Discovery of the deep-sea NEOM Brine Pools in the Gulf of Aqaba, Red Sea

Deep-sea brine pools are rare hypersaline, typically anoxic depressions known for hosting extremophile microbes and unique biodiversity. They are important natural laboratories for studying life in extreme environments and for preserving undisturbed sedimentary records due to exclusion of bioturbating megafauna. In the Red Sea region, brine pools have been attributed to dissolution of Miocene evaporites and are generally categorized into deep axial-trough pools (>1000 m) and shallower coastal-shelf pools (<850 m). The present study reports the discovery during the 2020 OceanXplorer cruise of the NEOM Brine Pools at 1770 m depth in the Gulf of Aqaba—the first such pools outside the Red Sea proper and only 2 km from shore. The research aims to characterize their physical setting, chemistry, biology, and sedimentology, to determine their provenance and whether they fit established categories or represent a new class. Their nearshore position suggests a unique capacity to archive regional events such as tsunamis, flashfloods, and seismicity.

Prior work documents brine pools in three main regions: the Gulf of Mexico, Mediterranean, and Red Sea, with the Red Sea hosting the largest number. Red Sea brine pools are commonly linked to dissolution of subsurface Miocene evaporites formed during early rifting. Most known Red Sea pools occur along the deep axial trough and are anoxic and warmer than ambient waters; two known shelf sites include the Thuwal Seeps (~860 m) and Afifi (~350 m). These pools host diverse microbial communities and macrofauna and have yielded microbes with bioactive compounds of therapeutic interest. Established chemical signatures include elevated Na+/Cl− ratios approaching halite dissolution values and reduced SO4 2−/Cl− ratios due to bacterial sulfate reduction. Microbial communities in Red Sea pools often include KB1, Bacteroidia, Clostridia, Deltaproteobacteria, Gammaproteobacteria, and archaeal Methanobacteria and Thaumarchaeota. Previous cores from Red Sea pools have been analyzed, but none were close enough to shore to archive terrestrial event beds robustly.

- Geophysical mapping: During the 2020 OceanX "Deep Blue" Expedition aboard the 87 m R/V OceanXplorer, hull-mounted multibeam sonars (Kongsberg EM712 and EM304, 70–100 kHz) provided bathymetry with ≥10 m across-track resolution, calibrated with CTD and XBT profiles. Additional geophysical data included sub-bottom and ADCP profiles. Pools were detected by dampened backscatter with an ROV-mounted 450–900 kHz imaging sonar. - In situ exploration and sampling: A 6000 m-rated Argus Mariner XL ROV and Triton submersibles were used to map pool geometry, collect imagery, and acquire samples. The main pool was located at 28°47.31587′N, 34°47.70937′E. - CTD and water sampling: A Sea-Bird 911+ CTD with a rosette of 12×12 L Niskin bottles was lowered through and below the brine-seawater interface. Water samples were collected 500 m and 5 m above the interface, at the interface, and below it. Profiles of salinity, dissolved oxygen, and temperature were recorded. Brine thickness was confirmed to 6 m in the pool center; the ROV could “land” on the brine surface to choreograph casts. - Laboratory analyses: Salinity via refractometer; chloride by Ag titration (0.1 N AgNO3; IAPSO standard). Total alkalinity and initial pH by potentiometric titration (Gieskes et al. methods). Elemental concentrations measured by ICP-MS/MS (Agilent 8900) with appropriate dilutions; standards included SPEX multielement solutions, IAPSO, and CASS-6. Oxygen isotopes (δ18O) were measured with a Picarro cavity ring-down spectrometer. - Sediment coring: A transect of five push cores was collected across the western rim to sample four concentric zones: (i) hemipelagic slope sediments, (ii) gray microbial “beach,” (iii) orange microbial “swash,” and (iv) brine pool interior. Four cores were 50 cm; one long core was 150 cm (refusal), recovered compressed from 135 cm to 85 cm and later corrected. - Core imaging and CT: The long core was CT scanned (Amira ZIB v2021.03) to reconstruct 3D lithology, grain-size distributions, clast maps, and orthoslices. - Porewater extraction: Rhizon samplers (0.22 µm) were inserted at 25 mm intervals; porewaters analyzed for alkalinity, salinity, chloride, and elements (acidified with 5% HNO3; stored at 4 °C). - XRF and XRD: Avaatech XRF core scanner provided elemental intensities at 2 mm spacing; elemental ratios (Rb/Zr, (Zr+Rb)/Sr, K/Fe) were used to identify siliciclastic inputs and turbidites. XRD on powdered samples at 25 mm spacing quantified mineralogy, emphasizing quartz+feldspar abundance. - Radiocarbon dating and age modeling: Six bulk sediment horizons from the long core were dated at Beta Analytic and calibrated using Marine20 with a local reservoir correction (ΔR = −447 ± 4) derived from U-Th coral ages. Bayesian age-depth modeling used the Bchron package; linear sedimentation rates were also calculated between tie points. High-density grain-flow layers (rapid depositional events) were excluded from sedimentation-rate calculations. - Organic geochemistry: δ13C and TOC/TN were measured on 27 samples; decalcified for organic C and N. Source of organic matter inferred from δ13C vs C/N domains. - Microbial community profiling: The top 20 mm of each core was sequenced for 16S rRNA; relative abundances of major aerobic and anaerobic taxa were assessed to delineate lateral gradients across zones. Contamination minimized by removing the outer surface of cores.

- Discovery and setting: A complex in the Aragonese Deep (Gulf of Aqaba) at 1770 m water depth includes one large pool (260 m × 70 m; ~10,000 m²; brine thickness up to 6 m) and three minor pools (<10 m²) within 50 m of the main pool. The complex lies adjacent to the Saudi toe-of-slope near the intersection of the Arona (strike-slip) and a NNW-trending normal fault; the Aragonese fault bounds the basin to the west. - Hydrography and chemistry: Ambient bathyal water above the pool is ~21.33 °C, 40 PSU, and ~180 µmol L−1 O2. Across the brine interface at 1769.46 m, salinity increases to ~160 PSU within 15 cm; dissolved O2 drops >75% at the interface to ~50 µmol L−1 by 20 cm into the brine and to <10 µmol L−1 by 50 cm (effectively anoxic). Temperature increases gradually by ~1.0 °C at 2 m into brine, indicating no proximal hydrothermal heating. - Brine provenance: Microbial staining and visual observations show multiple seep points along the toe-of-slope (up to 5 m above the pool), consistent with continuous brine seepage feeding the pool. Geochemically, NEOM brine exhibits low SO4 2−/Cl− (~2 mM/M) and elevated Na+/Cl− relative to seawater, consistent with dissolution of subsurface halite and pronounced bacterial sulfate reduction. δ18O of the brine is ~1‰, far lower than the ~16‰ expected for evaporation to 160 PSU, ruling out an origin by evaporation of modern seawater. - Rim zonation and biology: Four concentric zones occur: outer hemipelagic mud; gray microbial “beach” (periodically inundated by brine); orange microbial “swash” at the interface; and the brine pool proper. The bivalve Apachecorbula muriatica densely inhabits a ~20 cm belt at the interface; shrimp (Plesionika sp.), an eel (possibly Ariosoma), bigeye houndshark (Lago omanensis), and a flatfish (cf. Cynoglossus acutirostris) exploit the brine margin for feeding on stunned prey. - Microbial gradients: 16S rRNA profiling shows aerobic taxa (Gammaproteobacteria, Thaumarchaeota, Alphaproteobacteria, Nitrospira) dominate outside the pool and in the gray zone, while anaerobes (e.g., Deltaproteobacteria—sulfate reducers; fermenters; dechlorinators; methanogens) dominate the orange swash and pool interiors, aligning with brine anoxia and low SO4 2−/Cl−. - Stratigraphy and chronology: A 150 cm core beneath the brine records at least 1.2 kyr of undisturbed deposition (no age reversals). Average sedimentation rates (excluding rapid grain-flow layers) range 0.5–4.5 mm/yr. Three depositional motifs: • Coarse siliciclastic intervals (30% of core): angular sand-to-granule quartz-feldspar-rich lenses, including a 30 cm-thick event deposited ~525–650 yr BP. Ten such coarse intervals over the last ~1000 years, with the youngest ~50 yr BP. These are interpreted as tsunamites; the youngest aligns within dating uncertainty with the 1995 Mw ~7.2 Nuweiba earthquake and modest tsunami. • Stacked fining-upward turbidites (55%): mm–cm Bouma-like sequences of clay-to-silt with >30% carbonate and TOC 0.4–0.8%, attributed to turbidity currents/hyperpycnal flows from flashfloods; at least 50 sequences indicate recurrence ≥4 times per century. • A 20 cm tan clay with silt streaks (15%) deposited continuously between ~750–900 yr BP with highest carbonate (50–60%). - Chemostratigraphy and mineralogy: Elevated Rb/Zr, (Zr+Rb)/Sr, and K/Fe ratios mark coarse siliciclastic intervals; ranges: Rb/Sr 0.01–0.09; (Zr+Rb)/Sr 0.18–1.77; K/Fe 0.19–1.83, peaking at 48–77 cm (~600 yr BP). Quartz+feldspar abundance varies 11.4–89.1%, maximized in the thick coarse layer. TOC ranges 0.07–0.79% (avg 0.31%); TOC correlates strongly with TN (r2=0.99) and δ13C averages −19.71‰, indicating predominantly marine algal organic matter. - Tectonic context and sediment dynamics: The site experiences frequent seismicity (e.g., 1993 swarm; 1995 M7+ event) and lies on the steepest and deepest foreslope in the Gulf, favoring seismically induced turbidites and potential tsunami generation/impact. The sharp halocline may reflect recent brine emplacement potentially linked to fault activity. - Preservation: Anoxia eliminates bioturbation and brine density shields the bed from contour currents, enabling exquisite preservation of mm-scale event beds, making the pools exceptional archives of coastal processes and tectonic events.

The study set out to determine whether the newly discovered Gulf of Aqaba brine pools conform to known Red Sea categories or represent a distinct class, and to evaluate their potential as sedimentary archives of nearshore events. Findings show that the NEOM Brine Pools are the first brine pools identified outside the Red Sea proper and uniquely close to the coast (~2 km), fed by toe-of-slope seepage rather than direct association with axial spreading or shallow shelf seeps. Geochemically, they reflect subsurface halite dissolution and intense bacterial sulfate reduction, differentiating them from evaporative origins. The physical setting at a tectonically active pull-apart basin margin and their nearshore proximity facilitate delivery and preservation of terrestrial sediments and event beds. Core data reveal a high-resolution, at least millennial-scale record comprising frequent fine turbidites from flashflood-driven hyperpycnal flows, recurring coarse siliciclastic beds interpreted as tsunamites (with plausible ties to known earthquakes/tsunamis such as 1995 Nuweiba and a ~500–600 yr BP event from Tiran Straits), and deformation structures consistent with seismic shaking. The brine’s density stratification likely decouples incoming turbidity energy from the bed, aiding lamina preservation, while anoxia eliminates bioturbation. Microbial and metazoan communities mirror redox and salinity gradients, aligning with known Red Sea pool ecologies, yet the site’s biodiversity and ecological interactions (e.g., predators exploiting the brine margin) underscore its ecological significance. Collectively, the NEOM pools represent a new nearshore bathyal category, and their exceptional preservational conditions render them valuable archives for paleoclimatology and seismology in the Gulf of Aqaba.

The NEOM Brine Pools are the first deep-sea hypersaline brine pools discovered in the Gulf of Aqaba and the closest to shore known in the greater Red Sea system. They are fed by slope seepage and exhibit chemistry indicative of dissolved Miocene evaporites and strong microbial sulfate reduction. Their proximity to land, together with anoxia and strong density stratification, creates ideal conditions to trap and preserve event-driven sediments. A continuous, at least 1.2 kyr stratigraphy contains numerous fine turbidites from flashfloods, coarse siliciclastic beds likely deposited by tsunamis linked to regional seismicity, and intervals showing deformation consistent with earthquake shaking. These pools extend the geographic and genetic understanding of brine pool formation and demonstrate their utility as high-fidelity archives of climatic and tectonic events. Future work should include repeated temporal hydrographic surveys to assess brine dynamics, expanded coring to refine event chronologies, multiproxy discrimination of hyperpycnites vs seismites, and integrated tsunami and seismic modeling to link dated deposits with historical events.

- Radiocarbon dating used bulk sediments; recycled carbon effects cannot be unequivocally excluded, and calibration for material younger than ~100 years is limited, adding uncertainty to the youngest event ages. - The dissolved oxygen probe has a precision of ±18 µmol L−1, making it difficult to distinguish fully anoxic from severely hypoxic conditions near the interface. - The CTD conductivity sensor saturated at 120 PSU; true brine salinity (160 PSU) required lab determination, and vertical resolution at the halocline was insufficient to fully resolve the diffusion profile. - Assumption that all measured sulfur was present as sulfate may overestimate SO4 2− in near-anoxic brine where some S may be H2S; SO4 2−/Cl− could be even lower. - Molecular data from highly saline sediments may include preserved (“pickled”) rRNA, potentially biasing representation of active microbial communities. - Limited number of cores (five, with one long core) provide a localized record; broader spatial sampling would improve representativeness. - Discriminating hyperpycnites (flashflood deposits) from seismically induced turbidites (seismites) was beyond the scope and remains to be addressed by detailed event correlation and modeling. - Pools lacked a distinct multibeam signature, complicating remote detection and necessitating ROV-based mapping.