The K20 & GMASS Redshift Surveys

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

In collaboration with the observatories of Arcetri, ESO and Rome, the Bologna Observatory has participated to the ``K20 Survey''³ an ESO VLT Large Program (PI Cimatti, INAF-Arcetri Observatory) which started in 1999. This program aimed at deriving the redshift distribution of a galaxy sample complete at $K<20$, in order to obtain clues on the formation and evolution history of present-day massive galaxies (Cimatti et al. 2002). The sample consists of 525 galaxies selected from a sub-area of the Chandra Deep Field South (CDFS) and from a field around the quasar 0055-2659 ($z=3.7$). Seventeen VLT nights (FORS1, FORS2, ISAAC) have been allocated to this project over a period of two years (1999-2000). In addition, ultradeep FORS2 spectroscopy was obtained in November 2002 to derive information on previously unidentified EROs and $z>1.7$ galaxies. We have measured spectra for 92% of the galaxies down to a completeness level of $K \simeq 20$.

The analysis of the main scientific objectives is now completed and the results obtained have been presented in different papers. The survey and the sample are described in detail in Cimatti et al. (2002b). The main results can be summarized as follows: (1) discovery of a significant high-redshift tail of galaxies not consistent with most versions of semi-analytical hierarchical models of galaxy formation (Cimatti et al. 2002c). (2) The two classes of EROs with very different stellar populations, with old or dusty star-forming respectively, are about equally represented and have significantly different clustering properties (Cimatti et al. 2002a, Daddi et al. 2002). (3) The near-IR ($J$ and $Ks$ band) Luminosity Functions (Pozzetti et al. 2003) and the Galaxy Stellar Mass Function (GSMF, described in Fontana et al. 2004) are consistent with a mild evolution up to $z\simeq1.5$. In particular, massive and luminous elliptical galaxies were already in place at $z\simeq 1$ and they should have formed their stars and assembled their mass at higher redshift. At $z>1$, the evolution of the GSMF appears to be much faster. Moreover, essentially all galaxies with $M>10^{11}M_\odot$ are early type at $z<0.7$, 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 $\leq 2$. (4) We found a new population of massive galaxies at $z>2$, characterized by HST irregular morphology, and with multi-band photometry and average spectrum consistent with high star formation rates, large amounts of absorbing dust, young stellar populations and high metallicity (Daddi et al. 2004, De Mello et al. 2004). This population is consistent with being the progenitor of massive local ellipticals. (5) The spectroscopic identification of four galaxies with $1.6< z <1.9$, which have very red colors ($R-K_s>6$) and rest-frame mid-UV spectra and average spectrum with shapes and continuum breaks consistent with being dominated by old stars, and reproduced by a synthetic stellar populations of about $1-2$ Gyr. Also the ACS high-resolution images is typical of elliptical/early-type galaxies. This result has been presented in a paper published in Nature (Cimatti et al. 2004).

During the year 2005, the analysis of the spectroscopic K20 sample has been completed in Bologna, both studying the properties of 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 (Mignoli et al. 2005). The classification of galaxy spectra has been performed according to a parametric recipe that uses the equivalent widths of the two main emission lines ([OII] and H$\alpha$+[N II]) and two continuum indices (the 4000Å break index, and a near-UV colour index, $C(28-39)$).

We found that the average spectra of early-type galaxies at different redshift present only small but systematic differences, consistently with the ageing of the stellar population. Conversely, the star-forming emission line galaxies show a clear ``blueing'' of the continuum with increasing redshift. We compared the emission line galaxy composites with simple galaxy evolution models, finding that the galaxies at high redshift are more active (i.e. have a higher ratio between the current and the past average SFR) than those in the lower redshift bins. The observed values of the [OII] equivalent width also suggest a somewhat lower average metallicity for the low-$z$ objects, explaining the approximate constancy of the EW throughout the redshift intervals. This may be at least partly due to the well-known metallicity-luminosity relationship for star-forming galaxies.

Finally, we have completed the analysis of high resolution spectra, and HST or VLT imaging of a subsample of 15 early-type galaxies at $z\simeq 1$ (di Serego Alighieri et al. 2005). These objects are found to define a Fundamental Plane relation, that keeps a very small scatter and shows an offset with respect to the local one. Moreover, the FP at $z \sim 1$ also seems to have a different slope with respect to the local one, thus implying that the detected evolution with redshift of the galaxy mass-to-light ratio ($M/L$) depends on mass, being faster for less massive systems. A fairly good agreement is found, given the uncertainties, between stellar and dynamical masses, although a discrepancy (larger stellar than dynamical masses) seems to be shown by low mass galaxies. A comparison between the observed scaling relations (FP, Faber-Jackson and Kormendy relations) and those predicted by the hierarchical model GalICs (Hatton et al. 2003) shows a good qualitative agreement, but most of the model galaxies have too blue intrinsic colours and a higher dust extinction with respect to the data. The analysis of the stellar population of these early-type subsample and a similar analysis for a subsample of late-type galaxies and the Tully-Fisher relation at $z \sim 1$ is also in progress.

The same people who participate in the K20 survey are involved in a logically consequent and even more ambitious project: the ``Galaxy Mass Assembly ultradeep Spectroscopic Survey (GMASS)''. This project aims to investigate the nature and the evolutionary status of galaxies in the redshift range $1<z<3$, where recent results suggest that most of the galaxy mass was built.

The $z \sim 1-3$ is a problematic redshift range because galaxies become rapidly very faint and this difficulty is exacerbated in the so called ``redshift desert'' (i.e. $z \sim 1.3-2.3$) where spectroscopic redshifts are hard to derive, especially for red galaxies. Photometric redshifts provide little clues on the nature and evolutionary status of galaxies, and no constraints on their 3D clustering. The GMASS project is an ESO Large Program (# 173.A-0687, P.I. A.Cimatti) which was awarded observing time for 145 hours with FORS2@VLT (30 hrs. in P73 and 115 hrs. in P74). The project approach was to perform ultra-deep ESO VLT+ FORS2 multi-object spectroscopy with very long integration times (15-40 hours per mask) of infrared-selected galaxies ($m_{4.5}<23$) having high-quality photometric redshifts $z_{phot}>1.4$. The spectroscopy has been obtained either in the blue or in the red, depending on the SEDs of the target galaxies, their $z_{phot}$ and their location in diagnostic colour-colour diagrams. The GMASS target field (about 50 square arcminutes) is located in the GOODS-CDFS area and includes most of the Hubble Ultra Deep Field and a fraction of the K20 field. This field includes several galaxies with already known spectroscopic redshifts, and a large collection of multi-wavelength data: VLT ISAAC $JHK_s$ photometry, HST+ACS/NICMOS imaging, Spitzer $3-24\mu$m photometry, GALEX UV deep imaging, and radio and X-ray data. The observations were completed in 2005 and provided spectra of 208 objects. The spectroscopic data reduction and analyses have been carried out in Bologna and Arcetri. The scheme for the blue and red spectroscopic data reduction has been intensively tested and developed in Arcetri. Redshift measurements have been completed: the results for the blue masks are extremely good, with a high spectroscopic redshift identification efficiency ($\sim 90$%) and most targets with $z_{\rm spec}>1.4$, as expected from the $z_{\rm phot}>1.4$ cut. We have a few objects also at . In the red the spectroscopic success rate is lower. The final analysis has provided a redshift for most (182/208) of the observed objects. Of the 182 measured redshifts 146 are for objects in the selected sample while 36 are for objects observed as fillers. The selection has worked reasonably well, with only 12% of the main sample having $z_{\rm spec}<1.4$. The most striking feature of the GMASS z distribution is a peak at $z \simeq 1.6$ (33 objects from GMASS + 7 from literature). The determination of optimized photometric redshifts using the final photometric and spectroscopic sample and the analysis of the SEDs to derive the photometric stellar masses is in progress in Bologna, using different codes of stellar population models (i.e. Bruzual & Charlot 2003 and Maraston 2005). Further analyses, such as the metallicity estimates from UV absorption and morphological analysis, are in progress in Arcetri and Padova.