Outline of recent research activity

   Signals out of the noise: Properties of high redshift galaxies from radio stacking

How and when galaxies build up their stellar mass is still a major question in observational cosmology. While a general consensus has been reached in the last years on the evolution of the galaxy stellar mass function (e.g. [8,9,19,3.c] and references therein), the redshift evolution of the star formation rate (SFR) as a function of stellar mass still remains unclear. UV-derived SFRs suffer from a major unsolved issue: the amount of dust attenuation affecting the UV light in the inter-stellar medium [14,5.c]. Often the dust attenuation factor is the result of a highly degenerate fit to the multiwavelength photometry available or, when the photometric coverage is not sufficient, a median attenuation factor is usually applied to the whole galaxy sample.

An alternative estimate of the star formation, not biased by the galaxy's dust content, is provided by its radio continuum emission, due to processes dominated by young massive stars. By means of the radio-FIR correlation (e.g. Condon 1992, Kennicutt 1998, Yun et al. 2001) it is possible to estimate the total star formation rate in a galaxy from its radio luminosity. Thanks to their arcsecond resolution and relatively wide field of view, radio interferometric observations offer several advantages over present-day FIR facilities which are limited by their $\sim10''$ resolution, small field of view and relatively poor sensitivity. For this very reason radio continuum observations turn out to be an excellent tool for tracing the dust-unobscured star formation in the high redshift Universe.

However at $z>1$, even in the deepest present-day surveys (e.g. [1.s,2.s]), radio detections are likely to include a substantial population of AGNs, although extreme ULIRG/SMG-like starbursting galaxies do exist. Therefore the best way to explore, with existing radio facilities, the dust-unbiased SFRs of normal galaxy populations is to use a stacking analysis of the radio data, which allows the investigation of large samples of galaxies drawn from optical-NIR surveys, which are individually undetected in the radio. In this context the Cosmic Evolution Survey (COSMOS, Scoville et al. 2007), with its state-of-the-art multiwavelength coverage all the way from X-rays to radio of a 2 square degrees field, provides an ideal opportunity to build large high redshift galaxy samples with well characterized spectral properties.

I have taken advantage of the COSMOS database to select a large sample of $1<z<3$ star-forming galaxies, and derive dust-unbiased SFRs by stacking the 1.4 GHz radio data after removing the AGN contaminants [18]. I found that a single "universal" dust-attenuation correction cannot be applied to the whole sample under study. For instance, the generic factor of 5 often used to correct the UV light of Lyman-break galaxies (LBG) is appropriate only for objects with $M_*\sim 3\times 10^{10}\; M_\odot$ -which incidentally is very close to the median stellar mass of LBG galaxies - but would grossly underestimate the correction for more massive galaxies and, likely, overestimate the one for lower mass galaxies. I also found that the SFR of these $z\sim2$ star-forming galaxies increases almost linearly with the galaxy stellar mass, therefore their specific star formation rate (SSFR=SFR/$M_{*}$) is about the same at all masses: star forming galaxies at z$\approx$2 have a nearly exponential growth in mass with the same evolutionary timescales, $\tau$, at all stellar masses. Clearly, individual galaxies would enormously overgrow in mass if these SFRs were maintained for long time. One may speculate (see Renzini 2009) that this does not happen because many galaxies turn passive, and do so in a downsized fashion, because massive galaxies are the first to reach unsustainable SFR levels and to hexaust the available gas reservoirs: these galaxies are just at their dawn of downsizing.

   The faintest ever radio populations: galaxies, massive black holes and their coevolution

The nature of the $\mu$Jy radio sources is still an unsolved topic. The naive expectation that all sources below a few hundreds $\mu$Jy had radio emission mostly driven by star formation activity has been seriously reconsidered. The departure of radio source number counts from the power law slope of the bright radio sources is definitely due to the arising of a different population, but this latter is not, at least at the flux levels so far explored, a purely star forming population. In fact, a significant fraction of the sub-mJy radio sources have their emission dominated by Low Luminosity AGN. Also, the restframe optical color distribution of the whole faint radio population peaks on the "green valley", i.e. the relatively empty region in a magnitude vs. color diagram separating "red sequence" passive galaxies from "blue cloud" star forming systems. In the present understanding of galaxy evolution, the "green valley" phase is thought to be a relatively fast transition stage when the star formation is truncated/switched off and galaxies move from the blue cloud to the red sequence. Especially in the last few years, a particular importance in this process has been given to the galaxy central massive black hole activity (the so-called AGN feedback). The fact that a relevant fraction of the faint radio sources are found in this particular, and potentially short, phase of galaxy evolution is thus particularly intriguing.

Over the last years, I worked in collaboration with V. Strazzullo and F. Owen on the deepest VLA 1.4 GHz continuum survey carried out to date, the SWIRE Deep Field (SDF) in the Lockman Hole (Owen & Morrison 2008). The unprecedented depth (rms $\approx$ 2.7$\mu$Jy at the image center) of this survey allows us to see distant, faint radio populations which are unseen in other surveys. From a wide multi-wavelength dataset (11 passbands from GALEX NUV to IRAC 4.5$\mu$m) we assembled a multicolor catalog of sources in the field. Thanks to the arcsec resolution of the VLA image, we were able to match the vast majority (97%) of the radio sources with a optical/NIR counterpart. We then took advantage of the full multi-wavelength information and used the counterpart's observed SED to estimate accurate photometric redshifts for the radio sources, and to derive stellar masses and stellar population properties through comparison with Bruzual & Charlot (2003) spectral population synthesis models, as well as with a library of galaxy semi-empirical templates. Based on their rest-frame optical/NIR SED, we classified the radio sources in three sub-samples of quiescent (red sequence), starforming (blue cloud), and intermediate (green valley) galaxies. As main results of our investigation [1.s, 2.s], we have found that: 1) the relative contributions of low luminosity AGN and starforming galaxies to the $\mu$Jy population depend on the flux limit of the sample; 2) the fraction of starforming objects reaches 50% only at the faintest flux levels (14$\mu$Jy$<S_{1.4GHz}<$24$\mu$Jy); 3) at all flux levels a significant population of intermediate galaxies, with colors compatible with green-valley objects, is observed; 4) according to our analysis the radio emission in these intermediate objects is generally best explained by a relevant contribution from nuclear activity plus only a residual ongoing star formation.

During last summer, as part of an REU student program, we have started the analysis of the Chandra X-ray survey on the SDF. We adopted a stacking approach as most of the radio sources are too faint to be detected in the 70ksec X-ray image. We stacked X-ray counts for our subsamples of quiescent, starforming, and intermediate sources. We were able to obtain enough signal to derive hardness ratios and restframe luminosities for the samples at redshift $z\sim0.5$ . These point toward an AGN-dominated nature for both the red and green populations, while blue sources appear close to star formation-dominated systems. While this is still work in progress (Pannella et al. in preparation), the first results have nicely confirmed that deep radio continuum surveys are unique tools to spot samples of galaxies undergoing a very peculiar, but fundamental, phase of galaxy evolution, namely the moving from the blue star forming population to the passive red sequence. Follow up studies of this population will allow in the next future a better understanding of black hole feedback and its role on the host galaxy evolution.
 

     Mass growth and morphological evolution of galaxies in different environments

Galaxy formation and evolution have been a very actively debated topic of observational cosmology in the last years. It is only in the very recent years that models have been able to fully reproduce the observed galaxy stellar mass function up to high redshifts, thus apparently reconciling the theoretical bottom-up assembly of dark matter halos to the claimed top-down downsized assembly of galaxies (e.g. Cimatti et al. 2006). Still many hidden details are missed or not completely understood. Questions like: How much merging happens? Is it wet, dry or moist? When and how the black hole feedback is impacting the galaxy life? are still far to be answered in detail by models and verified by observations.

An effective way to constrain models is to understand where the stars were located, i.e. in which kind of galaxies, at different look-back times. This allows us to directly probe when/how galaxies assembled their stars and how their morphology (i.e., at least with some approximation, their dynamical status) evolves. Since merging is driving both mass assembly and dynamical evolution in a hierarchical scenario, these studies offer a direct probe of its role in galaxy evolution.

Starting with my PhD research, I studied the evolution of stellar mass content for galaxies of different morphological types and in different environments, using multiwavelength surveys like GOODS-S, FORS DEEP FIELD and COSMOS, for which high resolution, deep HST-ACS imaging allows an accurate description of the galaxy morphology at least up to redshift 1 [9,19]. In agreement with some other recent studies on the same subject, we have confirmed an evolution of the morphological mix, with the relevance of disk-dominated galaxies in the total stellar mass budget increasing with redshift. Such redshift evolution is also coupled to strong environmental dependence, since the morphological mix is also a function of the local comoving density: at all redshifts we probed, we found evidence of a morphology-density relation. Finally, while the stellar mass function of disk-dominated galaxies is consistent with being constant up to z$\approx$1.2, the stellar mass function of bulge-dominated systems shows a decline in normalization over the same redshift range by at least a factor of two. These observations point toward a scenario in which massive objects almost double their mass from redshift 1 to 0, at the same time evolving toward a bulge-dominated structure. Since the SFR of massive galaxies at $z\sim1$ is not high enough to double their mass over 7 Gyrs, this study suggests that merging and accretion events must play a key role in the mass pouring from disks to bulges, increasing the stellar mass, moving the galaxy to a bulge-dominated structure, and quenching the star formation possibly after that a merging-triggered burst exhausts the available gas. According to our data, this process has happened first, or more quickly, in the highest density regions of the Universe.

The newly available WFC3 on board HST will allow to push morphological evolution studies up to redshift 2, by surveying the optical restframe light of galaxies, and hence to look into the initial stages of the build up of the Hubble fork.