Looking at the modern GPS based map representing tectonic drift vectors, it looks like Africa and Eurasia are more and less drifting in a parallel way:

Looking at the Mediterranean and its surroundings, it is obvious that the African plate is being subducted below the Eurasian plate. The relative movement between those two plates consists therefore mainly of a northward component. The convergent boundaries are quite complex, as can be seen here:

This is a clip from the presentation found at: https://www.youtube.com/watch?v=cqK-CbuM3Eo (Geoscience Information for Teachers Workshop at EGU).
If we try to simplify this as much as possible, the two main features of the North Coast of Africa, and the Adriatic Sea should be looked at:

It is known that Italy has rotated and the Adriatic Sea has thereby been enclosed with slab reaching under both Italy and the Balkan Peninsula. The African Plate subducts below the Aegean Sea, Sicily and Turkey. Subduction from the north and the appearance of ocean floor due to spreading is omitted here. The lines resulting from the Convection Rolls Model can be seen, and of course the main purpose of this study is to compare that model with what is known about the geology of the area. For this short post, it is mainly pointed out that the North Coast of Africa coincides with the 32nd parallel where the convection rolls are aligned directly N-S. One aspect of that latitude is that two downwelling lines of two different layers coincide, and thereby also two upwelling lines of two layers coincide. This creates the opportunity for convergent boundaries to appear, if the downwelling part is the dominant factor (no-slip) and the upwelling part becomes neutral (slip). A consequence of this is that extension within the area is then mainly caused by the north component of drift of the Eurasian Tectonic Plate, as the subduction is mainly related to the conditions close to 32°N.












