Problems for modified Newtonian dynamics in clusters and the Lyα forest?

The observed dynamics of gas and stars on galactic and larger scales cannot be accounted for by self-gravity, indicating that there are large quantities of unseen matter or that gravity is non-Newtonian in these regimes. Milgrom's modified Newtonian dynamics (MOND) postulates that Newton's laws are modified at very low acceleration, and can account for the rotation curves of galaxies and some other astrophysical observations, without dark matter. Here we apply MOND to two independent physical systems: Lyα absorbers and galaxy clusters. While physically distinct, both are simple hydrodynamical systems with characteristic accelerations in the MOND regime. We find that, because MOND violates the strong equivalence principle, the properties of Lyα absorbers depend strongly on the (unknown) background acceleration field in which they are embedded. If this field is small compared to their internal accelerations, then the absorbers are more dense and about 10 times smaller than in Newtonian gravity with dark matter, in conflict with sizes inferred from quasar pair studies. If, however, the background field is rather large, then the absorbers take on properties similar to those predicted in the cold dark matter picture. In clusters MOND appears to explain the observed (baryonic) mass-temperature relation. However, given observed gas density and enclosed mass profiles and the assumption of hydrostatic equilibrium, MOND predicts radial temperature profiles that disagree badly with observations. We show this explicitly for the Virgo, Abell 2199, and Coma Clusters, but the results are general and seem very difficult to avoid. If this discrepancy is to be resolved by positing additional (presumably baryonic) dark matter, then this dark matter must have ∼1-3 times the cluster gas mass within 1 Mpc and about 10 times the gas mass within 200 kpc. This result strongly disfavors MOND as an alternative to dark matter.