Event #1805: Storegga slide tsunami, North Sea Doggerland

A prehistoric land mass once connecting Britain to mainland Europe may have been wiped out by a 5m-high tsunami, according to new research.

Some 8,000 years ago, a devastating subsea landslide off the coast of Norway generated a wave which overran the island of Doggerland, a low-lying Atlantis, which is now completely submerged.

“It would have completely inundated the landmass,” says Dr Jon Hill, one of the team of researchers from Imperial College London, who have submitted their findings to the journal Ocean Modelling. “Any humans living there would have suffered a catastrophic event.”

There could have been inhabitants on the island at the time the waves struck, though this is still unclear from the evidence. The Imperial team’s computer models suggest that Doggerland was mostly less than 5m above sea level at the time of the tsunami, suggesting flooding would have been extensive.

The trigger for the tsunami was a landslide named Storegga, where 3,000 cubic km of sediment collapsed in the North Sea.

“That’s a lot of sediment,” says Dr Hill. “Three hundred times more than all the rivers in the world hold in a year, which is what makes this event so unique. […]

The tsunami would have also reached the UK and the Imperial team’s is the first study to measure its impact on British shores. Previous studies had always looked at the effect of the Storegga tsunami on the Norwegian coast.

Professor Martin Bell, head of Archaeology at Reading University, said: “What we know from the work that has been done in Doggerland is that there are major episodes of sea change around that time, of which the tsunami was the most dramatic. We know a lot about the slide itself, but the extent of the impact on east-Britain has been less studied.

There are some archaeological sites that have been flooded there. There are one or two sites in Scotland with known Mesolithic sites under layers of gravel… they clearly come from the tsunami because we can tell using carbon dating.”

Previously, however, two axes from Neolithic times, a period post-Storegga, had been discovered in an area of the North Sea termed the Brown Bank, which would have been affected by the tsunami.

“The axes suggest there may have been a little land surviving into the Neolithic times,” says Bell. “It is difficult to prove conclusively because the land has been covered over completely… we can’t just relate it to sea level change to find where those islands were.”

The Storegga slide was a large, tsunamigenic submarine slide off the coast of Norway that occurred around eight thousand years ago. The volume of material in the slide has been estimated at 3000 cubic kilometres – enough to cover Scotland to a depth of 30m. The tsunami generated had run-up height of around 20m on the Norwegian coast, 10m in Shetland, a few meters on the Scottish mainland coast and reached as far as Greenland. Run-up height can be estimated in certain locations via tsunami deposits, but these are not preserved everywhere. Numerical simu- lation of the slide and resulting tsunami can therefore help constrain how the slide moved by matching observed run-up heights and providing run-up estimates at other locations. The most accurate simulations need high resolu- tions near the coasts, particularly near tsunami run-up observations, in the slide region and to accurately calculate effects such as funnelling. However, there is a need to cover the whole of the Norwegian-Greenland sea to avoid issues such as model coupling and doing so at a constant resolution is computationally expensive. One recourse is to use multiscale modelling where the resolution of the model can change by orders of magnitude across a simulated domain. Here, we present a multiscale model where resolution varies from 500m to 50km across the domain thereby accurately resolving the coastlines and changes in bathymetry, but is still able to model the entire Norwegian-Greenland sea with reasonable computational effort. The model used is Fluidity, a finite-element, un- structured mesh model. We use a rigid block slide to initiate the wave and then track its propagation across the domain. Using this model we show how mesh, coastline, and bathymetric resolution affect the simulated wave. We also show the effect of including palaeobathymetric changes, due to ice loading and unloading, which have been neglected in previous numerical studies of the Storegga slide. This is the most complete and detailed simulation of the Storegga slide to date.