble plain with sea whips.
All three zones are represented across the twenty frames reviewed, and the transitions are sharper than four frames could show. The shallow flat frames are dominated by meter-scale meandering brain coral domes standing shoulder to shoulder; the mid-slope frames are close to a monoculture of cream and gray leather corals with only brain and Pocillopora heads emerging through them; the deep frames are dim, magenta-and-mauve encrusting sheets on coralline buttresses falling away into blue.
The lobophylliid brain corals are photographed in close-up: two deep-maroon, thick-walled colonies pressed together on the flat, their fleshy ridged polyps unmistakable against surrounding tan branching coral.
Pocillopora thickets carpet whole frames in dense lavender and gray-purple, with cauliflower heads packed edge to edge.
A large continuous crimson coral field appears in one frame — deep red, knobby, and unbroken across the entire shot, an unusually vivid patch even by this site's standards.
Soft corals are more varied than the leather corals alone: one frame is dominated by rust-red and orange bushy soft corals of the Dendronephthya type, and a blue-purple sea whip stands off the rubble in another.
Deeper frames add foliose plating corals in whorled, lettuce-like tiers alongside the encrusting sheets and pale leather corals, in noticeably lower light.
One frame is the site's diversity showcase: leather corals, brown branching Acropora, a vivid yellow-green branching Acropora, gray brain corals, and encrusting sheets all within a single frame width.
Stressors
Beyond that trace and the bleached leather corals in Anderson's photographs, no bleaching was recorded at this site during the May 2024 survey, before the seasonal SST peak; NOAACoral Reef Watch recorded Alert Level 2 heat stress and bleaching was reported along this coast from July 2024, weeks after the surveys.
Bleaching was a trace: 0.2% of hard coral points.
The twenty-frame set shows no clear candidate for that trace: the small branching colony on the upper margin of the crimson coral field is pale gray but pigmented, like the colonies around it.
Typical recent thermal stress: mean annual maximum of 11.0°C-weeks (Degree Heating Weeks, 2021-2025, NOAA Coral Reef Watch; Gulf of Oman virtual station).
On the leather-coral slope, a large stark-white ruffled leather coral sits beside tan colonies of the same form, and a pale lobed leather coral stands among ruffled ones; Anderson calls these bleached soft corals (e.g., Image 21, Image 22). The published survey (Mayfield 2025, Platax) lists bleached soft corals among the point-count labels not seen in its photographs; the bleached leather corals here are Anderson's own photo observation, and the published values are unchanged. The brain coral domes and the Pocillopora thickets show no paling.
A crown-of-thorns starfish was photographed on this dive, lying on coralline-crusted dead framework beside clumps of soft coral (e.g., Image 23); no fresh white feeding scars are visible beside it.
Highlights included a resting green turtle, a banded sea snake, an ocellated nudibranch, and a camouflaged scorpionfish.
The ocellated nudibranch is the standout photograph of the set: a pale mint-green dorid covered in raised tubercles, each ringed in bright orange, with a black-and-orange mantle margin and two erect rhinophores, crawling over an encrusted rubble patch dotted with orange sponges.
A school of yellowfin goatfish was photographed in open blue water above the reef, and small cardinalfish shelter within the Pocillopora thickets.
A large sea cucumber lies stretched across the top of a Pocillopora thicket in one frame, with others on the adjacent sand.
Assorted wrasses, damselfish, and a butterflyfish move over the coralline buttresses in the deeper frames.
Published hard coral cover 39% (Mayfield 2025, Platax, Table 1; CoralNet point counts, mixed manual + automated, 30 points per photo, hard coral as a share of benthic points with water and fish excluded), with soft coral 22% and turf algae 12% on the same denominator.
The twenty frames make that 22% soft-coral figure look entirely credible at this site, unlike at Daymaniyat Site 1: whole frames here really are carpeted in leather corals and bushy red soft corals, with hard coral reduced to emergent heads.
Coral cover — external cross-reference (Reef Check). Two independent measurements, reported side by side and deliberately not averaged. Anderson Mayfield / CRD: 39% hard coral (2024). Reef Check: 14% at Novice Bay, Bandar Khayran, 0.06 km away (2017-2017, 1 survey). The two are 7 years apart (CRD later). Methods differ (CRD photographic/point-count index vs Reef Check point-intercept transect). Reef Check site positions are approximate; even with the positions coinciding, the two surveys may have sampled different patches of the same reef, which is the main caveat on this comparison.
Zonation is a further caveat on any single site figure: a frame from the brain coral flat and a frame from the leather-coral slope would give completely different hard coral answers.
How to read the number
The published 39% is a point count (mixed manual + automated) from Anderson Mayfield's own published work and carries the provenance of the paper it appears in.
CoralSCOP works by summing the area fractions of its segmentation masks, and it segments coral-shaped objects. Standing dead skeleton scores as coral, so recently dead colonies can count as live coral to a mask-based method.
The heavy soft-coral cover is the other hazard at this site: leather corals and Dendronephthya are neither hard coral nor bare substrate, and how each method treats them will drive much of the disagreement between the four.
Neither the computer vision outputs nor the photographic point counts are diver transects.
CV model comparison
Computer-vision estimates from this site's own photographs are listed in the comparison table below, alongside the headline rather than merged into it.
Headline is the reference value; every other row is listed beside it for comparison and none is averaged in. ⚠ The published Oman cover rests on 661 frames: 260 scored by hand and 401 classified automatically by CoralNet (Mayfield 2025, Platax, Table 2); the paper analyzed the images in random order across the seven sites and does not report the split for any one site. More infoCave Site is the ONLY real CoralNet comparison in this region. The two CoralSCOP values are paired — a photo counts only if both its edited and unedited versions were scored, so the edited-minus-unedited difference is not a frame-selection artifact. CoralSCOP cover is a mask-sum AREA fraction; no point denominator applies to it. The Claude value comes from blind scoring of edited photos (15 frames per site, 13 at Daymaniyat site 3); a separate, harder blind test on unedited photos gives a different, non-interchangeable number. CoralNet automated point classifier: n/a — every frame at this site was hand-scored by the author, not by CoralNet's classifier, so there is no CoralNet value.
The four computer-vision methods disagree by 24.3 pp (23.0–47.3%) and read on both sides of the published point count (38.5%); the headline is the published point count because it is the site's published value (mixed manual + automated: the paper scored 260 of its 661 Oman frames by hand and 401 with CoralNet's AI).
*Based on only 15 reef-habitat photos; this estimate is less certain.
Hard coral cover over time — this site and every independent station within 5 km. Programs are shown separately because they use different methods at different stations; the lines are not one comparable series.
Within 5 km, 7 independent survey stations recorded hard coral cover in 14 surveys between 2017 and 2026. The one station surveyed in more than one year increased by 10 points between its first and latest surveys. The longest record, BK big mount (submitted to MERMAID by Alhussaini A.*, 0.78 km away), went from 60% in 2024 to 70% in 2025. Since 2021, station-year means average 47% (range 34–70%; n = 7 station-years from 13 surveys). This site's own estimate (39%) sits within the recent range of station-year means up to 2 km away (34–70%); the methods differ, so the comparison is context, not a like-for-like test.
Collected by other programs using different methods (Reef Check uses point-intercept transects; MERMAID is a data platform: each MERMAID row names the organization that did the survey, and its method is in the row's hover text. *Conducted by citizen scientists), so these are context for the figures above, not a like-for-like comparison. Where a station was surveyed more than once, the cell shows either the range of its values or, as a single value, the mean of surveys made in one year. Station names are not dive-site names.