The fourth geological mystery in this series of articles concerns thrust faults and thrust nappes – geological structures that are prevalent in much of our mountainous landscape. It is also about the art of seeing what you cannot quite bring yourself to believe.
By Fridtjof Riis,
Established 28 October 2015
To the west and north, the Hardangervidda mountain plateau borders the Hårteigen, Hallingskarvet and Hardangerjøkulen mountains. The top layer in these landmarks is several hundred metres thick and consists of hard gneiss rocks. The gneiss rocks are lying on top of shale and limestone, which are less resistant to erosion. This layer of gneiss is part of the thrust nappes in the Caledonian mountain range that extends through Scandinavia from Jæren to Hammerfest.
The sketch shows how a thrust nappe is defined.
Sketch of thrust nappe and reverse fault
Reverse faults are formed where the Earth's crust is under compression. When a reverse fault moves, older beds are moved on top of younger ones.
A thrust fault is a reverse fault that follows the boundaries of the beds. The beds in the hanging wall are pushed, and settle above the beds in the footwall.
A thrust nappe consists of rocks that have been pushed several kilometres. The lower boundary of a thrust nappe is defined by a thrust fault. In technical language, this thrust fault is often called a decollement or basal detachment. A typical thrust nappe rests more or less horizontally on top of a younger sequence of strata.
As a postgraduate student in the 1970s, I learned about the thrust nappes in Southwest Norway when Johan Naterstad, Roy Gabrielsen and Arne Solli were my guides in the Hardangervidda-Ryfylke project. Much of my geological experience up to that point stemmed from the Oslo area, which is known for its fossil-rich sequence of Cambro-Silurian strata. Locals with an eye for rocks are very familiar with the folds that characterize this sequence of beds. The layers have been folded because they are located in the outer part of the Caledonian mountain range. It was however a big change for me to go from the folds in Eastern Norway to the much more intense deformation that is seen in the thrust nappes nearer the core of the Caledonian mountain range. What kind of forces were able to push large mountain masses more than a hundred kilometres on top of younger layers?
The mountain problem
Two mountains stand out as the the Scandinavian reference sites of thrust nappes where older types of bedrock lie on top of Cambro-Silurian sedimentary rocks. They are Hårteigen on the Hardangervidda plateau in Norway and Åreskutan in Jämtland, Sweden.
As regards Åreskutan, this phenomenon (“the mountain problem”) was described and interpreted as thrust faulting by Alfred Törnebohm in the 1890s, following a huge surveying effort spanning more than 20 years.
Törnebohm compared the fossils in the Trondhjem field with the similarly aged fossils at the base of Åreskutan and argued that the thrust nappes on the Norwegian side must have been transported at least 100 kilometres relative to the Swedish Cambro-Silurian sequence.
At that time, thrust faults on a smaller scale had been proven in the European Alps and in Scotland, but to many people it seemed physically impossible that such large distance thrust faulting could have taken place.
Törnebohm had set off a discussion that was to continue for more than 50 years, raising questions that are still not fully answered. The beginning of this discussion has been well summed up in The Making of a Land – Geology of Norway.
When plate tectonics had its breakthrough in the 1960s, geoscience found an explanation for why large shortenings had occurred in the mountain ranges. The Caledonian mountain range, the Alps and Himalaya are all examples of continental plates colliding in such a way that one plate is pushed underneath the other one.
Hårteigen (data source www.norgeskart.no Kartverket-Geodata)
In the Hårteigen landmark on Hardangervidda, thrust faulted bedrock gneiss is lying on top of younger rocks. Hans Reusch had entertained the same ideas as Törnebohm, and when Törnebohm had published his interpretation of Åreskutan, geologists Reusch, J.Rekstad and K.O. Bjørlykke travelled to Hardangervidda to test the thrust fault theory. Their conclusion was that the gneiss rocks had been overthrust, but they found no evidence of such huge thrust distances as Törnebohm had published. An alternative theory for Hårteigen at that time was that the gneiss rocks had been formed by magma that had intruded into the shales and transformed them. Such theories lived on until plate tectonics was generally accepted. The problem with these theories was that no intrusive rocks had been observed anywhere in the shales.
Thrust nappes in the big mountain ranges
In collision mountain ranges such as the Himalayas or the Scandinavian Caledonides, there is a central zone where the crystalline basement is involved in the deformation.
During the collision, and as the profile is compressed and shortened, a stack of thrust nappes forms from the outer edge of the collision zone and further out on the foreland. In an active mountain range like the Himalayas, powerful earthquakes will occur in this stack due to the thrusting process.
Thrust nappes contain material from the basement that was involved, mixed with the sediment beds that were originally positioned above this basement. Material from the central collision zone will be at the top of the stack.
The nappe stack slides on decollements located in the sediment beds above the stable basement, so that the shortening is absorbed by the thrust nappes rather than deeper down in the Earth's crust.
One talks of “thin skin tectonics” in the stack of thrust nappes and “thick skin” in the zone where basement is involved.
Schematic profile of collision mountain range with dimensions like those in the Himalayas. The stack of thrust nappes shows a possible distribution of basement (blue) and younger sediments (green).
In the Himalayas, the distance from the tallest mountains to the undisturbed sediments on the Indian Plate is about 250 kilometres.
In the Scandinavian Caledonides, we find a similar stack of thrust nappes preserved in a belt that extends from the area where the basement is involved in the deformation and approx. 100 kilometres towards the east. The outer part of the thrust nappe has mostly been removed by subsequent erosion, but it has been preserved in the Oslo Rift. There one can confirm that the belt of thrust nappes and thrust faults is about as wide as that in the Himalayas.
The inverse stratigraphy
In the Scandinavian Caledonides deep erosion has exposed the deep parts of a 400 million year old mountain range that was of a similar size to the Himalayas. Here, a tectonic stratigraphy is now described where the nappe stack is divided into different units. Regionally, one often speaks of the lower, middle and upper nappe units (See The Making of a Land – Geology of Norway, 2013).
The figure shows different styles of deformation in the lower units of the nappe stack in the rocks within the Ritland Crater.
|5 Thrust nappe (deformed basement gneiss)
4 Overthrust rocks (deformed sandstone)
2 Black shale
1 Shale and siltstone
The lower thrust nappe units in the rocks within the Ritland Crater. Yellow line: the bottom thrust boundary, at the top of an upwards coarsening Cambrian-Ordovician sandstone unit. The sandstones in the thrust nappe are deformed by folds and reverse faults (white), while the in situ part of the sandstone is not much deformed. Light red line: lower boundary for the next thrust nappe, which contains crushed bedrock types. The height of the profile is about 150 m.
Above the local sandstone thrust nappe there is a thrust nappe made up of schistose to foliated gneiss rocks. This thrust nappe covers a large area and cannot have a local source because the old basement surface has been preserved in this area. It can be classified as belonging to the middle nappe units.
In the upper part of the nappe stack, the thrust movement has been absorbed by thicker zones with ductile deformation, the minerals have recrystallized, and they typically have extensive lineation in the direction of movement.
The upper thrust zones in the nappe stack have been formed at a higher temperature and pressure than what existed at the bedrock surface where they are now positioned.
The upper nappes must therefore have been formed when they were deeply buried closer to the collision zone, while during the last transport stage they were resting passively on the lower thrust nappes.
The reverse stratigraphy can be explained as follows: While the collision was going on, the stack of thrust nappes moved progressively onto the Baltic Shield, always with the most active zone of motion at the bottom of the stack.
The basement windows
Inside the Caledonian mountain range is a belt characterised by basement windows as shown in the figure.
Map of Scandinavia with basement windows (red).
H marks the window of Haugalandet.
In this belt, the Precambrian crystalline basement was involved in large dome or fold structures because of the collision. To the west of the belt, the basement was strongly involved in the deformation, while it was not much affected in eastern areas. Between the green and the blue lines in the figure, the bedrock is mainly built up of thrust nappes.
The area for my postgraduate thesis at Nedstrand borders the basement window at Haugalandet (H in the map). The basal Cambrian sandstone at Nedstrand rests on the basement, which has been folded up to a height of more than 600 metres along this flank.
Inside the basement window, the contact between the older gneiss and Cambrian metasediments has been cut by thrust faults that create wedges of phyllite inside the basement. Under such conditions, bedrock wedges have been pulled loose and transported further as thrust nappes.
A thrust nappe is born
The most spectacular example of a disconnected thrust nappe at Nedstrand is found where the Vatsfjord meets the Yrkefjord.
The Gaupefjellet mountain contains a 300-400 m thick body of Precambrian gneiss overlying Cambrian basal strata. Towards the north, this body seems to be connected with the basement window; towards the south on the Nedstrand peninsula, it terminates as a wedge inside Cambrian-Ordovician phyllite.
The local thrust nappe in the Gaupefjellet mountain. Red line: the top of autochthonous basement Yellow: the thrust boundary underneath detached basement. White: Shale/phyllite above the detached basement. Gaupefjellet mountain is about 400 m high.
The gneiss was disconnected from the basement window together with the overlying phyllite and thrust to the south-east. In Nedstrand, thin zones of mylonitised gneiss with small bodies of less altered gneiss are located inside the phyllite package, marking the thrust boundary towards the south and east.
Sketch of the thrust nappes in the Boknafjord area. Black lines: Sole thrusts between different nappe units. Thrust nappe 1 and 2 (typically 500-1000 m thick) consist of strongly deformed basement gneisses (orange) and Cambrian-Ordovician shale/phyllite. These nappes may be sourced from the areas around the Haugalandet basement window (thrust distance less than 100 km), while nappe unit 3 has been transported farther. Square shows location of Gaupefjellet.
The sliding surfaces
There was another aspect of the “mountain problem” in the Hardangervidda-Ryfylke project which bothered us, and which Johan Naterstad kept coming back to: How can thrust nappes move across vast distances in such a way that the deformation is only absorbed in a thin decollement surface?
This is particularly true for thrust faults which occur under a low degree of metamorphism, at burial depths down to between 5 and 10 km. Under such conditions, the movement can take place along sliding surfaces or very thin thrust zones, as we see examples of e.g. in the Hjelmeland mountains east of the Ritland Crater.
The thrust zone is marked by the base of the hammer and consists of a white weathering layer of extremely deformed mylonite. Dark rock below is autochthonous, bedded sandstone, light rocks above consist of deformed sandstone.
If the thrust occurs under pressure and temperature conditions where extensive recrystallisation of minerals is taking place, zones of fine-grained mylonite will form. At shallower depths one can expect crushing and movement of sediment grains.
The thrust faults are created over a long time as the sum of many earthquake movements, which must then have reactivated the same sliding surfaces repeatedly.
In modelling the mechanics of thrust movements, it is still considered an open question why the friction in the sliding surfaces during major earthquakes is so low that the large thrusts can occur.
A common hypothesis is that an overpressure builds up, e.g. in clay zones, that can carry much of the superimposed weight. Sliding surfaces can also become slippery because they have been lubricated with salt. Such factors might obviously affect the shape of the thrust nappes and facilitate large thrust distances, but thrusting is such a typical feature of collision zones that it is unreasonable to imagine it is directly dependent on local conditions in the sequence of beds involved.
The gravitational forces created by the topography in the mountain range contribute to the shear stress along the sliding surfaces, but they do not reduce the friction.
Landslides can also create thrust faults
When I was involved in studying the Storegga slide in the early 2000s, we discussed the significance of a compression zone on the slope in the middle part of the landslide (yellow area in the figure). This compression zone was described in publications by Petter Bryn and his colleagues in 2005.
Seismic profiles through the compression zone show interesting analogies with the geology in thrust nappes. The thrust movement arose because vast landslide masses collided into the side wall of the landslide.
The landslide masses compressed the sediments beyond the wall. The basal sliding surface used by the slide was mobilized and formed a decollement surface, and the overlying masses were compressed over a distance of more than 20 kilometres.
The phase of the landslide that created the compression zone also caused a tsunami in the North Sea Basin and the Norwegian Sea. It is a reasonable interpretation that the thrust movement in the compression zone was triggered by this landslide and took place in the most powerful stage of the landslide.
If so, the time scale for the triggering and mobilization could be comparable to a major earthquake. The regularity of the structures that were formed, and a seemingly higher degree of compression towards the west, would indicate that the triggering of the whole structure happened as one event, and that the sliding surface had very little friction.
Seismic profile through the compression zone in the middle of the Storegga landslide, location shown with a blue line in the map on the right. The orange reflector shows the sliding surface approx. 600 m below the seabed. A train of fold structures linked to reverse faults that are connected to the sliding surface, has been created on the seabed. Some of the faults have been interpreted (thin black line). The belt with folds and reverse faults is more than 20 km wide. In the thrust direction, the movement ends along the sliding surface in a slightly larger fold and a reverse fault which have transferred the movement to shallower sliding surfaces about 300 m and 200 m below the seabed (blue line). In the map on the right, the deep part of the Storegga landslide (phase 2) is shown in green. Protruding heights in the landslide prevented the southernmost slide masses from turning north, and they therefore ran into the wall and formed the compression zone (in yellow). The map is modified according to Bryn et al. 2005.
A neglected factor?
The studies of the Storegga landslide showed that the sliding surfaces were formed in contourite sediments. They have lower shear strength and higher sensitivity than the glacial sediments in the landslide area, but not extremely low values.
Could there be a hitherto neglected factor which helps reduce the friction on sliding surfaces both in powerful earthquakes and major landslides?
One option would be to take a closer look at processes in materials that are subjected to shock deformation.
This will be one of the topics in the next mystery.
It seems appropriate to conclude by giving credit to Alfred Törnebohm, who chose to trust and publish his observations, even though geoscience at that time had no way of explaining them.
To me, he was the first person to exemplify that geoscience advances when we believe what we see.
It is not always enough to see what we believe.