High-resolution bathymetric and shallow seismic data along the northeast Red Sea margin reveal a previously disregarded mechanism for carbonate platform drowning at a steep-flanked rift basin. At the seafloor, salt extrusions highlight the influence of extensional salt tectonics, with a salt flow from the southern flank of the Al Wajh carbonate platform that likely originates from below. Salt-flow direction, morphology, and kilometer-sized slumps and rotated blocks indicate platform-margin disintegration and rafting of platform blocks toward the southwest. The outlines of several smaller detached or semi-detached carbonate platforms to the south of the main platform can be refitted to the larger platform margin by counter-moving the direction of mass wasting. Several platforms, reaching heights above the seafloor of up to 650 m, are partially or fully submerged in the mesophotic zone and appear to be in danger of drowning. We conclude that the southern outer rim of the Al Wajh platform is breaking apart owing to salt withdrawal, which indicates that carbonate platforms on top of salt sequences grow on mobile ground, leading to platform disintegration, basinward rafts, and the demise of broken-off, smaller pieces of platform. Salt displacement also controls the growth geometries of individual platform rafts, with keep-up reef growth (growth rate equal to sea-level rise) and drowning occurring in close spatial proximity. Therefore, the interplay between salt diapirism and platform growth is not limited to platforms growing on the apexes of diapirs and is more complex than previously thought.

Drowned carbonate platforms and reefs have been common in the geological record since the Cambrian (Schlager, 1981). They represent valuable archives of palaeo-environment and climate (Weissert et al., 1998) with significant economic importance as hydrocarbon reservoirs on passive margins (Hendry et al., 2021) and in tectonically active regions, such as in Southeast Asia (Vahrenkamp et al., 2004). However, their demise by drowning is still a geological paradox because the production of carbonate factories in the euphotic zone and the related growth and accumulation rates normally exceed rates of eustatic sea-level rise (Schlager, 2005). Several mechanisms for drowning have been proposed. These mechanisms can include insufficient autotrophic carbonate factories that cannot keep up with the relative sea-level rise ensuing from the eustatic sea-level increase and/or subsidence (DiCaprio et al., 2010), and environmental drivers, such as (volcano-) clastic and nutrient input, changes in salinity and biota, platform-margin collapse, climate variability (Weissert et al., 1998; Van Tuyl et al., 2019), and strong currents (Betzler et al., 2021). Mathematical models also indicate a combination of initial water depths and the response time of carbonate to an increase in eustatic sea-level rise as a cause of drowning (Kim et al., 2012). However, the effect of salt tectonics on platforms in rift basins has received little attention until now.

In our study, high-resolution bathymetric and shallow seismic data from the northeast Red Sea (Fig. 1A) reveal a case of active carbonate platform drowning in one of the most prolific carbonate provinces of the world, associated with salt extrusions and platform-margin collapse features (Fig. 1B). Unlike previous static models, these features are interpreted based on a dynamic salt tectonic model that indicates the breakup and rafting of platform fragments toward the center of the Red Sea basin. Therefore, rafting platforms or parts are drowning due to salt displacement and eustatic sea-level change.

The Red Sea has evolved into a slowly spreading ocean basin (Tapponnier et al., 2013). After the onset of continental rifting during the Oligocene, the basin was flooded during the early Miocene, and restricted conditions followed that led to the deposition of an evaporite sequence 2–4 km thick. This sequence includes mobile mid-Miocene salt (Hughes and Johnson, 2005), which forms one of the largest evaporite basins worldwide (Evans, 2006). The evaporites are unconformably overlain by Pliocene-Pleistocene marine carbonates and continental siliciclastics from the basin margin.

In extensional salt tectonic settings like the Red Sea, the overburden is stretched and thinned above the salt, leading to differential loads and areas of mechanical weakness (Jackson and Hudec, 2017). These conditions are ideal for diapirism, which may evolve into down-building structures exposed at the sediment surface (Vendeville and Jackson, 1992). The Red Sea margins are influenced by the basinward dip of the mid-Miocene salt layers (Heaton et al., 1995; Orszag-Sperber et al., 1998; Rowan, 2014; Tubbs et al., 2014) and characterized by steep-sloped shallow-water carbonate platforms and fringing reefs. Their appearance and arrangement were previously interpreted to be controlled by rift tectonics, siliciclastic input through onshore drainage systems (wadis), and salt diapirs (Dullo and Montaggioni, 1998). Salt flows are present along the axis of the basin and at the distal margins within the modern central Red Sea (Mitchell et al., 2010).

The Al Wajh platform is a major landattached carbonate platform in the Red Sea (Fig. 1A; Petrovic et al., 2022). It is located just offshore of the largest drainage basin of western Saudi Arabia (Abdelkareem et al., 2020), which controlled the intermittent input of siliciclastic sediments during the late Pleistocene (Rowlands et al., 2014). The onset of the distal Al Wajh platform margin is influenced by its position on the edge of a tilted rift block (Dullo and Montaggioni, 1998).

Bathymetric data and seafloor imagery were acquired during two research cruises in 2019 with RV Thuwal off Saudi Arabia (Fig. 1B) using a Falcon Seaeye remotely operated vehicle (ROV) and a ship-mounted Kongsberg EM 710-MK2 multibeam echo sounder. Qimera software (https://qps.nl/qimera/) was used for data processing and to calculate a bathymetric grid with a resolution of 25 m, covering water depths of –1800 m to –100 m relative to sea level. The Red Sea Development Company (TRSDC; Riyadh, Saudi Arabia) provided lidar data that were integrated with the bathymetric data and cover water depths of –100 m to +5 m. The merged data were analyzed using Fledermaus Pro Software (https://www.qps.nl/fledermaus/) and visualized using a scientific color scheme (Crameri et al., 2020).

Seismic data were obtained in November 2021 with the RV Explorer (Fig. 1B) using an HMS-620LF Bubble Gun (Falmouth Scientific, Inc.) as an acoustic source and a 24 channel MicroEel analogue streamer (Geometrics, Inc.). The streamer had an active length of 100 m and a group interval of 4.167 m. Data were acquired using Geode and SmartSeis seismographs (Geometrics). Shot distance was set to 16.7 m, and the sample rate to 4 kHz. The seismic data were processed using ProMax software (https://www.landmark.solutions/SeisSpace-ProMAX), including geometry setup, static correction, noisy trace editing, multiple reductions, normal-moveout and dip-moveout correction, spherical divergence, surface-consistent gain correction, and time migration. Before interpretation, the seismic lines were tied to the bathymetry using Petrel software (https://www.software.slb.com/products/petrel).

The seafloor south of the Al Wajh carbonate platform is rugged, consisting of shallow (<120 m water depth) carbonate platforms separated by 450-m-deep linear troughs (Figs. 1B and 1C). The platforms are 1–10 km long and have steep sides facing submarine canyons. In map view, their shapes are scalloped, with concave and convex margins. The shallow platform margins outline interior lagoons, which results in an atoll-like appearance (Fig. 2A). Some of the carbonate platforms are semi-attached to the main Al Wajh platform, and others are isolated (IP in Fig. 2A). The semi-attached platforms have rounded to elongated shapes, whereas isolated platforms are elongated and branching, with sizes ranging from 1 km2 to around 25 km2. The platforms have heights of up to 640 m above the seafloor and steep slopes with angles of 75° to 80° (Fig. 1C). Most platforms are entirely submerged. In contrast, others are largely submerged below 40 m water depth (Fig. 2A). Below this depth, vigorous coral growth is absent, and red algal incrustation is abundant. In contrast, below 100 m, reef-like structures are colonized by light-independent corals, such as dendrophylliidae, and fleshy algae turfs surrounded by bioclastic sand flats, indicating that these platforms have not managed to stay within a zone of vigorous coral growth during the recent sea-level rise and are in danger of drowning.

In contrast, the main Al Wajh platform exhibits healthy reef growth along its rim, with corals thriving down to a 10 m water depth (Petrovic et al., 2022). Only the southern ends of the two southern isolated platforms extend into the euphotic zone and exhibit active shallow-water coral growth (Fig. 2A). The largest isolated platform, with dimensions of ~11 km × 6 km, has a complex morphology (Figs. 2A and 2B). The southern part of the platform has an overall smooth morphology, few internal depressions, and dozens of active pinnacle and patch reefs, and culminates at the southernmost end in a more continuous reef structure. However, toward the north, the morphology is more rugged and characterized by north-south and northeast-southwest plateaus 40 m deep that line elongated depressions and reach water depths of 120 m (Fig. 2B). Normal faults delimit these depressions (Figs. 2B2E). The deep plateau restricted by northwest-southeast–striking faults in line P2 is reminiscent of a northeast-southwest extension and is interpreted as a graben and horst system. The vertical displacement of the platform strata in line P1 appears to be related to normal or lateral movements in the northeast-southwest direction, and the fault is proposed to serve as a transfer fault marking the northwest boundary of the extensional faults.

Toward the south, platform slopes and basin floors (B1–B3) are locally covered by lobate bodies with extensions of several kilometers (Fig. 1B). We interpret these features as active salt extrusions (Mitchell et al., 2010; Smith and Santamarina, 2022). In a proximal position concerning the Red Sea margin, such extrusions (S1–S3 in Fig. 1) are fed from underlying salt diapirs located along basinward-dipping faults (Fig. 1B), as described by Ge et al. (1995). In a distal position (S4), extrusions are fed by passive salt diapirs inside shallow graben systems, a process discussed by Hudec and Jackson (2007).

Sea-level variations, tectonic processes (riftbasin extension and uplift), and fan delta formations might have influenced the initial platform topography on a kilometer scale, as discussed for other platforms in the Red Sea (Bosence, 2005). A significant effect of karstification on the modern platform morphology was proposed for the Al Wajh platform (Purkis et al., 2010). Over the past 1.1 m.y., the Red Sea has been in an arid climate with only short, intermittent periods of pluvial conditions, primarily during sealevel highstands (Nicholson et al., 2020). Hence, the exposed platforms were likely exposed to an arid climate. During pluvial conditions, the platforms were probably covered by seawater. Therefore, karstification is expected to play a minor role in platform evolution. Furthermore, it is unlikely that the complex, kilometer-scale platform morphology results from differential reef growth or karstification of an exposed reef morphology. In addition, differential reef growth makes climatic and oceanographic variations unlikely as controlling factors, considering the scale of the Al Wajh platform.

At the southern Al Wajh platform margin (Fig. 1), an array of curved ridges was identified by a curved, south-dipping detachment fault (Fig. 3A). The hanging wall is sliced by many curved (listric) normal faults. A southwest-northeast–running seismic line (P3 in Figs. 3A and 3B) shows seaward-dipping normal faults cutting into the southern Al Wajh platform margin, whereas the sediments in the subsequent trough dip landward with a clockwise rotation along a listric normal fault. We interpret this as a submarine landslide due to its association with the prominent rotated strata in the trough as deposited in a ramp basin (sensu Rowan, 2014; Jackson and Hudec, 2017). This basin evolved due to active load-related withdrawal (expulsion) of the underlying mid-Miocene salt, which caused a basinward collapse of the salt overburden. In summary, the data indicate fragmentation of the southern platform margin, driven by underlying salt displacement due to heavy loading from sediments accumulating at the Red Sea margin. The shapes of individual platforms resemble dispersed puzzle pieces (see the Supplemental Material1) that have rotated and moved at least 2 km seaward and are separated by curved linear troughs. The absence of recent mass-wasting deposits in the troughs between the platforms might indicate a currently stable tectonic position.

Salt tectonics, especially diapirism in evaporitic basins, has primarily been associated with reef development and providing substrate for platform growth (Dullo and Montaggioni, 1998). Salt diapirs form seafloor highs, which lead to the development of reefs and carbonate platforms after conditions are optimized for shallow-water carbonate production (Bosence, 2005). The growth potential of Triassic carbonate platforms was the greatest in areas of salt withdrawal (Strauss et al., 2020). The results prove that salt withdrawal toward the basin can also result in margin fragmentation and the subsequent partial or total drowning of individual platforms along basin margins for extensional salt tectonic settings (Fig. 4).

Coral growth rates in optimum water depths vary between 5 mm/yr and 16 mm/yr (Dullo and Montaggioni, 1998), and rates of salt dislocation range from >1 mm/yr to 4000 mm/yr (Jackson and Hudec, 2017), potentially outpacin carbonate accumulation. In addition, the fragmentation of platform pieces and differential subsidence leading one part to drown and another to grow (Fig. 2B) indicate that salt flow is not uniform (Jackson et al., 2014), which adds spatial complexity. An intricate basement relief likely influences irregular extension patterns (Pichel et al., 2019). For the case study discussed, this theory awaits confirmation by deep seismic data.

To our knowledge, this process of platform decay and drowning (Fig. 4) has not been recognized anywhere else in platform sequences. We propose that potential ancient analogues are the Albian platforms in the subsurface offshore Angola (Eichenseer et al., 1999) and Gabon (Dupré et al., 2007). These shallow-water carbonates are capped by deep-water facies with drowning and are interpreted to be caused by a eustatic sea-level rise. However, basinward rafting on Aptian salt and the possible demise caused by salt displacement were not considered.

The southern margin of the Al Wajh carbonate platform is fragmenting, with the detached platform pieces drowning or partially drowning. Curvilinear troughs (grabens at depth) separate the platform fragments filled by landward-dipping sediments. The fragments form smaller platforms of their own, rafting on salt layers from their original position toward the deep Red Sea. Several of them have subsided, at least in part, below the euphotic zone, which indicates a danger of drowning. These observations imply that salt displacement must be considered a critical process in basins with thick evaporite sequences that controls platform development and demise in addition to rift tectonics and diapirism.

1Supplemental Material. A supplemental figure showing rafting dispersed isolated platforms of the southern Al Wajh area based on Figure 2A, and restoration of the present-day situation to a pre-extensional state. Please visit https://doi.org/10.1130/GEOL.S.21743408 to access the supplemental material, and contact editing@geosociety.org with any questions.

The project was funded by King Abdullah University of Science and Technology through baseline support to V. Vahrenkamp. We thank The Red Sea Development Company (Riyadh, Saudi Arabia) for providing the shallow-water bathymetric data set and publishing permission. In addition, we thank Halliburton-Landmark (Houston, Texas) and Schlumberger (Houston, Texas) for providing university grants for their software packages. We also thank the three anonymous reviewers who helped to improve this manuscript with constructive comments.

Gold Open Access: This paper is published under the terms of the CC-BY license.