The faint blue galaxies problem
No-evolution models of galaxy number counts (GNCs) consider that the local galaxy populations have experienced no change in their numbers and intrinsic physical properties (such as in color or luminosity) with time. Therefore, these models predict GNCs in any band by displacing the local galaxy mixing in redshift. The only change in the apparent magnitude of the galactic populations with redshift will be due to the considered cosmology (usually known as K-corrections).
In the '90s, the first number counts data using CCD detectors and automatic detection algorithms were obtained. They showed an excess of galaxy number over the prediction of no-evolution models in the fainter magnitudes (B>~22 mag), whose origin was unknown (Tyson et al. 1988). This excess becomes more noticeable in bluer bands (Koo 1986 ApJ, 311, 651; Songaila et al. 1990, ApJ, 348, 371; Jones et al. 1991, MNRAS, 249, 481), being nearly zero in redder NIR bands (J, H, K).
The next GNC models considered that galaxies evolve in luminosity according to the evolution of their stellar populations. The "pure-luminosity evolution" models (PLE models) associate a given standard star formation history to each local galactic population. Tracing the change in luminosity of each population back in time due to the imposed star formation history, the global change in luminosity for that galactic population can be derived. When observational GNCs were compared with the GNCs predicted by PLE models, the excess in the fainter magnitudes was, although less noticeable, still present (Gardner et al 1996, MNRAS, 282, L1; Pozzetti et al. 1996, MNRAS, 281, 953). This excess of galaxies over no-evolution and PLE models at faint, blue magnitudes is traditionally known as the "faint blue galaxy problem".
Thanks to the high spatial resolution of the HST, morphological GNCs were obtained (i.e., GNCs for different morphological galactic types). Several studies (as, e.g., Odewahn et al. 1996, ApJ, 472, L13) could identify this excess: they were basically late-type galaxies and merging systems (Sd/Irr+M, see figure below). Later spectroscopic studies have shown that the galaxies responsible of the excess in faint ,blue magnitudes are in fact a mixing of three different galactic populations: irregulars at low z, blue dwarfs with star formation bursts at z~0.3-0.6, and peculiar galaxies at high redshift (z~1-3). But the weight of each population in the excess of GNCs is still unknown.Therefore, the origin and evolution until z=0 of the FBGs responsible of the blue excess is still a mistery.
In general, GNC models need to consider a strong evolution in the luminosity function (Metcalfe et al. 1995, MNRAS, 273, 257) and/or a high merging galaxy rate in order to reproduce this excess in the blue bands (Hogg et al. 1997, MNRAS, 288, 404; Gardner et al. 2000, ApJ, 542, 79; Fried et al. 2001, A&A, 376, 788). But, usually, these models overestimate the GNCs at faint magnitudes in redder bands. Why is this excess of blue galaxies exist? How can we reconcile the excess of blue. faint galaxies (over no-evolution GNC models) with the galactic populations in the NIR GNCs, apparently non-evolutive?
|Figure 1: Morphological GNCs in the B band in several fields, according to the morphological type. Panels: a) Total GNCs; b) GNCs of E/S0's; c) GNCs of Sabc's; d) GNCs of Sd/Irr/M's. the different symbols represent the morphological GNCs obtained for the different fields. Blue lines correspond to the prediction of a no-evolution GNCs model, using the local luminosity function obtained by Marzke et al. (1994). Notice that the excess of galaxies over the no-evolution model in the faint magnitudes is basically due to Irr/M galaxies. Figure taken from Odewahn et al. (1996, ApJ, 472, L13).|