The similarity between the drift vectors and the convection rolls model is striking. Explaining it thoroughly is more difficult.
The vectors are found in the report https://www.lmi.is/wp-content/uploads/2019/09/skyrsla.pdf
It is obvious that the drift vectors only approach the alignment of the convection rolls in Iceland. The uppermost rolls are aligned from NE to SW, and in the NW part of Iceland these rolls should be driving the drift. Rolls opposing the drift should be decoupled, and rolls following the drift should be coupled.
This article explains the basics of coupling and decoupling. https://www.researchgate.net/publication/310896846_A_mantle_convection_perspective_on_global_tectonics
Different layers can get coupled and decoupled, so the combined effect of lower layer can be felt at the surface with movements of the tectonic plate. An intersection layer is found between convection rolls layers, and between the upper most convection layer and the ductile lower most part of the tectonic plate as well.
With these factors in mind, it will be possible to analyse and trace the geological history.
To further show the double nature of the two rift zones of the southern half of Iceland, this map should be added:
The eastern margins of West Iceland Volcanic Zone and East Iceland Volcanic Zone are 3° apart from east to west along each parallel. The pulling effect of the North Americal Plate to the NW causes rifting, as the local small convection rolls underneath flow into opposite direction. This creates and area with rather northerly oriented drift vectors.
At Vatnajökull, the Eurasian plate slides along the side of the East Volcanic Zone, contributing to the fact that the geology of East Iceland has much more holistic appearence than that of West Iceland.
Magic Magma 