Two minutes without friction

In geology, we have to consider both the long and short time scales. Events that change the course of geological development can occur very quickly. Large landslides, earthquakes and meteorite impacts have a time scale of just a few minutes.

A few minutes of extreme energy release, when rocks are crushed by shock deformation.

It can be difficult to interpret the traces of such catastrophic events. They are rare, and if there are any eyewitnesses, they are usually occupied with survival rather than geological observations during these few minutes.

In order to interpret the traces of catastrophes of the past we must combine fieldwork with models and laboratory studies of how rocks and sediments behave under shock.


Cock in the scree


Rockslide in Gloppedalsura.

Rockslide in Gloppedalsura.
Central lobe with large blocks, and lobe on left that flowed along the bottom of the valley on a relatively low gradient down towards Gloppevatnet.


The Gloppedalsura scree in Rogaland County (photo above) is one of the largest screes north of the Alps. There are legends about it. People said that the rockslide buried a farm, and that a cock could be heard inside the scree for seven years.

During the landslide in Gloppedalen, which actually occurred just after the inland ice sheet retreated, the rock material filled up the entire bottom of the valley and flowed along the valley for a distance. It appears as if the friction was less than one would expect.

There are examples of rockslides with more extreme movements than the Gloppedalsura scree. A friend of mine recently returned home from summer holiday in Canada and showed me a reprint of the geological report from 1904 about the Frank Slide in Alberta.


Frank slide. Source: Google Earth  

Frank slide.
Source: Google Earth


He had never before seen a rockslide like this even though he grew up in western Norway. The rock mass consisted of fractured limestone that fell from a height of approximately 900 meters down into the valley and continued at a speed of about 50 kilometres an hour over the river and nearly two km up the slight slope on the other side before it suddenly came to a halt.

Rocks with a diameter of several metres flew through the air, and the entire landslide area was covered by a layer of 15-20 metres of rock. Part of the mining town Frank was buried by the rock mass.

The geologists reported that the shape of the rock mass resembled a surface which could have been formed by viscous fluid flow. It appears as if this rock mass moved at a lower friction than in “normal” rockslides with less energy.

Wyoming is the home of Heart Mountain where blocks of Carboniferous limestone rest on top of a Paleocene basin. The limestone is most likely remnants of a package several thousand metres thick that moved at least 40 kilometres.

The tectonic boundary between these units has been interpreted as both a slope failure and thrust boundary through the years. Today, this Paleocene rockslide is considered the largest ever proven on land.


Mjølnir and Ritland: Structures that needed an explanation

In the late-1980s, I worked on interpreting the Norwegian Petroleum Directorate’s 2D seismic data to map the unopened areas in the Barents Sea.

One of the seismic lines showed an uncommon structure in the Jurassic layers near the seabed. For lack of a better explanation, I added it to the map as a plug of magmatic rocks.

I was not entirely pleased with this solution as there was no known volcanic activity in this area.

A few years later, Steinar Gudlaugsson came up with an alternative interpretation. He proved that there was a 40-km ring-shaped structure around the plug, and claimed that the plug was the central peak in an impact crater.

Subsequent drilling operations and analyses showed shock deformation of quartz grains and that Steinar was correct. The Mjølnir crater was proven. And I wondered: Why do the large impact craters have a central peak, and why was there not a deep crater depression around Mjølnir?


Map of the Mjølnir Crater in the Barents Sea, showing the base Cretaceous surface. 

Map of the Mjølnir Crater in the Barents Sea, showing the base Cretaceous surface.
The central peak is approximately 200 m higher than the circular structure, 40 km in diameter. The crater was formed in unconsolidated sediments at the Jurassic/Cretaceous boundary and most of it is today buried below younger strata. Kilde: 


If you come from Hjelmeland in Ryfylke and go into the Ritland structure, the landscape opens towards a slope of alum shale. This is surrounded by basement, more than 300 metres underneath the sub-Cambrian peneplain in the area.

The first time I visited the area with colleagues in the mid-1980s, we were looking for fossils, but were also surprised to find a sedimentary sequence of strata with thick conglomerates and sandstone underneath the alum shale.

Over the next 15 years I visited the structure many times on hikes, without understanding the geometry of the structure and how it could have formed.

I tried to interpret it in light of what I already knew: the sub-Cambrian peneplain, basal conglomerate, folds and thrust faults, but what I saw did not correspond with anything I had seen in the mountain range before.

In 2000, I mapped the extent of the thick conglomerate layers located between the shale and basement. It became clear that these were fans created by slides that were older than the shale. Moreover the map of the slides defined a circular structure that was nearly three kilometres in diameter.

It was time to mobilise the geologists at the University of Oslo, home of the Norwegian community for researching meteorite craters.

Johan Naterstad showed me cores through the entire stratigraphic sequence in the Gardnos crater. Henning Dypvik conducted several rounds of fieldwork and excursions. And guided me through the technical literature. The pieces were starting to fall into place.


View from the north over the Ritland crater. 

View from the north over the Ritland crater.
Green ring: Boundary for sediments and thrust nappes deposited after the crater was formed. White line: Boundary for the preserved part of the crater wall. White arrow (on left) points to the crater rim.


Dry rock that flows

In 1989, Jay Melosh published a paper on crater formation that should be read by everyone who is interested in shock deformation and the short time scale in geology. He compiled and compared data from craters on Earth, the planets and moons in the solar system, and modelled the physical processes.

During the actual crater formation, the rock behaves as a “Bingham fluid”. This means that for stress under a certain level, it will undergo plastic deformation, in the event of shearing movements, but above this level it will flow like a viscous fluid.

One can differentiate between the initial, or transient, crater shape that is formed over a few seconds upon collision between the Earth and the projectile from space, and the final, modified crater.

Over the course of a minute or two the rock that is crushed in the collision will flow back down into the transient crater depression. This period of time is called the modification phase, and leads to the final shape of the crater.

The larger the projectile, the longer time the collision and modification phase will take.

In large craters, this backflow creates a central peak, just like when raindrops hit a water surface and the water comes back up from the impact points. The flow is quick. The speed is about the same as one would observe in large rockslides, in the order of 50 km/hour.

The processes are all the same on Earth, on Mars and the moons in the solar system, and do not depend on water being present.

Melosh also writes about ejecta: particles that are ejected from the crater due to the collision.

Near the crater edge one would expect to find a continuous sheet of ejecta. Further out are deposits of boulders and rocks that have followed ballistic trajectories and that decline in size as you move away from the crater edge. Think about what happens when you throw a rock into water: A sheet of water comes up along with a spurt of droplets.


Ejecta tell us about the energy of the impact.  

Ejecta tell us about the energy of the impact.
15-50 metre thick mega-blocks (yellow line) that have been thrown nearly 1 kilometre away from the crater edge. The sub-Cambrian peneplain is shown with a red line.


The truly pioneering aspect of Melosh’ work is that he also introduced and calculated a physical model for how the flow takes place.

The idea is that the rock around the collision point is broken up and the fragments collide with each other under the powerful tremors triggered by the vast energy from the collision. The motion of the rock can be modelled as a fluid flow, but in detail it flows as a result of innumerable violent collisions between the fragments it is split into.

The flow stops when the tremors fade out – after one or a few minutes - and the crater will then solidify in its final shape. The process is widely accepted in the communities that research impact craters and is called acoustic fluidisation. The rock behaves like a liquid even though it is dry.


In the field with a model

Henning Dypvik from the University of Oslo found shock structures in quartz grains in samples we collected from the Ritland structure in 2007. The finding led to a larger project financed by the Research Council of Norway to learn more about this crater that had now been definitively proven.


Microscope image of quartz grain with shock lamellae.

Microscope image of quartz grain with shock lamellae.
Such lamellae can only be formed under the extreme pressures created by an impact.


Valerij Shuvalov from the Russian Academy of Sciences in Moscow now joined the team to create a mathematical model for the crater formation.

It was difficult to prove the impact crater origin because it has been modified by geological processes over more than 500 million years. The impactor that formed the Ritland crater occurred in the Early or Middle Cambrian and collided with the shallow sea that covered the sub-Cambrian peneplain at that time.

The crater was later infilled with marine sediments, buried under Caledonian thrust nappes and partially eroded and exposed again during the ice ages.

When Elin Kalleson, Abdus Azad, the master’s students, Henning and I went into the field, it was important to keep track of exactly where we were in relation to the original crater. The combination of fieldwork with an exact model we had placed into a terrain model was very useful.


Model of the crater (Shuvalov), merged with a terrain model.

Model of the crater (Shuvalov), merged with a terrain model.
Seen from the west. Orange dots show the lower boundary of post-impact slides and green dots show the base of Middle Cambrian shale. 


The Ritland area turned out to contain much more than we could have anticipated. It shows unique profiles from central parts of the crater, along the crater wall, through the crater rim and four kilometres through ejecta, all the way to where the rock material from the crater becomes too small to recognise without a microscope.

Such profiles are very important because the physical models for crater formation are largely built on photos from planets and moons, on laboratory experiments and mathematical modelling. On Earth, the models can be calibrated with studies in the field.


The southern crater wall of the Ritland crater.

The southern crater wall of the Ritland crater.
Basement gneiss crushed by the impact. Aerial photo from drone (A. Deryabin). The exposure is about 50 m long.



Energy and tremors – not only in impact craters

The tremors from a major earthquake last for about one minute. The energy from powerful earthquakes is so large that it forces the axis of the Earth to tilt slightly.

The breccias that form in a tectonic crushing zone can in some cases resemble the breccia found in the wall of the Ritland crater.

Seismologists have trouble explaining why the displacements during the major earthquakes appear to occur virtually without friction even at considerable depths where very little fluid is present.

One day in the field I asked Elin Kalleson whether it was possible to use the acoustic fluidisation theory to analyse the movement in the slip zone of the fault during the most powerful earthquakes. “Sure,” she said. “Of course. Jay Melosh published an article on this in Nature in 1996. I can give you the reference.”

The energy triggered in the Storegga slide caused a tsunami – and made its own impact field. This was published by Petter Bryn and his colleagues in 2005 and was described as an “impact zone” where the seabed topography flattens at a water depth of about 2500 metres.

The slide debris travelled very fast down the submarine slope and settled on a flatter part of the seabed. The figure shows this impact zone, which is approx. 80 kilometres wide.

In seismic cross-sections, the soft sediments in this “crater area” are transparent, no clear bedding can be seen. The impact zone is surrounded by mounds of diatomitic mud, which has very low density and therefore significant buoyancy. There is a low compression ridge outside the mound.

Which mechanisms transfer the forces from the fallen slide debris to the compression ridge in these soft sediments?


3D view looking down along the Storegga slide towards the “impact area”

3D view looking down along the Storegga slide towards the “impact area”


The acoustic fluidisation theory can be used on rock that is exposed to extreme tremors in a short period of time (a few minutes).

Within the community that researches the development of our solar system, this theory is generally accepted and used for impact craters and major landslides. It seems to me as if the theory is less known within classic geology.

The Ritland project has shown that the conformity between geological observations and physical models is very good, both in crater and ejecta.

I believe it would be very useful to test and attempt to calibrate this theory for other natural phenomena where large volumes of energy are released in a short amount of time, such as landslides and major earthquakes. There are a myriad of mysteries here that await explanation.

Topics: Geology