The fjords in North Iceland are rather large and wide. They look at first as randomly distributed with various shape. Comparing with the convection currents underneath, the pattern is not random at all. The convection rolls division line extending from the Reykjanes Ridge marks the three eastern fjords of Eyjafjörður, Skjálfandi and Öxarfjörður. The crossings between that line and the other equator lines of the next layer below mark the estuary of the relevant rivers flowing into the fjords. The next division line north of the Reykjanes Ridge line works in the same way, marking Hrútafjörður, Húnafjörður and Skagafjörður. The result is a ladder-shaped pattern for the fjords, when compared with the mantle convection rolls underneath.

The two layers forming the surface in this way (with help from the glaciers above during the ice age) are actuallly the subducted convection rolls, at this latitude only found close to 410 km depth. How can the upper rolls be a secondary factor? Actually they also have their say. The lower set of convetion rolls at the north coast seem to have a certain degree of coupling, and therefore the Tjörnes Fracture Zone mainly follows the division lines of those lower layers. In the same way, the lower layers have this effect on large scale formation of topography of the North of Iceland.