Homa J. Lee

Pleistocene slope instability of gas hydrate‐laden sediment on the Beaufort sea margin

Abstract The Beaufort Sea continental slope is disrupted by a belt of massive bedding‐plane slides and rotational slumps. This zone coincides with a region of sediment containing gas hydrate, an inclusion compound of gas and water. Quantitative studies suggest that elevated pore‐fluid pressures generated as a by‐product of gas hydrate disassociation during repeated episodes of eustatic sea level lowering during Pleistocene time were a major cause of these slides. Eustatic sea level fall causes reduced pressures acting on seafloor sediment. In oceanic areas underlain by sediment with gas hydrate, the reduction of sea level initiates disassociation along the base of the gas hydrate, which, in turn, causes the release of large volumes of gas into the sediment and creates excess pore‐fluid pressures and reduced slope stability. A quantitative approach was taken to predict excess pressures at the base of the gas hydrate zone, and the stability of the overlying sediment for a variety of sediment types. Our stud...

Distinguishing sediment waves from slope failure deposits: Field examples, including the 'humboldt slide', and modelling results

Abstract Migrating sediment waves have been reported in a variety of marine settings, including submarine levee–fan systems, floors of fjords, and other basin or continental slope environments. Examination of such wave fields reveals nine diagnostic characteristics. When these characteristics are applied to several features previously attributed to submarine landslide deformation, they suggest that the features should most likely be reinterpreted as migrating sediment-wave fields. Sites that have been reinterpreted include the ‘Humboldt slide’ on the Eel River margin in northern California, the continental slope in the Gulf of Cadiz, the continental shelf off the Malaspina Glacier in the Gulf of Alaska, and the Adriatic shelf. A reassessment of all four features strongly suggests that numerous turbidity currents, separated by intervals of ambient hemipelagic sedimentation, deposited the wave fields over thousands of years. A numerical model of hyperpycnal discharge from the Eel River, for example, shows that under certain alongshore-current conditions, such events can produce turbidity currents that flow across the ‘Humboldt slide’, serving as the mechanism for the development of migrating sediment waves. Numerical experiments also demonstrate that where a series of turbidity currents flows across a rough seafloor (i.e. numerical steps), sediment waves can form and migrate upslope. Hemipelagic sedimentation between turbidity current events further facilitates the upslope migration of the sediment waves. Physical modelling of turbidity currents also confirms the formation and migration of seafloor bedforms. The morphologies of sediment waves generated both numerically and physically in the laboratory bear a strong resemblance to those observed in the field, including those that were previously described as submarine landslides.

Numerical analysis of the mobility of the Palos Verdes debris avalanche, California, and its implication for the generation of tsunamis

Abstract Analysis of morphology, failure and post-failure stages of the Palos Verdes debris avalanche reveals that it may have triggered a significant tsunami wave. Our analysis of the failure itself indicates that the slope is stable under aseismic conditions but that a major earthquake (with a magnitude around 7) could have triggered the slide. A post-failure analysis, considering the debris avalanche as a bi-linear flow, shows that peak velocities of up to 45 m/s could have been reached and that the initial movement involved a mass of rock less than 10 km wide, 1 km long and about 50–80 m thick. Initial wave height estimates vary from 10 to 50 m. Tsunami waves propagating to the local shoreline would be significantly smaller. Such a range demonstrates our lack of proper knowledge of the transition from failure to post-failure behavior related to mass movements. Further investigations and analyses of terrestrial and submarine evidence are required for a proper hazard assessment related to tsunami generation in the Los Angeles area.

Sediment Mass-Flow Processes on a Depositional Lobe, Outer Mississippi Fan

ABSTRACT SeaMARC 1A sidescan-sonar imagery and cores from the distal reaches of a depositional lobe on the Mississippi Fan show that channelized mass flow was the dominant mechanism for transport of silt and sand during the formation of this part of the fan. Sediments in these flows were rapidly deposited once outside of their confining channels. The depositional lobe is formed of a series of long, narrow sublobes composed of thin-bedded turbidites (normally graded siliciclastic sand and silt, 20 cm thick on average), debris-flow deposits (soft clay clasts up to 5 cm in diameter in a siliciclastic silt matrix, 48 cm thick on average), and background-sedimentation hemipelagic muds. The mass flows most likely originated from slope failure at the head of the Mississippi Canyon or on the outer ontinental shelf and flowed approximately 500 km to the distal reaches of the fan, with debris flow being the dominant flow type. An analysis that uses the geometry of the confining channels and strength properties of the debris-flow material shows that these thin debris flows could have traveled hundreds of kilometers on extremely small sea-floor slopes at low velocities if the flowing medium behaved as Bingham fluids and were steady-state phenomena.

Characteristics of a sandy depositional lobe on the outer Mississippi fan from SeaMARC IA sidescan sonar images

SeaMARC IA sidescan sonar images of the distal reaches of a depositional lobe on the Mississippi Fan show that channelized rather than unconfined transport was the dominant transport mechanism for coarse-grained sediment during the formation of this part of the deep-sea fan. Overbank sheet flow of sands was not an important process in the transport and deposition of the sandy and silty sediment found on this fan. The dendritic distributary pattern and the high order of splaying of the channels, only one of which appears to have been active at a time, suggest that coarse-grained deposits on this fan are laterally discontinuous.

Upper Pleistocene turbidite sand beds and chaotic silt beds in the channelized, distal, outer-fan lobes of the Mississippi fan

Cores from a Mississippi outer-fan depositional lobe demonstrate that sublobes at the distal edge contain a complex local network of channelized-turbidite beds of graded sand and debris-flow beds of chaotic silt. Off-lobe basin plains lack siliciclastic coarse-grained beds. The basin-plain mud facies exhibit low acoustic backscatter on SeaMARC IA sidescan sonar images, whereas high acoustic backscatter characteristic of the lobe sand and silt facies. The depth of the first sand-silt layer correlates with relative backscatter intensity and stratigraphic age of the distal sublobes (i.e., shallowest sand = highest backscatter and youngest sublobe). The high proportion (>50%) of chaotic silt compared to graded sand in the distal, outer-fan sublobes may be related to the unstable, muddy, canyon-wall source areas of the extensive Mississippi delta-fed basin slope. A predominace of chaotic silt in cores or outcrops from outer-fan lobes thus may predict similar settings for ancient fans.

Geotechnical characteristics and slope stability on the Ebro margin, western Mediterranean

Abstract Sedimentological and geotechnical analyses of core samples from the Ebro continental slope define two distinct areas on the basis of sediment type, physical properties and geotechnical behavior. The first area is the upper slope area (water depths of 200–500 m), which consists of upper Pleistocene prodeltaic silty clay with a low water content (34% dry weight average), low plasticity, and high overconsolidation near the seafloor. The second area, the middle and lower slope (water depths greater than 500 m), contains clay- and silt-size hemipelagic deposits with a high water content (90% average), high plasticity, and a low to moderate degree of overconsolidation near the sediment surface. Results from geotechnical tests show that the upper slope has a relatively high degree of stability under relatively rapid (undrained) static loading conditions, compared with the middle and lower slopes, which have a higher degree of stability under long-term (drained) static loading conditions. Under cyclic loading, which occurs during earthquakes, the upper slope has a higher degree of stability than the middle and lower slopes. For the surface of the seafloor, calculated critical earthquake accelerations that can trigger slope failures range from 0.73 g on the upper slope to 0.23 g on the lower slope. Sediment buried well below the seafloor may have a critical acceleration as low as 0.09 g on the upper slope and 0.17 g on the lower slope. Seismically induced instability of most of the Ebro slope seems unlikely given that an earthquake shaking of at least intensity VI would be needed, and such strong intensities have never been recorded in the last 70 years. Other cyclic loading events, such as storms or internal waves, do not appear to be direct causes of instability at present. Infrequent, particularly strong earthquakes could cause landslides on the Ebro margin slope. The Columbretes slide on the southwestern Ebro margin may have been caused by intense earthquake shaking associated with volcanic emplacement of the Columbretes Islands. Localized sediment slides on steep canyon and levee slopes could have been caused by less intense shaking. In general, the slope is stable under present environmental loading conditions and is fundamentally constructional. Nevertheless, rapid progradation caused by high sedimentation rates and other processes acting during low sea-level periods, such as more intense wave loading near the shelfbreak, may have caused major instability in the past.

Quantitative controls on submarine slope failure morphology

Abstract The concept of the steady‐state of deformation can be applied to predicting the ultimate form a landslide will take. The steady‐state condition, defined by a line in void ratio‐effective stress space, exists at large levels of strain and remolding. Conceptually, if sediment initially exists with void ratio‐effective stress conditions above the steady‐state line, the sediment shear strength will decrease during a transient loading event, such as an earthquake or storm. If the reduced shear strength existing at the steady state is less than the downslope shear stress induced by gravity, then large‐scale internal deformation, disintegration, and flow will occur. If sediment exists at a state that is on or below the steady‐state line, disintegration and flow will typically not occur. Confirming these concepts, studies of subaerial landslides show an association between disintegrative flows and void ratio‐effective stress states above the steady‐state line. Nondisintegrative landslides are associated ...

The Saguenay Fjord, Quebec, Canada: integrating marine geotechnical and geophysical data for spatial seismic slope stability and hazard assessment

In 1996 a major flood occurred in the Saguenay region, Quebec, Canada, delivering several km 3 of sediment to the Saguenay Fjord. Such sediments covered large areas of the, until then, largely contaminated fjord bottom, thus providing a natural capping layer. Recent swath bathymetry data have also shown that sediment landslides are widely present in the upper section of the Saguenay Fjord, and therefore, should a new event occur, it would probably expose the old contaminated sediments. Landslides in the Upper Saguenay Fjord are most probably due to earthquakes given its proximity to the Charlevoix seismic region and to that of the 1988 Saguenay earthquake. In consequence, this study tries to characterize the permanent ground deformations induced by different earthquake scenarios from which shallow sediment landslides could be triggered. The study follows a Newmark analysis in which, firstly, the seismic slope performance is assessed, secondly, the seismic hazard analyzed, and finally an evaluation of the seismic landslide hazard is made. The study is based on slope gradients obtained from EM1000 multibeam bathymetry data as well as water content and undrained shear strength measurements made in box and gravity cores. Ground motions integrating local site conditions were simulated using synthetic time histories. The study assumes the region of the 1988 Saguenay earthquake as the most likely source area for earthquakes capable of inducing large ground motions in the Upper Saguenay region. Accordingly, we have analyzed several shaking intensities to deduce that generalized sediment displacements will begin to occur when moment magnitudes exceed 6. Major displacements, failure, and subsequent landslides could occur only from earthquake moment magnitudes exceeding 6.75. 8 2002 Elsevier Science B.V. All rights reserved.

Geotechnical Analyses of Submarine Landslides in Glacial Marine Sediment, Northeast Gulf of Alaska

Glaciation is the most important process contributing sediment to the northeast Gulf of Alaska. Large sediment failures within the Holocene glacial-marine sediment of the continental shelf have been identified on slopes as gentle as 0.5°. The major offshore processes responsible for sediment failure in the Gulf of Alaska are earthquake and storm wave loading coupled with cyclic shear strength degradation. A normalized soil parameter (NSP) approach can yield shear strength parameters that are somewhat independent of coring disturbance by normalizing these parameters by appropriate consolidation stresses. The NSP approach also appears capable of aiding in the extrapolation of surficial sediment properties to the subsurface. Laboratory tests using the NSP approach, supplemented with in-place vane shear data, reveal that for these glacial-marine sediments, clayey silt with a natural water content between 35% and 45% is most susceptible to cyclic loading. Cores that contain more of this susceptible clayey silt roughly correlate with locations of sediment failure features. A simplified analysis shows that in water depths shallower than 35 m, maximum storm waves would produce shearing stresses greater than stresses induced by maximum earthquakes. In deeper water, earthquakes would produce greater stresses. Differences in failure morphology are difficult to relate to advanced geotechnical parameters but likely relate to observed variations in plasticity, slope angle, water depth, or variations in consolidation state.