The K20 Redshift Survey

People involved at OAB: Lanzoni, Mignoli, Pozzetti, Zamorani.

The Bologna Observatory participates, with the observatories of Arcetri, ESO and Rome, in an ESO VLT Large Program (PI Cimatti, Arcetri Obs.), which started in 1999. This program (called K20 ) aims at deriving the redshift distribution of a galaxy sample complete at $K<20$ in order to obtain stringent clues on the formation and evolution of present-day massive galaxies (Cimatti et al. 2002). The sample consists of about 500 galaxies selected from a sub-area of the Chandra Deep Field (CDF) and from a field around the quasar 0055$-$2659 ($z=3.7$). Seventeen VLT nights have been allocated to this project over a period of two years. The survey makes use of both optical (FORS1/FORS2) and near-IR (ISAAC) spectroscopy. The observations were completed in 2000. In addition, ultradeep spectroscopy was obtained in November 2002 with FORS2 to derive information on previously unidentified EROs and $z>1.7$ galaxies. We have provided spectra of 92% of the galaxies down to a completeness level of about $K<20$. Moreover, the K20 survey is triggering several follow up studies in which we are involved. High resolution spectroscopy (PI Cimatti) was recently obtained with FORS2 to study the kinematics of $z>1$ galaxies in order to estimate their dynamical masses. The CDFS sub-area of the K20 survey is also a target of the HST+ACS GOODS Treasury Programme (PI Giavalisco) and of the SIRTF GOODS (PI Dickinson) Legacy Programme.

The scientific analysis is still in progress and the main results obtained up to now can be summarized as follows (details can be found also in the previous reports):

• A major result was the discovery of a significant high-redshift tail (32% and $\sim9$% of galaxies at $z>1$ and $z>1.5$ respectively), not consistent with current versions of hierarchical models of galaxy formation (Cimatti, Pozzetti et al. 2002).
• We have obtained, for the first time, a spectroscopic sample of EROs. We found that the two classes of old and dusty star-forming galaxies in the ERO population, based on both spectroscopic (Cimatti et al. 2002) and HST morphological properties obtained from GOODS images (Cimatti et al. 2003), are about equally populated. Conversely they show very different stellar populations, as well as different clustering and correlation length.
• We have carried out the analysis of the near-IR ($J$ and $Ks$ band) Luminosity Functions down to redshift $\sim 1.9$ (Pozzetti et al. 2003). The data are consistent with a mild luminosity evolution, in particular, red and early-type galaxies dominate the bright-end of the LF, and their number density shows at most a small decrease ($<30\%$) up to $z\simeq 1$, thus suggesting that massive elliptical galaxies were already in place at $z\simeq 1$ and they should have formed their stars and assembled their mass at higher redshift. Current versions of hierarchical models (Cole et al. 2000; Kauffmann et al. 1999) significantly overpredict the density of low luminosity galaxies at $z\le1$ and underpredict the density of luminous galaxies at $z\ge1$, whereas passive evolution models are more consistent with the data up to $z\sim 1.5$.

Figure 9: Cosmological Stellar Mass Density for objects of $M_*\simeq 10^{11}M_\odot$ in the K20 survey (Fontana et al. 2004, in press), compared with the theoretical models, divided according to the relevant IMF: upper panel: Salpeter (1955), central panel: Gould et al. (1996), lower panel: Kennicutt (1983). Filled squares and filled triangles represent different Mass estimates. The empty circle and square at $z=0.1$ represent the local values of Cole et al. 2001. The models plotted are: Menci et al. 2002 (M02), Menci et al. 2004 (M04) and Pozzetti et al. 2003 (PLE) in the upper panel; Nagamine et al. 2001 (N01) in the central panel, and Cole et al. 2000 (C00), Somerville et al. 2004a (S04a and S04b) and Granato et al. 2004 (G04) in the lower panel. The relative value with respect to the local value of Cole et al. 2001 is shown in the right-hand y-axis.

This year we have completed, in collaboration with the Rome Observatory, the study of the Galaxy Stellar Mass Function and its evolution to $z\simeq2$ (Fontana, Pozzetti et al. 2004, in press). We found that the typical mass-to-light ratio of massive early type galaxies is larger than that of less massive ones, suggesting that their stellar population formed at higher $z$. The final K20 galaxy sample spans a range of stellar masses from $10^9M_\odot$ to $10^{12}M_\odot$: massive galaxies ( $\geq10^{11}M_\odot$) are common at $0.5<z<1$, and are detected also up to $z\simeq2$. The Galaxy Stellar Mass Function shows only a mild evolution (i.e. by 20-30%) up to $z\simeq 1$. At $z>1$, the evolution in the GSMF appears to be much faster: at $z\simeq2$, about 35% of the present day stellar mass in objects with $M_*\simeq 10^{11}M_\odot$ appears to have assembled. Moreover we found that at $z<0.7$, all galaxies with $M>10^{11}M_\odot$ are early type, while at higher $z$ a population of massive star-forming galaxies progressively appears. The predictions from hierarchical models range from severe underestimates to slight overestimates of the observed mass density at $\leq 2$ (see Fig. 9).

From ultradeep spectroscopy we have furthermore investigated the population of $z>2$ galaxies (Daddi et al. 2004). We found a new population of massive galaxies, with HST irregular morphology, and with multi-band photometry and average spectrum consistent with high star formation rate absorbed by a large amount of dust, young stellar populations and high metallicity (De Mello et al. 2004, in press). This population is consistent with being the progenitor of massive local ellipticals.

At present, in Bologna, we are completing the analysis of the spectroscopic K20 sample, both studying the properties of the single object spectra, and constructing average templates for different spectral classes and/or different redshift bins, with the aim of characterizing a possible spectral evolution. In addition, we are collaborating with Arcetri in the study of the star formation history using different SFR indicator ([OII], H$\alpha$, UV flux) up to $z\simeq1.5$.

Finally, in collaboration with the Arcetri Observatory (Alighieri, Cimatti, Vernet) we are involved in the analysis of the high resolution spectra of a subsample of early and late type galaxies at $z\simeq 1$, recently obtained. In particular we are carrying out the spectral analysis of early type galaxies using high resolution model templates (Bruzual & Charlot 2003) to derive new constraints on their stellar population and on the epoch of formation of massive elliptical galaxies. Moreover, we are comparing the dynamical properties, derived from high-resolution spectra of both elliptical and spiral galaxies, with those obtained by the hierarchical model GalICs (Hatton et al. 2003). From observations we found well defined scaling relations, as the Tully-Fisher and the Fundamental Plane. Preliminary results show that model galaxies define scaling relations similar to those observed, even if the zero point and scatter appear not to be exactly the same. A more detailed comparison and the interpretation of results in terms of the capability of the hierarchical scenario in describing galaxy formation is still in progress.