Constraints on high-z disk formation

figure 1 from Robertson & Bullock (arXiv:0808.1100)

The authors investigate the claim by Genzel et al. (2006) that the
z=2.4 "disk" galaxy BzK-15504 formed very rapidly and early on (with
a correspondingly rapid accretion of mass), but shows no evidence of
a merger (because its velocity asymmetry is low). Using their
simulations of gas-rich disk mergers and taking into account noise
and PSFs, the authors figure out how their merger remnants would
appear viewed through SINFONI. In all four of their simulations
(each with different initial configurations), galaxies that would be
observationally classified as "disks" appear 100-150Myr after the 
merger. The above figure shows one particular simulation that they
claim matches the kinematic properties of BzK-15504 (shown below for
comparison, from Genzel et al. 2006) remarkably well.