A submarine landslide, also known as a subaqueous landslide, is a large-scale downslope movement of sediment, rock, or both along the seafloor, typically occurring on continental slopes or other inclined marine margins where gravitational forces exceed the material's shear strength.[1] These events can involve immense volumes of material—ranging from thousands to tens of thousands of cubic kilometers—and may travel distances exceeding 100 kilometers, far surpassing most terrestrial landslides in scale.[2] Often triggered by earthquakes, they displace overlying seawater, generating tsunamis that pose significant hazards to coastal populations.[3]Submarine landslides are a fundamental geomorphic process shaping ocean basins, redistributing vast quantities of sediment from shelves to deep-sea fans and contributing to global carbon and nutrient cycles.[1] They occur worldwide, particularly along tectonically active margins like the Pacific Ring of Fire and passive margins such as the Atlantic, with evidence preserved in seafloor bathymetry, seismic profiles, and sediment cores dating back millions of years.[2] Common triggers include seismic shaking, which reduces sediment strength; rapid sediment loading from rivers or glacial melt; oversteepening of slopes by erosion or tectonic uplift; and dissociation of gas hydrates, leading to excess porepressure.[1] Other factors, such as storm waves or volcanic activity, can initiate smaller events nearshore, while larger failures may propagate as debris flows or turbidity currents.[4]The hazards from submarine landslides extend beyond tsunamis, which can amplify wave heights dramatically—reaching tens of meters locally and propagating across oceans—to include damage to seafloor infrastructure like submarine cables, pipelines, and offshore oil platforms.[2] Notable historical examples illustrate their destructive potential: the 1929 Grand Banks earthquake off Newfoundland triggered a submarine landslide that severed transatlantic telegraph cables and generated a tsunami killing 28 people[5]; the 1998 Papua New Guinea event, caused by a magnitude 7.0 quake, produced a 15-meter tsunami that claimed over 2,200 lives[2]; and prehistoric giants like the Storegga Slide (~8,150 years ago) off Norway displaced enough material to inundate coasts across the North Atlantic,[6] as well as more recent events such as submarine landslides triggered by the 2024 Noto Peninsula earthquake in Japan[7] and the 2025 Tracy Arm landslide in Alaska, which generated a local tsunami.[8] While small landslides occur frequently (potentially annually in some regions), massive events are rarer, with recurrence intervals of thousands to tens of thousands of years, though climate change and human activities may increase risks through sea-level rise and coastal development.[2] Ongoing research integrates geophysical mapping, numerical modeling, and paleoseismic data to improve hazardassessment and mitigation strategies.[1]
Fundamentals
Definition and Characteristics
A submarine landslide is a gravity-driven mass movement of sediment or rock on the seafloor, initiated when downslope forces exceed the resisting shear strength of the material, often along weak layers such as overpressured sediments or fractured rock.[9] These events are ubiquitous along continental margins, volcanic island flanks, and submarine deltas, occurring at water depths ranging from shallow coastal zones to abyssal plains.[10] Unlike subaerial landslides, which typically require steeper slopes, submarine variants frequently develop on gentle gradients below 2°, facilitated by the buoyant support of water that reduces effective stress and promotes failure over vast distances.[9][11]Key characteristics include their immense scale and mobility: individual events can mobilize volumes of hundreds to thousands of cubic kilometers—up to 20,000 km³ in extreme cases like the Agulhas Slide—far surpassing most terrestrial landslides by three orders of magnitude.[10][9] Runout distances often exceed 140 km, and in some instances surpass 1,000 km, with transformation into high-velocity flows such as debris avalanches or turbidity currents enabling rapid propagation at speeds potentially reaching tens of meters per second.[11] Submarine landslides exhibit diverse morphologies, including translational slides (coherent block movement), rotational slumps (curved failure surfaces), and fragmented debris flows (matrix-supported clast transport), depending on material cohesion, water content, and substrate properties.[9] They are identified through geophysical surveys revealing headwall scarps, evacuation zones, and lobate deposits, often using multibeam bathymetry and seismic reflection profiles.[11][12]These landslides play a critical role in marine geomorphology by transferring vast sediment loads from shelves to deep basins, shaping seafloor topography over geologic timescales.[10] Their low-frequency but high-impact nature distinguishes them, with recurrence intervals spanning thousands to millions of years for large events, though smaller failures are more common on steep slopes exceeding 5°.[12]Buoyancy and hydrodynamic forces unique to the subaqueous environment allow for prolonged flow durations and complex transformation pathways not typically seen on land.[11]
Global Occurrence and Scale
Submarine landslides are ubiquitous features of the seafloor, occurring along continental margins, volcanic island slopes, and intra-slope basins worldwide, with documented evidence spanning all major ocean basins from polar to equatorial regions. They are particularly prevalent in areas of rapid sediment accumulation, such as glaciated margins (e.g., off Norway and Alaska), river-dominated systems like the Amazon and Mississippi fans, and tectonically active zones including the Indonesian and Japanese subduction zones. A global compilation of large submarine landslides (volumes >1 km³) identifies 68 events over the past 175,000 years, distributed across diverse settings: 15 on glaciated margins, 36 on river fans, and 17 on sediment-starved or other slopes. Smaller landslides are far more abundant; inventories from high-resolution seafloor mapping reveal thousands of features globally, with the Euro-Mediterranean Submarine Landslide Database cataloging nearly 8,000 events from the Late Miocene to the present across European margins and adjacent basins. These distributions highlight how advances in multibeam bathymetry have unveiled previously unrecognized widespread occurrence, especially for smaller-scale failures.The scale of submarine landslides spans six orders of magnitude in volume, from micro-slumps of <10^{-6} km³ to colossal events exceeding 10^4 km³, with over half of documented cases falling between 0.001 and 0.1 km³. Their size-frequency distribution follows a power-law relationship with an exponent of approximately -0.5, meaning small landslides are numerous but giant ones dominate total mobilized sediment. The Storegga Slide off Norway, dated to ~8,200 years ago, exemplifies extreme scale, evacuating 2,400–3,000 km³ of material across an area of ~95,000 km² and with a runout exceeding 800 km, making it one of the largest known mass movements on Earth. Similarly, the 1929 Grand Banks landslide displaced ~200 km³, propagating over 700 km along the Sohm Abyssal Plain and generating a transatlantic tsunami. Areas affected range from <1 km² for localized slumps to >10^5 km² for basin-scale failures, underscoring their capacity to reshape continental slopes dramatically.Frequency analyses indicate that large submarine landslides (>1 km³) occur episodically, with an average recurrence of ~2 events per millennium during periods of post-glacial sea-level rise, but without strong global clustering tied to sea-level fluctuations, earthquakes, or climate cycles—consistent with a random Poisson process. Smaller events are more frequent, potentially recurring on decadal to centennial scales in active regions, though detection biases limit precise global rates. Overall, submarine landslides drive a substantial portion of global sediment flux to the deep ocean, with associated turbidity currents transporting volumes rivaling or surpassing the ~19 Gt/year delivered by all rivers combined; individual events, like the 1929 Grand Banks flow, can carry up to 25 times the annual riverine input. This episodic transport influences abyssal plain sedimentation, carbon burial, and geohazards far beyond their source regions.
Causes and Triggers
Internal Geological Factors
Internal geological factors refer to intrinsic properties of the seabed and underlying sediments that precondition slopes for failure, independent of external dynamic triggers such as earthquakes or storms. These factors primarily involve sediment characteristics, porefluid dynamics, and structural weaknesses that reduce shear strength along potential failure planes.[13]One key internal factor is the composition and consolidation state of sediments, particularly in rapidly deposited fine-grained materials like clays and silts, which exhibit low shear strength due to high water content and incomplete consolidation. Such underconsolidated sediments, often formed during highstand or lowstand phases of sea-level change, can develop excess pore pressures that decrease effective stress and promote instability. For instance, in areas with sedimentation rates exceeding 1-2 m per millennium, these conditions can lead to slope angles as low as 1-2° becoming susceptible to failure.[12][14]Weak layers within the sediment column, such as clay-rich contourites or gas hydrate-bearing zones, further exacerbate vulnerability by acting as low-friction detachment surfaces. These layers form due to diagenetic processes or biogenic gas accumulation, creating planes where failure initiates retrogressively. In the Kumano Basin, for example, overpressured mud layers associated with rapid Pliocene deposition have been identified as critical weak zones in large-scale slides.[15][14]Tectonic structures, including faults and folds, contribute by oversteepening slopes and fracturing bedrock, which facilitates the development of concave or planar rupture surfaces. In tectonically active margins, such as offshore southern California, approximately 75% of submarine landslides occur within 10 km of Quaternary faults, where differential uplift creates gradients exceeding 5-10° and promotes sediment instability. These internal features often result in landslides volumes reaching thousands of cubic kilometers, as seen in prehistoric events where fault reactivation tilted strata without seismic input.[12][13]
External Dynamic Triggers
External dynamic triggers refer to transient external forces that destabilize submarine slopes, often acting on preconditioned sediments to initiate failure. These triggers contrast with internal geological factors by involving rapid, episodic loading or shaking from environmental or anthropogenic sources. They are critical in submarine landslide dynamics, as they can rapidly elevate shear stresses or reduce effective stresses, leading to liquefaction or cyclic degradation of slope materials.[16]Earthquakes represent the most prominent external dynamic trigger for submarine landslides, particularly those of moderate to high magnitude (M > 6.5) that generate intense ground shaking. Seismic waves propagate through the seabed, increasing pore water pressures and inducing cyclic loading that diminishes frictional resistance along failure planes. This mechanism is well-documented in historical events, such as the 1929 Grand Banks earthquake (M 7.2), which triggered a massive submarine landslide off Newfoundland, Canada, displacing over 200 km³ of sediment and generating a tsunami that killed 28 people. More recently, the 2024 Noto Peninsula earthquake (M 7.6) in Japan caused multiple submarine landslides along the Toyama Deep-Sea Channel at depths of 1,540–1,700 m, with headwall scarps up to 45 m high, as evidenced by bathymetric surveys and cable damage from resulting turbidity currents. These events highlight how earthquakes can trigger landslides even on slopes with low gradients (<2°), provided weak layers like overpressured clays are present.[16][17][18]Storm-induced waves and currents serve as another key external trigger, primarily affecting shallow-water slopes (<200 m depth) where wave orbital velocities impart significant shear stresses on the seabed. Intense storms generate cyclic loading that can liquefy cohesionless sediments, similar to seismic effects but limited to nearshore environments. A notable example is Hurricane Camille in 1969, which triggered submarine landslides in the Mississippi Canyon, Gulf of Mexico, damaging offshore oil platforms through wave-induced failures on deltaic slopes. Such triggers are exacerbated in regions with high sediment accumulation, like continental shelves during tropical cyclones, where wave heights exceeding 10 m can exceed critical shear thresholds for slope stability.[16][19]Anthropogenic activities constitute an emerging class of external dynamic triggers, often inducing localized but impactful landslides through direct seabed disturbance or rapid loading. Construction projects, such as dredging, trenching for pipelines, or port expansions, can remove lateral support or increase pore pressures, precipitating failures. The 1979 Nice submarine landslide, off the French Riviera, was linked to underwater sand deposition for airport extension, which overloaded the slope and caused an approximately 10 million m³ (10 × 10^6 m³) slide that generated a localized tsunami affecting coastal infrastructure.[20] Similarly, the 1996 Finneidfjord event in Norway involved blasting and excavation for a harbor, triggering a rockslide that entered the sea and produced a 10 m run-up tsunami. These cases underscore the role of human interventions in amplifying natural instabilities, particularly in engineered coastal zones.[16]
Processes
Initiation and Failure
Submarine landslides initiate when the shear stress induced by gravitational forces along a potential failure plane exceeds the shear strength of the sediment, often preconditioned by geological factors such as the presence of weak layers or rapid sediment accumulation on continental slopes.[21] These weak layers, typically composed of overpressured clays or gas hydrate-bearing sediments, reduce the overall stability, allowing even gentle slopes (less than 2°) to fail over distances exceeding 100 km.[9] Transient triggers, including earthquakes, storm waves, or undercutting by turbidity currents, provide the additional loading or strength reduction needed to surpass the critical threshold, as observed in events where seismic shaking amplifies pore pressures.[9] For instance, in the 2020 Congo Canyon landslide, failure began at the slope toe due to erosion by a high-velocity turbidity current (~5 m/s), highlighting how external hydrodynamic forces can initiate collapse without significant seismic input.[22]Failure mechanisms primarily involve progressive weakening and deformation within the sediment mass, leading to either localized or extensive mobilization. Shear band propagation (SBP) is a dominant mode, where a shear band forms along a weak layer and extends upslope, driven by the release of gravitational potential energy that overcomes residual shear strength at the band's tip.[21] This process can propagate hundreds of meters, as modeled in numerical simulations using large deformation finite element analysis, where the band's growth depends on the softening index of the sediment (ranging from 1 for brittle to infinity for ductile behavior).[21] In contrast, slab failure occurs when the sliding layer breaks into rigid blocks after initial shear band development, limiting retrogression if the layer's resistance exceeds that of the underlying weak zone, as seen in the Licosa Slide in the Tyrrhenian Sea.[21]Retrogressive failure, a common upslope-extending process, involves the formation of horsts and grabens through extensional tectonics, where the collapsing headwall creates a tensile zone that pulls additional material into motion.[21] This mechanism is quantified by criteria such as the critical drop ratio (e.g., 0.574 for certain clay slopes), beyond which the failing mass cannot support further extension without additional weakening.[21] In the Congo Canyon case, retrogression proceeded in discrete pulses over 15 minutes, with velocities escalating from 1.5 to 5.8 m/s across 125–350 m, interspersed with brief downslope reversals, indicating a hybrid initiation that combines toe undercutting with progressive upslope collapse.[22] Overall, these failures transition from stable to catastrophic states via a "progressive failure" phase, where incremental strain softening accumulates until a fully softened zone forms, enabling rapid mobilization.[9]
Propagation and Flow Dynamics
Submarine landslides propagate through a series of dynamic processes that transform initial slope failures into extensive downslope movements, often spanning tens to hundreds of kilometers. Propagation typically begins with retrogressive failure, where the headwall recedes upslope as material fails sequentially, driven by excess pore pressure and reduced shear strength in underlying weak layers. This evolution can lead to translational sliding of intact blocks or disintegration into fragmented debris, with flow velocities reaching up to 68 m/s along the major axis in modeled scenarios.[23] The process is highly influenced by bathymetric confinement, which channels flow and promotes acceleration on steep slopes (>5°), while gentle gradients facilitate prolonged run-out.[16]Flow dynamics are governed by the rheology of the mobilized sediment and interactions with ambient seawater, resulting in distinct regimes such as cohesive debris flows and dilute turbidity currents. In clay-rich flows, high cohesion and viscosity enable long-distance propagation on low slopes (1-3°), with hydroplaning at the base reducing basal friction and enhancing mobility, as quantified by lower H/L ratios (height of fall to run-out distance) compared to subaerial landslides.[16] Sand-rich flows, conversely, exhibit granular behavior with vertical mixing and particle segregation, transitioning into turbidity currents where suspended sediment concentrations (0.1-7%) drive self-acceleration through downslope gravity and flow-induced turbulence.[16]Waterentrainment further dilutes the flow, promoting deposition as velocity decreases.[24]Numerical models provide critical insights into these dynamics, with depth-averaged approaches like BingClaw employing Herschel-Bulkley rheology to simulate cohesive flows and validate against events such as the 1929 Grand Banks landslide, capturing run-out distances and velocity profiles.[16] Frictional-collisional models, such as the Voellmy type, incorporate water resistance for granular flows, while three-dimensional smoothed particle hydrodynamics (SPH) methods address complex evolutions from block sliding to debris avalanches.[16] These simulations highlight how initial slip surfaces as small as 100 m² can rapidly expand to over 100 km² through unstable growth, underscoring the role of dynamic weakening in propagation.[23]
Deposition and Resulting Features
During the propagation phase, submarine landslides transition from high-mobility flows to depositional regimes as velocity decreases and frictional resistance increases, leading to the settling of sediment-laden material on the seafloor. This deposition often occurs in phases, beginning with the emplacement of a cohesive headwall or blocky debris followed by finer-grained distal lobes, influenced by the initial failure volume, slope gradient, and water depth. In many cases, the flow incorporates ambient seafloor sediments, effectively doubling the deposit volume through entrainment and remolding.[25]Resulting features include prominent morphological elements such as slide scarps, evacuation channels, and lobate debris aprons at the flow terminus, which can extend tens of kilometers downslope. These deposits typically exhibit blocky, hummocky surfaces with irregular blocks up to hundreds of meters in diameter, interspersed with smoother, finer-grained turbidite sheets derived from post-failure suspension settling. Sediment characteristics vary from matrix-supported conglomerates with high water content in proximal zones to laminated siltstones and megaturbidites in distal areas, often showing evidence of internal deformation like shear zones and rotated blocks.[12][26][25]In confined basins or fjords, deposition can shoal the seafloor by tens of meters, creating new accommodation space that funnels subsequent sediment flows and alters local bathymetry for millennia. For instance, the 2015 Taan Fiord landslide in Alaska deposited approximately 65 million cubic meters of blocky material up to 70 meters thick over 6 kilometers, overlain by two megaturbidite layers totaling 76 million cubic meters, which advanced the adjacent glacier by 300 meters.[25] Similarly, the Pleistocene San Nicolas slide off Southern California formed a 232 square kilometer debris apron up to 300 meters thick with a 22-kilometer runout, draping Pliocene sediments and influencing regional slope stability.[12]Large-scale events can induce long-term regime shifts in sedimentation patterns, such as the ca. 21 Ma Mafia mega-slide in the Indian Ocean, which covered 11,600 square kilometers and created escarpments and topographic lows that promoted coarse-grained turbidite and debrite infilling for over 15 million years.[27] These features not only record the event's scale but also precondition adjacent slopes for future failures by loading unconsolidated sediments.
Hazards and Impacts
Tsunami Generation and Coastal Effects
Submarine landslides generate tsunamis by rapidly displacing large volumes of overlying seawater through the motion of the sliding mass, creating an initial sea surface depression followed by a compensatory elevation that forms propagating waves.[28] The efficiency of this energy transfer is low, typically ranging from 0.1% to 15% of the landslide's potential energy, depending on factors such as the density ratio between the slide material and water (ideally <1.2) and the submergence depth ratio (>0.4).[28] Unlike seismic tsunamis, which result from broad crustal deformation and produce long-wavelength waves with minimal dispersion, landslide-induced waves have shorter wavelengths, exhibit significant frequency dispersion, and are highly directional, leading to more localized impacts.[29]The initial wave formation occurs in three phases: an impulsive stage where the accelerating slide pushes water outward, a relaxation phase as the slide decelerates, and a final adjustment as the water surface stabilizes.[28] For sub-critical slides (Froude number <<1, common in deep water), the maximum wave elevation scales with the product of slide volume and acceleration divided by the square of the wave speed.[28] As these waves propagate toward the coast, they undergo shoaling in shallow water, where critical or super-critical slide motion (Froude number ≥1) can amplify amplitudes due to reduced wave speeds and decreased radial damping.[28] Propagation models, such as Boussinesq-type equations, account for dispersion, showing that energy decays more rapidly (proportional to r^{7/6} for submerged slides) compared to the slower decay of seismic waves.[29]Upon reaching coastal areas, landslide tsunamis often produce high near-field run-up heights and severe inundation, particularly in confined geometries like bays or narrow shelves, with waves arriving within minutes to tens of minutes and little warning time.[30] For instance, the 1929 Grand Banks submarine landslide off Newfoundland, triggered by a magnitude 7.3 earthquake, generated waves up to 13 meters that inundated coastal communities, causing 28 deaths and approximately $14 million in damage (adjusted to 2017 values).[30] Similarly, the 1998 Papua New Guinea event, involving a submarine slump, produced run-up heights of about 15 meters, contributing to over 2,200 fatalities through rapid coastal flooding.[28] These effects highlight the hazard's potential for catastrophic local impacts, including erosion of beaches, destruction of harbors, and disruption of marine ecosystems, though far-field propagation is limited due to quick energy dissipation.[29]
Infrastructure and Environmental Consequences
Submarine landslides pose significant risks to offshore infrastructure, including oil and gas platforms, submarine pipelines, telecommunications cables, and renewable energy installations such as offshore wind farms. These events can cause direct physical damage through sediment displacement and high-velocity debris flows, leading to structural failures or burial of equipment. For instance, in southern California, nearly 1,500 identified submarine landslides, including large-scale features like the 232 km² San Nicolas slide, threaten seabed cables and moorings essential for global communication and navigation.[31] Similarly, historical incidents have demonstrated pipeline disruptions due to landslide-induced vibrations and fatigue, as seen in cases affecting subsea oil and gas transport.[32] In the Gulf of Mexico, underwater mudslides have displaced pipelines by hundreds to thousands of feet, highlighting the economic and operational vulnerabilities in high-risk areas.[33]Beyond immediate structural threats, submarine landslides can indirectly impact coastal and nearshore infrastructure through seabed instability, complicating anchoring for floating platforms or port facilities. Mitigation strategies often involve site-specific geohazard assessments to avoid landslide-prone zones during design, though post-event recovery remains challenging due to the inaccessibility of deep-sea environments.[32] These hazards underscore the need for integrated engineering approaches that account for dynamic marine geology, as failures can result in substantial economic losses from repair and downtime.[34]Environmentally, submarine landslides disrupt benthic habitats by burying organisms and altering sediment composition, leading to reduced biodiversity and shifts in community structure. In the Laxmi Basin of the Arabian Sea, a massive landslide event resulted in mass transport deposits that showed decreased diversity and abundance of benthic foraminifera compared to surrounding undisturbed sediments, with taxa like Cibicidoides and Epistominella particularly affected due to changes in organic matter availability and dysoxic bottom waters.[35] Similarly, following the 2016 Kaikōura earthquake, which triggered turbidity flows and landslides in New Zealand's Kaikōura Canyon, deep-sea macrofauna abundances plummeted, with nematodes and polychaetes nearly absent initially; recovery involved pioneer species like spionid polychaetes and bivalves, achieving partial community similarity after 4–6 years but with persistent differences in infaunal structure.[36]Additionally, these events can destabilize gas hydrate deposits, potentially releasing methane—a potent greenhouse gas—into the ocean and atmosphere, exacerbating climate change. Offshore Norway, a documented slope failure induced rapid gas hydrate decomposition through pressure drops, facilitating fluid escape and methane venting that could influence regional carbon cycles.[37] Such releases, while not always massive, contribute to long-term ecological stressors, including acidification and oxygen depletion in affected marine ecosystems, with recovery times spanning years to decades depending on disturbance scale and local conditions.[38]
Notable Examples
Prehistoric Mega-Landslides
Prehistoric mega-landslides represent some of the most voluminous submarine mass-wasting events in Earth's history, often involving sediment displacements exceeding 1,000 km³ and extending hundreds of kilometers across continental margins or oceanic basins. These events, occurring prior to recorded human history, were typically triggered by a combination of tectonic activity, rapid sediment loading, oversteepening of slopes, and glacial or volcanic influences, leading to retrogressive failures that reshaped seafloors and generated basin-wide turbidite deposits. Their study relies on seismic profiling, bathymetric mapping, and sediment core analysis, revealing headwall scars, slide blocks, and debris flows that persisted as geohazards for millions of years.[6][39]One of the most extensively studied examples is the Storegga Slide complex in the Norwegian Sea, which includes the main Storegga event approximately 8,150 years before present (BP) and an earlier Nyegga Slide around 20,000 BP. The Storegga Slide mobilized an estimated 1,300–2,300 km³ of glacial and marine sediments across a ~300 km-wide evacuation scar, propagating ~800 km into the Norway Basin at speeds potentially exceeding 100 km/h. Triggered by seismic activity and fluid overpressuring during post-glacial sea-level rise, it produced a paleotsunami with run-up heights exceeding 20 m along coasts from Scotland to Norway, inundating low-lying areas and contributing to the submersion of Doggerland in the North Sea. The Nyegga precursor, with ~900–1,100 km³, likely stemmed from ice-sheet retreat destabilizing sediment lobes, though its tsunami impact remains less documented due to contemporaneous lower sea levels.[6]In the Pacific, the Nu'uanu Landslide off Oahu, Hawaii, stands as a prime illustration of volcanic island flank collapse, dated to between 2.1 and 1.78 million years ago (Ma) based on magnetostratigraphy of debris deposits. Originating from the northeastern flank of the Koolau shield volcano, it involved the failure of ~230 km of submarine slope, forming a debris field spanning the Hawaiian Deep to the Hawaiian Arch and incorporating fragmented volcanic rocks transported via debris avalanche. Paired with the adjacent Wailau Landslide from Molokai (~1.4 Ma), the combined complex highlights recurrent instability in hotspot-driven island chains, with paleomagnetic data indicating coherent block rotations during failure. These events underscore the role of intrusive volcanism and flank oversteepening in generating mega-scale slides capable of trans-oceanic sediment redistribution.[40]Further exemplifying tectonic influences, the Halibut Slide in the North Sea Basin occurred between 64 and 62 Ma during the Paleocene, triggered by mantle plume uplift reactivating intra-plate faults beneath Cretaceouschalk platforms. This epicontinental event displaced a decompacted volume of 1,450 km³ over a basal surface of ~7,000 km², with a runout length of 290 km from an intra-shelf high east of Scotland. Its morphology includes intact slide blocks and chaotic distal flows, influencing regional sediment routing for up to 10 million years and demonstrating how far-field stresses can destabilize low-gradient shelves.[39]The Agulhas Slump off southeastern Africa represents an extreme in scale, a post-Pliocene composite failure spanning 750 km in length and 106 km in width, with a volume surpassing 20,000 km³—the largest documented submarine slump. Situated on a sheared continental margin, it features a prominent glide plane scar separating proximal allochthonous masses from distal spreads into the Transkei Basin, constrained by basement ridges. Formed through multiple retrogressive phases on overpressured sediments, the slump's structure reflects interactions between sheartectonics and depositional loading, with its toe region impounded against oceanic crust. Such mega-events highlight the long-term hazards of passive margins, where ancient slides can precondition modern slopes for renewed instability.[41]
Modern and Recent Events
One of the earliest well-documented modern submarine landslides occurred on November 18, 1929, off the Grand Banks of Newfoundland, triggered by a magnitude 7.2 earthquake.[42] The event involved a massive slump displacing approximately 100 km³ of sediment along the St. Pierre Slope, evolving into a turbidity current that severed 12 transatlantic telegraph cables over 600 km away.[43] This landslide generated a tsunami with waves up to 13 meters high along the Burin Peninsula, resulting in 28 fatalities and significant coastal flooding.[44]In 1957, a large submarine landslide was recently identified near the epicenter of the magnitude 8.6 Andreanof Islands earthquake in the Aleutian Islands, Alaska, occurring on March 9.[45] The structure spans over 10 miles across the southern slope of the Aleutian Shelf, approximately 10 miles northwest of the quake's origin, and is believed to have been triggered by the seismic shaking.[45] Preliminary analysis suggests it may have contributed as a secondary source to the earthquake's trans-Pacific tsunami, though further dating is required to confirm its exact role and timing.[45]The 1998 Papua New Guinea event, following a magnitude 7.0 earthquake on July 17 near Aitape, exemplifies a submarine slump generating a deadly local tsunami.[46] The landslide, involving sediment failure at depths of around 1-2 km, produced waves up to 15 meters high that inundated a 20 km coastal stretch, killing over 2,200 people and displacing thousands.[47] Detailed seismic and bathymetric surveys later confirmed the slump's volume at about 4 km³, highlighting the hazard of earthquake-induced mass wasting in tectonically active margins.[48]A more recent example is the December 22, 2018, flank collapse of Anak Krakatau volcano in the Sunda Strait, Indonesia, which included a significant submarine component.[49] Triggered by prolonged eruptive activity that added over 54 million tons of material to the unstable southwestern flank, the failure displaced 0.116 km³ of submarine debris at depths of 100-120 meters, forming blocky megablocks extending 1.5 km into the adjacent caldera.[49] This event generated a tsunami with runup heights reaching 13 meters on Sumatra and Java coasts, causing 437 deaths, over 14,000 injuries, and widespread displacement.[49] The incident, captured by modern instrumentation, provided unprecedented data on volcanic submarine mass movements and their tsunami potential.[49]
Research and Mitigation
Monitoring and Detection Methods
Monitoring and detection of submarine landslides rely on a combination of remote geophysical surveys, passive acoustic and seismic techniques, and in-situ instrumentation to identify precursors, ongoing deformation, and post-event features. These methods are essential for assessing slope stability in submarine environments, where direct observation is challenging due to water depth and remoteness. High-resolution bathymetric surveys using multibeam echosounders have become a cornerstone for mapping landslide morphology and monitoring changes over time, enabling the detection of scarps, deposits, and evacuation zones with centimeter-scale precision.[24]Seismic and acoustic monitoring provides critical insights into landslide dynamics by capturing signals from initiation to propagation. Passive geophysical networks, including broadband seismometers and hydrophones deployed on the seafloor or onshore, detect low-frequency acoustic emissions and ground motions associated with mass movements, allowing for real-time event location and timing through triangulation. For instance, hydrophone arrays have recorded landslide velocities of 10–25 m/s over distances up to 7,000 km, as observed in events along the Hawaiian Ridge.[50] In the Gulf of Mexico, onshore seismic data from 2008–2015 identified 85 landslide events triggered by storms or salt tectonics, highlighting the role of such systems in preconditioning analysis.[51] Side-scan sonar and sub-bottom profilers complement these by imaging seafloor disruptions and sediment layers, revealing buried slide planes at depths of 8–12.5 m in tidal channels.[52]In-situ geotechnical monitoring employs instruments like inclinometers, pressure gauges, and piezometers to measure slope deformation, pore pressure, and sediment strength directly. These are often integrated into long-term observatories for multi-parameter data collection, such as vertical and lateral displacements tracked via array displacement meters with MEMS accelerometers. A 75-day observation in China's Zhoushanarchipelago demonstrated wave-induced accelerations during storms, with significant wave heights of 0.65 m correlating to heightened landslide activity.[52] Acoustic Doppler current profilers (ADCPs) further aid detection by monitoring turbidity currents generated by slides, estimating flow velocities and sediment concentrations in water columns.[24]As of 2022, advances incorporate distributed acoustic sensing along fiber-optic cables and machine learning for data fusion from multi-source platforms, enhancing early warning capabilities. As of 2025, further developments include ocean-bottom seismometers for documenting landslide evolution in real time and deep learning-based semantic segmentation for automated landslide identification from bathymetric data.[24][53][54][22][55] Seminal reviews emphasize the evolution from qualitative mapping to quantitative, real-time systems, as outlined in early classifications of submarine mass movements.
Modeling and Hazard Assessment
Modeling of submarine landslides involves numerical simulations to predict slope failure initiation, material flow dynamics, and associated hazards such as tsunamis. These models integrate geophysical data like bathymetry and soil properties to assess stability and runout. Key approaches include slope stability analyses and dynamic flow simulations, often validated against historical events like the 1929 Grand Banks landslide.[16][56]Slope stability modeling employs methods such as the infinite slope analysis, where the factor of safety (FS) is calculated asFS = \frac{s_u}{\gamma z \sin \theta \cos \theta}with s_u as undrained shear strength, \gamma as total unit weight, z as depth to the failure plane, and \theta as slope angle. More advanced techniques use limit equilibrium methods, the Newmark sliding block model for seismic triggering, and finite element (FE), finite difference (FD), or material point method (MPM) simulations for complex geometries. A notable application is Hovland's 3D slope stability model for cohesive-frictional soils, which identified 23 potential landslides in Palu Bay during the 2018 Sulawesi event, with volumes up to 0.07 km³ and thicknesses of 5–20 m, aiding in tsunami run-up predictions of 3.5–13.5 m.[16][56]Dynamic propagation models simulate landslide mobility and deposition, using depth-averaged equations with rheologies like Herschel-Bulkley in the BingClaw model, which has been validated for the Storegga Slide with runout distances exceeding hundreds of kilometers. Frictional-collisional models, such as the Savage-Hutter or Voellmy formulations, account for granular flow behaviors, revealing higher mobility in submarine settings compared to subaerial landslides due to lower height-to-length (H/L) ratios. These models are crucial for estimating volumes ranging from 10⁻⁶ to 10⁵ km³ and integrating with tsunami simulations using nonlinear shallow-water (e.g., MOST, HySEA) or Boussinesq equations (e.g., Neowave).[16][57]Statistical approaches enhance hazard understanding by analyzing global databases of landslide scars, showing volume-frequency distributions following a power-law with an exponent of ~1/2 across six orders of magnitude, though lognormal fits better for siliciclastic margins. Probabilistic hazard assessments employ event trees, Bayesian networks, or magnitude-frequency curves to derive empirical hazard curves, linking landslide size to earthquake magnitudes above M4.5 and sedimentation rates. For instance, inverse power-law assumptions are tested against lognormal models to predict sizes decreasing non-linearly with earthquake frequency.[58][16][59]Hazard assessment follows structured workflows: site characterization via multibeam echo-sounding and seismic data, slope stability screening, dynamic modeling, and impact evaluation including tsunami generation. Recommended steps include constructing databases from bathymetric surveys, source-specific simulations, and probabilistic tsunami hazard assessments (PTHA) using empirical wave parameter methods. Examples include semi-probabilistic analyses at Bjørnafjorden, calibrated with undrained shear strength of 8 kPa and Savage number ~6, and Bayesian Monte Carlo simulations for the Stones oil field, emphasizing credible worst-case scenarios and sensitivity analyses for offshore infrastructure. Multidisciplinary efforts, such as susceptibility mapping in the South Sandwich Islands via GIS, further identify high-risk zones for tsunamis.[16][57][60]