Carl B. Harbitz

Submarine landslides: processes, triggers and hazard prediction

Huge landslides, mobilizing hundreds to thousands of km3 of sediment and rock are ubiquitous in submarine settings ranging from the steepest volcanic island slopes to the gentlest muddy slopes of submarine deltas. Here, we summarize current knowledge of such landslides and the problems of assessing their hazard potential. The major hazards related to submarine landslides include destruction of seabed infrastructure, collapse of coastal areas into the sea and landslide-generated tsunamis. Most submarine slopes are inherently stable. Elevated pore pressures (leading to decreased frictional resistance to sliding) and specific weak layers within stratified sequences appear to be the key factors influencing landslide occurrence. Elevated pore pressures can result from normal depositional processes or from transient processes such as earthquake shaking; historical evidence suggests that the majority of large submarine landslides are triggered by earthquakes. Because of their tsunamigenic potential, ocean-island flank collapses and rockslides in fjords have been identified as the most dangerous of all landslide related hazards. Published models of ocean-island landslides mainly examine ‘worst-case scenarios’ that have a low probability of occurrence. Areas prone to submarine landsliding are relatively easy to identify, but we are still some way from being able to forecast individual events with precision. Monitoring of critical areas where landslides might be imminent and modelling landslide consequences so that appropriate mitigation strategies can be developed would appear to be areas where advances on current practice are possible.

Model simulations of tsunamis generated by the Storegga Slides

Abstract A mathematical model based on the hydrodynamic shallow water equations is developed for numerical simulation of water waves generated by the submarine Storegga Slides on the Norwegian continental slope. The equations are solved numerically by a finite difference technique. Computations of wave amplification effects reveal run-up heights for the Second Storegga Slide between 3 and 5 m in exposed areas along the eastern coast of Greenland, Iceland and Scotland and the western coast of Norway. The calculated run-up heights agree remarkably well with possible tsunami wave heights deduced from geological evidences along the eastern coast of Scotland. The generated wave heights are strongly dependent on the acceleration of the slide. The effects of shear stress at the interface between the water and the slide body, has turned out to be important.

Submarine slope stability on high-latitude glaciated Svalbard–Barents Sea margin

Abstract Slope stability is evaluated at two locations on high latitude, deep sea fans along the Svalbard–Barents Sea margin, based on available samples and using an “infinite slope” analysis. The stability evaluation uses the Mohr–Coulomb failure criterion, and a semi-analytical approach based on Gibsons formulation for determining the excess pore pressure build-up due to sedimentation. The main results are presented in the form of contour plots of slope safety factors in a diagram with axes of time and thickness of deposit. The results show that during rapid sedimentation, which mostly takes place during periods of maximum glaciation with the ice front located along the shelf edge, slope failure will occur with a frequency varying between 95 and 170 years. Only part of the upper sedimented layer will be mobilised (10–30 m), while the remaining thickness (40–70 m) will remain at the initial sedimentation site. These results may explain why the continental slope is characterised by relatively uniform sediment thickness from upper to lower slope.

Submarine landslide tsunamis: how extreme and how likely?

A number of examples are presented to substantiate that submarine landslides have occurred along most continental margins and along several volcano flanks. Their properties of importance for tsunami generation (i.e. physical dimensions, acceleration, maximum velocity, mass discharge, and travel distance) can all gain extreme values compared to their subaerial counterparts. Hence, landslide tsunamis may also be extreme and have regional impact. Landslide tsunami characteristics are discussed explaining how they may exceed tsunamis induced by megathrust earthquakes, hence representing a significant risk even though they occur more infrequently. In fact, submarine landslides may cause potentially extreme tsunami run-up heights, which may have consequences for the design of critical infrastructure often based on unjustifiably long return periods. Giant submarine landslides are rare and related to climate changes or glacial cycles, indicating that giant submarine landslide tsunami hazard is in most regions negligible compared to earthquake tsunami hazard. Large-scale debris flows surrounding active volcanoes or submarine landslides in river deltas may be more frequent. Giant volcano flank collapses at the Canary and Hawaii Islands developed in the early stages of the history of the volcanoes, and the tsunamigenic potential of these collapses is disputed. Estimations of recurrence intervals, hazard, and uncertainties with today’s methods are discussed. It is concluded that insufficient sampling and changing conditions for landslide release are major obstacles in transporting a Probabilistic Tsunami Hazard Assessment (PTHA) approach from earthquake to landslide tsunamis and that the more robust Scenario-Based Tsunami Hazard Assessment (SBTHA) approach will still be most efficient to use. Finally, the needs for data acquisition and analyses, laboratory experiments, and more sophisticated numerical modelling for improved understanding and hazard assessment of landslide tsunamis are elaborated.

Submarine mass-wasting on glacially-influenced continental slopes: processes and dynamics

Abstract Submarine slides and debris flows are common and effective mechanisms of sediment transfer from continental shelves to deeper parts of ocean basins. They are particularly common along glaciated margins that have experienced high sediment flux to the shelf break during and after glacial maxima. During one single event, typically lasting for a few hours or less, enormous sediment volumes can be transported over distances of hundreds of kilometres, even on very gentle slopes. In order to understand the physics of these mass flows, the process is divided into a release phase, followed by break-up, flow and final deposition. Little is presently known regarding release and break-up, although some plausible explanations can be inferred from basic mechanics of granular materials. Once initiated, the flow of clay-rich or muddy sediments may be assumed to behave as a (non-Newtonian) Herschel-Bulkley fluid. Fluid dynamic concepts can then be applied to describe the flow provided the rheological properties of the material are known. Numerical modelling supports our assertion that the long runout distances observed for large volumes of sediments moving down gentle slopes can be explained by partial hydroplaning of the flowing mass. Hydroplaning might also explain the sharp decrease of the friction coefficient for submarine mass flows as a function of the released volume. The paper emphasizes the need for a better understanding of the physics of mass wasting in the submarine environment.

Modeling tsunamis from earthquake sources near Gorringe Bank southwest of Portugal

The Azores-Gibraltar fracture zone with the huge bathymetric reliefs in the area southwest of Portugal is believed to have been the source of large historic tsunami events. This report describes simulations of tsunami generation and propagation from sources near the Gorringe Bank. The well-documented 1969 tsunami event is examined both with a ray-tracing technique and with finite difference models based on various shallow water equations. Both methods show that the most likely source location is southeast of the Gorringe Bank near the epicenter location determined from seismic data. The tsunami source is calculated by formulas given by Okada [1985] for surface deformation of an elastic half-space caused by faulting. Observed wave amplitude and travel time and values computed from an initial wave field according to Okada [1985] formulas show acceptable agreement for most stations along the coast of Portugal and Spain. However, in order to explain a large primary wave with downward displacement observed on the coast of Morocco, an alternative source model with a larger area of downward displacement has been introduced. This also leads to a better overall fit with observed travel time. Implications for disastrous events, as the one in 1755, are also discussed. Linear hydrostatic shallow water models are used for most of the simulations, but the importance of nonlinearity and dispersion is examined with the Boussinesq equations. The sensitivity of the solution to changes in the location and the strength of the source is discussed, and a series of grid refinement studies are performed in order to assess the accuracy of the simulations.

On the characteristics of landslide tsunamis

This review presents modelling techniques and processes that govern landslide tsunami generation, with emphasis on tsunamis induced by fully submerged landslides. The analysis focuses on a set of representative examples in simplified geometries demonstrating the main kinematic landslide parameters influencing initial tsunami amplitudes and wavelengths. Scaling relations from laboratory experiments for subaerial landslide tsunamis are also briefly reviewed. It is found that the landslide acceleration determines the initial tsunami elevation for translational landslides, while the landslide velocity is more important for impulsive events such as rapid slumps and subaerial landslides. Retrogressive effects stretch the tsunami, and in certain cases produce enlarged amplitudes due to positive interference. In an example involving a deformable landslide, it is found that the landslide deformation has only a weak influence on tsunamigenesis. However, more research is needed to determine how landslide flow processes that involve strong deformation and long run-out determine tsunami generation.

Numerical simulation of mud-rich subaqueous debris flows on the glacially active margins of the Svalbard-Barents Sea

Abstract Seismic images and sediment core data from the Bear Island and Isfjorden fans localized along the Svalbard–Barents Sea continental margin, reveal an interesting depositional system consisting of stacked debris flow lobes. The frequent release of debris flows was associated with large volumes of sediment rapidly delivered to the shelf break during periods of maximum glaciation. The compositions of the lobes for both fans are similar, consisting of mainly clay and silt. The data show, however, a dramatic difference in runout distances for the two areas. Isfjorden debris lobes are 10–30 km in length whereas Bear Island lobes are 100–200 km in length. Even more intriguing is the fact that the large runout distances on the Bear Island fan occurred on slopes less than 1° whereas the Isfjorden fan flows occurred on slopes greater than 4°. Depth-averaged non-linear one-dimensional equations for balance of mass and linear momentum are applied to simulate the subaqueous debris flow. The equations are solved by the numerical model BING, describing the flow as a visco-plastic Bingham fluid. The model is employed to study the effect yield strength, viscosity and bathymetry have on debris flow runout. The study shows that the large runout distances can be achieved on the Bear Island fan by visco-plastic flows with sufficiently low yield strength. High yield strength sediments require an additional mechanism, such as hydroplaning, to reach measured runout distances. Most importantly, this study shows the necessity of good rheological measurements for accurate numerical modeling of subaqueous debris flows.

Submarine mass movements and their consequences : 6th International Symposium

We present a short review of the state-of-the-art numerical tools that have been used for modeling landslide-generated waves. A comparative study is conducted on the physical properties of earthquake- and landslide-generated waves suggesting that both dispersion and nonlinearity effects may be neglected for the former waves whereas they may be considered for the latter ones. We introduce landslide tsunami models and group them into three classes: (1) models treating the moving mass as a fluid, (2) models estimating the initial water surface, and (3) models fed by the transient seafloor deformation. Selection of a particular model from the list of models introduced here depends on: (1) the dimensions of the source, (2) the available computing capacities, (3) availability of fine bathymetric grid, and (4) the purposes of the modeling.

On probability analysis in snow avalanche hazard zoning

Abstract The reduced societal acceptance of living in regions exposed to snow avalanches, and the increased economic consequences when houses are located within a hazard zone, highlight the uncertainty concerning avalanche run-out prediction. The limitations of today’s zoning procedures are especially pronounced in potential avalanche terrain where there are few observations of snow avalanches, where old buildings are present in the potential run-out zone, and where the local climate does not favour severe snow accumulation. This paper combines a mechanical probabilistic model for avalanche release with a statistical/topographical model for avalanche run-out distance to obtain the unconditional probability of extreme run-out distance. For the mechanical model, a first-order reliability method (FORM) and Monte Carlo simulations are compared. The interpretation of the statistical/topographical model either as an extreme value model or as a single value model is discussed. Furthermore, both a classical approach where the probability of an avalanche occurring is a constant, and a Bayesian approach with stochastic probability, are compared. Finally, example applications in hazard zoning are presented, with emphasis on how the influence of historical observations, local climate, etc., on run-out distance can be quantified in statistical terms and how a specified certainty level can be found from constructing confidence intervals for, for example, the most likely largest run-out distance during various time intervals.