In a recent paper, Kaplinghat and Turner (2001) (KT) advertise that MOND can be derived naturally in the CDM paradigm. They actually proceed to produce a more limited result: Every galaxy should have a transition radius, $r_t$, below which baryons dominate, and above which dark matter (DM) takes over; the acceleration at $r_t$ is nearly the same for all galaxies; and due to a coincidences this is of order $a_0\sim cH_0$. This follows from their tacit, intermediate result, whereby CDM halos of galaxies have a very nearly universal acceleration profile $a(r)\approx v^2(r)/r\approx A\hat a(r/\ell)$, where A is universal, and only the scale $\ell$ varies from halo to halo. (This remains so when baryons are added because they assume a universal baryon-collapse factor.) The KT scenario is phenomenologically wrong--observed galaxies are simply not like that. For example, it precludes altogether the existence of LSB galaxies, in which the acceleration is everywhere smaller than $a_0$. The phenomenologically sound outcome--i.e., the role of $a_0$ as a transition acceleration in high-surface-brightness galaxies--pertains to only a small part of the statement of MOND. There are several other, independent roles that $a_0\sim cH_0$ plays in MOND phenomenology, and other predictions of MOND, not related to the value of $a_0$, that are not explainable in the KT scenario. The results of KT also disagree with those of CDM simulations, which, as they now stand, do not reproduce any aspect of MOND phenomenology.
Read full abstract