The unstable mountainsides of Åknes and Hegguraksla in Møre and Romsdal in Western Norway poses a great risk due to possibility of sliding out into the fjord and thus initiating a tsunami. The Åknes/Tafjord project was therefore initiated in 2004 to examine, monitor and enable a system that would give an early warning if the slide was to occur.

Inspections of the mountain sides are intended to reveal the scope and volume of the unstable areas, as well as to discover the mechanisms and movement in the rock mass. The volume of the Åknes is approximately 30-40 million m3, possibly even more, and moves annually about 3 - 10 centimetres.

Animation of innundation of waves (Hellesylt) after a rock slide of 54 mill. m3 at Åknes.

The participants in the Åknes/Tafjord project have previously warned that a landslide from the Åknes mountainside can cause a tsunami in Sunnylvsfjorden. Such a tsunami could be a threat to several communities, such as Stranda, Hellesylt and Geiranger. The admonitions have resulted in the establishment of a surveillance system that became fully operative in 2007. NGI contributes with equipment and data to sustain the system.

The surveillance system includes extensometers, measurements by laser, GPS, ground radar and instrumentation in boreholes.

Landslide with a potential tsunami
At International Centre for Geohazards (ICG) ,experts from NGI, NGU, UiO, NTNU and NORSAR cooperate closely and have from 2005 until today conducted analysis of potential landslides and tsunamis from the Åknes. Laboratory experiments at UiO and at SINTEF's laboratory in Trondheim have also been carried out.

The results confirm what was previously believed; that there is a recognized danger of a landslide with a following tsunami, but also that it is possible to find measures to reduce the risk.

A preparedness centre (see their Norwegian homepage) has also been established in the Stranda municipality. The centre will receive surveillance data from the site at Åkneset and Hegguraksla in Tajord and analyze the situation. This process enables the centre to send out warnings and arrange an evacuation if the probability of a slide has been found to increase.

Åkneset in the Synnylvs-fjord leading to Hellesylt and Geiranger (photo: Marc Derron, NGU).	Field survey of the open crack at Åknes, Prof. Bjørn Nilsen, NTNU and PhD students Guro Grøneng, NTNU and Vidar Kveldsvik, NGI.

A Model of the Landslide
Vidar Kveldsvik, at the Natural Hazard Department at NGI, has developed a numerical model of the landslide prone area. This has been part of his PhD studies, completed in June 2008. The model has been used to examine the factors that can cause a slide and to estimate the volume of slide mass that possibly could fall into the underlying fjord. 

 - The worst case scenario is that a large part of the mountainside falls out at the same time, Kveldsvik says. A reason to fear such a slide is that sliding surfaces as far as 120 meters below the surface of the headland has been found. Nevertheless, it is also possible that a slide will occur more gradually. In such an event a catastrophic tsunami is not likely to occur.

The data collected show that a higher groundwater level in the mountainside could dramatically increase the possibility of a slide taking place. Increasing groundwater level can be the result of intense precipitation, potentially combined with snowmelt. Through the same model Kveldsvik has also concluded that a large earthquake (1000 yr earthquake) probably would trigger a rockslide. Earthquakes are, in contrast to other factors that could cause a slide, impossible to detect and predict in advance.

Finally, the results suggest some actions that could prevent the slide from happening. The likelihood of a slide could be considerably reduced if the mountainside was to be drained so that the groundwater level could be kept stable. Drainage could be possible through a tunnel system inside the mountain. 

Illustration of analysis of tsunami generated by a rock slide at Åkneset 100 seconds after slide of 54 mill. m3 have hit the water. The colours indicates the wave heights (scale in metres on the right side).

Maximum surface elevation for a slide at Åkneset (yellow point) of 54 mill. m3. Please observe that the colour scale is logarithmic.

Tsunami analysis
Dr. Sylfest Glimsdal, at NGI's Natural hazard division, participated in improving and refineing the numerical models describing a potential tsunami resulting from a slide from Åkneset. A total of 12 different slide scenarios with different slide volumes and velocities have been  modelled and analysed. The new numerical models are more advanced and detailed than the initial analyses in 2005.

-Nevertheless, Glimsdal says; the results still show that a slide from the Åknes can cause a catastrophic tsunami in the nearby fjords. The calculations also show that the wave run-up could potentially be several meters high even tens of kilometres away from the site itself.

Maximum surface elevation in Hellesylt after innundation due to a slide with volume 54 mill. m3. The shoreline is shown with a turcoise coloured line.

  

Physical model
In addition to the numerical models, NGI has also participated in wave tests at the University of Oslo, where the Hydrodynamics Laboratory at the Mathematical institute has a wave path  that has been used to examine potential wave heights. These results are also used at SINTEF in Trondheim which has built a physical model in a scale of 1:500 of the fjord system and mountainside in the Synnylvsfjord and Geirangerfjord. To ensure that results from the physical model can be used to verify the numerical model, NGI's results have been applied when building the new model.

Video showing comparioson of numerical analysis and laboratory simulations.