Faint Galaxies in Deep Advanced Camera for Surveys Observations

We present the analysis of the faint galaxy population in the Advanced Camerafor Surveys (ACS) Early Release Observation fields VV 29 (UGC 10214) and NGC4676. Here we attempt to thoroughly consider all aspects relevant for faintgalaxy counting and photometry, developing methods which are based on publicsoftware and that are easily reproducible by other astronomers. Usingsimulations we determine the best SExtractor parameters for the detection offaint galaxies in deep HST observations, paying special attention to the issueof deblending, which significantly affects the normalization and shape of thenumber count distribution. We confirm, as claimed by Bernstein, Freedman andMadore (2002), that Kron-like magnitudes, such as the ones generated bySExtractor, can miss more than half of the light of faint galaxies, whatdramatically affects the slope of the number counts. We present catalogs forthe VV 29 and NGC 4676 fields with photometry in the g,V and I bands. We alsoshow that combining the bayesian software BPZ with superb ACS data and newspectral templates enables us to estimate reliable photometric redshifts for asignificant fraction of galaxies with as few as three filters. After correctingfor selection effects, we measure slopes of 0.32+- 0.01 for 22 < g < 28,0.34+-0.01 for 22< V <27.5 and 0.33+-0.01 for 22 < I < 27. The counts do notflatten (except perhaps in the g-band), up to the depth of our observations. Wefind that the faint counts m_{AB}> 25.5 can be well approximated in all ourfilters by a passive luminosity evolution model based on the COMBO-17luminosity function (\alpha=-1.5), with a strong merging rate following theprescription of Glazebrook et al. (1994), \phi^*\propto (1+Qz), with Q=4.