Submarine landslide tsunamis – generation mechanisms and hazard assessment
Presented by Dr. Carl Bonnevie HARBITZ on 25 Sep 2013 from 09:00 to 09:30
Type: Oral presentation, full-paper
Track: Oral presentations
Submarine landslides may occur along any passive or active continental margin and at different water depths. Most landslides that cause tsunamis result in more local effects than comparable earthquake-induced tsunamis, due to different source characteristics. However, enormous submarine landslides exhibiting volumes of several thousands of km3 may cause tsunamis with more widespread effects. Volcanic flank collapses may also cause tsunamis inducing distant destruction, although their tsunamigenic potentials are disputed. The landslide parameters governing the tsunami generation can all gain extreme values. However, the hazard posed by a landslide is not necessarily proportional to its size. The high mobility of submarine landslides may be explained partly by the large volumes involved and partly by the landslide/water interaction. Material properties, including clay rheology, are of great importance for the dynamics of most events. The quantification of the landslide parameters is complicated by the transformation of the landslide from a huge slab to smaller blocks, then to a highly viscous fluid and – in many cases – to a turbidity current. The stages of flow evolution are connected to different flow regimes that require different modeling approaches. Many submarine landslides develop retrogressively (i.e. they are released progressively upwards from the toe). Depending on the time lags between releases of the individual elements, this process normally reduces the tsunamigenic power. Mechanical analyses of release, disintegration, and flow mechanisms will help understanding the dynamics of past events. Laboratory experiments (and the important discussion on how they relate to the corresponding natural phenomena) are particularly important for submarine landslides that are difficult to observe at full scale. Wave frequency dispersion, which is normally more pronounced for landslide than for earthquake tsunamis owing to the relatively short length scale of landslide thickness variations relative to the water depth, makes the propagation more complex. Generally, the leading-order wave is reduced due to dispersion. On the other hand, the limited wavelengths of landslide tsunamis favor amplification due to shoaling. Moreover, the landslides are difficult to observe and monitor; hence, they are apparently unpredictable, which in turn makes the tsunami consequences more extreme. Most tsunami hazard assessments have been scenario-based and focused on earthquake tsunamis. More recently, however, a Probabilistic Tsunami Hazard Assessment (PTHA, largely inherited from Probabilistic Seismic Hazard) approach has been developed. However, insufficient sampling and changing conditions for landslide release are major obstacles in transporting a PTHA approach from earthquake to landslide tsunamis. It is further expected that the landslide tsunami risk is dominated by the large return periods, generally carrying the largest uncertainties. Hence, the more robust Scenario-Based Tsunami Hazard Assessment (SBTHA) approach will probably still be most efficient to use. Advanced numerical models to simulate submarine landslides or tsunamis are already in practical use. Combining models for the evolution of retrogressive landslides with the more sophisticated (dispersive) tsunami propagation models will be a huge step forward in the field of landslide tsunami research.