[Previous] | [Session 106] | [Next]
J.M. Lotz (Johns Hopkins University), H.C. Ferguson (STScI)
We use a Bayesian - maximum likelihood analysis of the Hubble Deep Field to constrain the epoch of dwarf galaxy formation. Late formation of dwarf galaxies arises as a natural consequence of proposed solutions to the "over-cooling" problem in hierarchical structure formation. Although dwarf-sized halos are among the first objects to collapse out of a cold dark matter dominated universe, photo-ionization from the inter-galactic UV background and stellar feedback at early epochs may suppress or delay significant star formation in dwarf galaxies until redshifts ~ 1. Such late-forming dwarf galaxies may make up a portion of the population of the faint blue galaxies observed at intermediate redshifts. Previous attempts to understand the nature of the faint blue galaxy population have fit the binned number counts, luminosity functions, color and size distributions and compared the results to a handful of possible scenarios. Our approach sums the likelihood of observing each object in the HDF catalog given a dwarf galaxy formation scenario and computes the total likelihood of the given dwarf formation scenario. The parameters of the input model are then varied, and the model with the maximum total likelihood is determined. This technique does not bin the data in any way, tests a wide range of input model parameters, and allows us to quantify the goodness-of-fit and constraints on dwarf galaxy evolution.
The author(s) of this abstract have provided an email address for comments about the abstract: jlotz@pha.jhu.edu