[abridged] New near-infrared surveys, using the HST, offer an unprecedented opportunity to study rest-frame optical galaxy morphologies at z>1 and to calibrate automated morphological parameters that will play a key role in classifying future massive
datasets like EUCLID or LSST. We study automated parameters (e.g. CAS, Gini, M20) of massive galaxies at 1<z<3, measure their dependence on wavelength and evolution with redshift and quantify the reliability of these parameters in discriminating between visually-determined morphologies, using machine learning algorithms. We find that the relative trends between morphological types observed in the low-redshift literature are preserved at z>1: bulge-dominated systems have systematically higher concentration and Gini coefficients and are less asymmetric and rounder than disk-dominated galaxies. However, at z>1, galaxies are, on average, 50% more asymmetric and have Gini and M20 values that are 10% higher and 20% lower respectively. In bulge-dominated galaxies, morphological parameters derived from the rest-frame UV and optical wavelengths are well correlated; however late-type galaxies exhibit higher asymmetry and clumpiness when measured in the rest-frame UV. We find that broad morphological classes (e.g. bulge vs. disk dominated) can be distinguished using parameters with high (80%) purity and completeness of 80%. In a similar vein, irregular disks and mergers can also be distinguished from bulges and regular disks with a contamination lower than 20%. However, mergers cannot be differentiated from the irregular morphological class using these parameters, due to increasingly asymmetry of non-interacting late-type galaxies at z>1. Our automated procedure is applied to the CANDELS GOODS-S field and compared with the visual classification recently released on the same area getting similar results.
[abridged] The mass-size relation of early-type galaxies (ETGs) has been largely studied in the last years to probe the mass assembly of the most massive objects in the Universe. In this paper, we focus on the mass-size relation of quiescent massive
ETGs (Mstar/Msol > 3*10^10) living in massive clusters (M200 ~ 10^14 Mstar) at 0.8< z <1.5, as compared to those living in the field at the same epoch. Our sample contains ~ 400 ETGs in clusters and the same number in the field. Therefore, our sample is approximately an order of magnitude larger than previous studies in the same redshift range for galaxy clusters. We find that ETGs living in clusters are between ~30-50% larger than galaxies with the same stellar mass residing in the field. We parametrize the size using the mass-normalized size, gamma=Re/Mstar^0.57. The gamma distributions in both environments peak at the same position but the distributions in clusters are more skewed towards larger sizes. Since this size difference is not observed in the local Universe, the size evolution at fixed stellar mass from z~1.5 of cluster galaxies is less steep ((1+z)-0.53pm0.04) than the evolution of field galaxies ((1+z)-0.92pm0.04). The size difference seems to be essentially driven by the galaxies residing in the clusters cores (R<0.5*R200). If part of the size evolution is due to mergers, the difference we see between cluster and field galaxies could be due to higher merger rates in clusters at higher redshift, probably during the formation phase of the clusters when velocity dispersions are lower. We cannot exclude however that the difference is driven by newly quenched galaxies which are quenched more efficiently in clusters. The implications of these results for the hierarchical growth of ETGs will be discussed in a companion paper.
We quantify the effects of emph{morphological k-correction} at $1<z<2$ by comparing morphologies measured in the K and I-bands in the COSMOS area. Ks-band data have indeed the advantage of probing old stellar populations for $z<2$, enabling a determi
nation of galaxy morphological types unaffected by recent star formation. In paper I we presented a new non-parametric method to quantify morphologies of galaxies on seeing limited images based on support vector machines. Here we use this method to classify $sim$$50 000$ $Ks$ selected galaxies in the COSMOS area observed with WIRCam at CFHT. The obtained classification is used to investigate the redshift distributions and number counts per morphological type up to $zsim2$ and to compare to the results obtained with HST/ACS in the I-band on the same objects from other works. We associate to every galaxy with $Ks<21.5$ and $z<2$ a probability between 0 and 1 of being late-type or early-type. The classification is found to be reliable up to $zsim2$. The mean probability is $psim0.8$. It decreases with redshift and with size, especially for the early-type population but remains above $psim0.7$. The classification is globally in good agreement with the one obtained using HST/ACS for $z<1$. Above $zsim1$, the I-band classification tends to find less early-type galaxies than the Ks-band one by a factor $sim$1.5 which might be a consequence of morphological k-correction effects. We argue therefore that studies based on I-band HST/ACS classifications at $z>1$ could be underestimating the elliptical population. [abridged]
