For what it's worth, my theory (everyone else is assuming so I may as well) ...
Using Schubinho's calculations as a guide it hit the ground flat but slightly tail first, enough to loosen the tail but not to actually separate it, slid for a bit until the tail cone and subsequently the whole tail section came off (from here on there's lots of debris from stuff falling out the rear and from the fuselage breaking apart). The tail tumbles at least once, hence the damage on it, until it stops. The wings and centre section being the strongest part of the aircraft and still possible powered by at least one engine remain virtually intact until they hit the fence. In effect the rest of the fuselage has broken up around the wings and and wing box.
The only thing that argues against this in Schubinho's picture is that, if the track is of the belly of the aircraft there should be tracks from the engines sliding along the ground
(Looks like Takata beat me to it!)