Researchers at the University of Barcelona and the Institute of Marine Sciences (CSIC) are trying to unravel the distribution of submarine landslides in the continental margins of Europe and the Mediterranean region, both in time and space, to improve our prediction capability of these natural hazards, including potential tsunami generation.

Our perception of offshore natural hazards dramatically changed on December 26, 2004 after the big Sumatran earthquake and subsequent tsunami. Many people became suddenly aware of the potential hazard that tsunamis may pose, and offshore earthquakes and tsunamis are now readily associated. However earthquakes are not the only mechanism by which tsunamis may form. On July 17, 1998 another earthquake was followed by a tsunami that affected the northern coast of Papua-New Guinea and caused about 2000 fatalities. The earthquake characteristics failed, however, to explain the tsunami run up, and scientists suggested a submarine landslide, presumably triggered by the earthquake, as being ultimately responsible for the tsunami. In many other instances, a similar relationship between earthquakes, submarine landslides, and tsunamis has been found. But in some others an earthquake was not necessary to trigger a submarine landslide that caused a tsunami.

Significant improvements in our understanding of the mechanisms of tsunami propagation and run up have been made in the last decade and are now modeled in a variety of approaches. However, and most critically, scientists are not able to predict when earthquakes or submarine landslides might occur. Yet, probabilistic approaches exist to determine seismic hazards based on extensive catalogs of earthquakes from instrumental records. Our knowledge of submarine landslide occurrence is particularly crude as we lack the instrumental records that we can produce shortly after earthquakes occur. The offshore catalogue of slope failures and their recurrence rates is largely incomplete. Consequently, our ability to estimate the hazard from submarine slope failures, including the potential creation of a tsunami, is rather low. But, submarine landslides do not represent a potential geohazard only for their tsunami creating potential. We are making increasing use of the marine environment, and numerous reports exist showing damage to coastal and offshore infrastructures, including ports, cables, and pipelines, by submarine slope failures. In addition, with oil reserves onshore and in shallow water declining, the oil industry is moving onto the continental slope, where available data indicates a higher hazard for slope failure.

Our lack of knowledge on submarine landslides is further exacerbated by the inaccessibility of the marine environment, particularly the deep environment. Scientists use large research vessels and sophisticated geophysical tools to overcome this. These tools are a key to identifying the magnitude and areal distribution of submarine slope failures. But, some age constraints are also necessary. For this, scientists need to acquire seafloor sediment samples which can then be tested using a variety of methods to determine the ages of the sediments and the dates of the submarine slope failures. Most research vessels use a variety of sediment samplers that allow a maximum penetration of about 50 m into the seafloor. In most instances this provides access to the submarine events of about the last 20,000 years. These data, however, are often not sufficient to estimate recurrence rates. This is particularly the case in areas where submarine landslides occur. These areas often have high sedimentation rates reducing the time interval that we are able to sample. The only means presently available to science for sampling sediments deep in the sedimentary column is through drilling vessels similar to the ones being used in the oil industry.

The Integrated Ocean Drilling Program is an international initiative through which scientists at the University of Barcelona and the Institute of Marine Sciences (CSIC), coordinating an international research team, expect to drill thick sedimentary sequences where submarine slope instability is a recurrent phenomenon. One of the objectives of such experiments is to try to determine how the Earth's natural climatic variations affect the stability of the seafloor. Preliminary results suggest that warming climate periods are especially prone to exceptionally large submarine slope failures. An accurate determination of this relationship will be quite useful in determining future scenarios given the changing climate resulting from human industrial activities. In addition to providing the necessary data to estimate recurrence rates, deep sediment samples may help scientists establish the physical properties of the layers on which the slope failures have occurred and where they might occur again in the future. It will also allow scientists to determine the conditions under which these layers formed and improve our knowledge of their mechanical behavior. Recent work is also trying to address the ground conditions prior to submarine slope failure, or early in the failure process, to determine the signs that might indicate imminent slope failure. In this regard, scientists are trying to establish offshore shallow and deep sub-seafloor observatories to monitor the deformations and the relationship to pore fluid flow pathways in continental margins.

The integrated study of submarine landslides and offshore geohazards requires a multidisciplinary research approach involving geophysics, geology, civil engineering, biology, and social sciences. It is a scientifically and technologically challenging field requiring large infrastructures to address the problems. The scientific results will encompass both fundamental and applied research, and will not in all instances have an immediate application. To tackle these questions, we therefore need a combined effort from our research agencies, the EU, and industry to better understand submarine landslides and the associated geohazards. (Roger Urgeles Esclasans, University of Barcelona,