The pair of convection rolls under the continental shelf of Norway along 66°N is shown here:
The Convection Rolls
You can go further and visualize the next convection roll to the east, under the Norwegian mountains, as they span more or less the width of one convection roll.
The symmetry around the coastal point of Iceland at 66°N, 30° to East and West, is coherent with the analytical result that the span of one convection roll of lower mantle is 30° along each latitude. Still, the West Fjords Peninsula is not located over the division line of large convection rolls of lower mantle, but the tectonic drift has resulted in this symmetry.
The Symmetry Along 55°N of Coastal and Continental Shelf Points
The coast of Norway is 6° farther to the East (the span of 2 upper mantle convection rolls), where the main division line between lower mantle convection rolls is found. These are down-welling lines closest to the tectonic plate at 120 km depth below Earth’s surface. It is therefore to be understood that at those three points, the down-welling division lines also mark the division lines between topographic and bathymetric features.
The volcano Eyjafjallajökull became famous in 2010 when it erupted and airports all over Europe were closed. It is located at the western end of the mountain range of Eyjafjöll, and when revealing the mystery where volcanoes are located, the reason for the location of Seljalandsfoss becomes understandable. An EW-axis is found through Eyjafjallajökull, and the caldera of Katla as well.
Seljalandsfoss
The black line found on this map shows how the axis is formed from one corner of the relevant polygon to the other. Seljalandsfoss represents the end of the explicit part of fissure swarm of Eyjafjallajökull.
The volcanic zones of Iceland are the key to geology. A list is found in the newly published book, Náttúruvá á Íslandi:
Here the 7 volcanic zones are shown, names in Icelandic.
When this is compared with the convection rolls model, a theoretic pattern does emerge:
The volcanic zones of Iceland
A: The Snæfellsnes Volcanic Belt is basicly found end to end of two polygons, passing through the third, similarly as B and H.
B: The Reykjanes Volcanic Belt is based on one polygon, but the second part extending out to the ocean forms a continuity, so it resembles A and H.
C: The West Volcanic Zone covers two polygons in a similar way as F, but on most maps it does not fully cover the polygons from east to west. It then turns north around 65°N, coming to an end rather abruptly.
D: The Mid Iceland Volcanic Zone of Hofsjökull does provide a continuity of volcanic activity from the WVZ (C) to the EVZ (F). According to the system, it should be related to C, as the rifting process should be originated from there. The system is connected to a system added with dotted outlines, marke I, of Skagafjörður.
E: The South Iceland Volcanic Belt is found as a continuation of F, but is not a rift zone, as it does all belong to the Eurasian Tectonic Plate and is presumed to drift with it as a whole. The same type of tension is found theoretically to occur within the two polygons firstly for the one of Eyjafjallajökull and Katla, and secondly for the Vestmannaeyjar Islands polygon.
F: The East Volcanic Zone is the probably most clearly defined of any of the volcanic zones. The west side and east side are accurately marked by the sides of three polygons, extending to the southern part of the fourth polygon of Askja.
G: The North Volcanic Zone is oriented north to south, with volcanic systems arranged in a right hand en echelon pattern. The main polygons form the central axis, but at the sides, a large number of polygons are involved, as in the north, the network of division lines becomes more complex than in the southern half of Iceland.
H: The Öræfajökull Volcanic Belt has been quite obscure due to the fact it is mainly found under the glacier Vatnajökull. It convers two polygons, very much like A and B, extending between Öræfajökull and Snæfell.
I: This is actually an extinct volcanic zone, but as it was active rather recently (less than 800,000 years ago) I want to include it. This is the Skagafjordur Volcanic Belt, responsible for rifting within the fjord for a million years or so, according to Árni Hjartarson who wrote his PhD essay on the topic. The central axis is found exactly 3° west of the central axis of the NVZ (G), which is logical according to the convection rolls system, where similar conditions are repeated every 3° from east to west along each parallel.
The South Icelandic Seismic Zone and a part of the Tjörnes Fracture Zone are marked on the map for reference. The Grímsey Oblique Belt also connects with the Kolbeinsey Ridge north of Iceland, not shown on the maps.
This drawing can be best understood as a section of the equator plane of the Earth. Then all physical factors are combined in a 2D plane. Convection currents of lower mantle and outer core are represented with circles. Real flowlines are of course different, filling up the space of the cell froming the section of a convection roll. Developing a 3D model, the interaction of rotation of the Earth itself (and thereby centrifugal force) and convection currents have to be considered, in addition to the Coriolis effect along with the curvature of the Earth. As all those factors have been taken into account, the Convection Rolls Model has been completed.