Over counting in first bin of longitudinal profile

@Nikos showed in Karlsruhe an issue with the first bin of the longitudinal profile where a very high number of particles appeared that did not make sense. This is caused by backscattered particles that are behind the initial point of the shower axis. This particles are given by the shower axis a projected grammage of zero, being counted always only in the first bin of the longitudinal profile. The images show a longitudinal profile without the fix (left) and with a fix that does not count tracks with initial and final grammage of zero (right), referred to the injection point, in this case at 15 km above the surface of the Earth.

The shower run was (with Fluka): ../bin/corsika -A 1 -Z 1 -E 10000000 --emcut 0.01 --hadcut 0.01 --mucut 0.01 -f testLongProfileBugFix2 --emthin 1e-06 --max-weight 100 --injection-height 15000. --observation-level 0. -s 2

The plot on the right, with the fix, shows the first bin with 6 hadrons. This is caused by some neutrons from what I have seen. They start at zero and propagate x amount of grammage. By how the shower axis works, they probably come from above the injection point, so my first guess is that a backscattered particle interacted before that initial point and produced the neutrons that travel downwards again, but not 100% exactly sure.