# Posts Tagged spectroscopic redshifts

## Recent Postings from spectroscopic redshifts

### Photometric redshifts and clustering of emission line galaxies selected jointly by DES and eBOSS

We present the results of the first test plates of the extended Baryon Oscillation Spectroscopic Survey. This paper focuses on the emission line galaxies (ELG) population targetted from the Dark Energy Survey (DES) photometry. We analyse the success rate, efficiency, redshift distribution, and clustering properties of the targets. From the 9000 spectroscopic redshifts targetted, 4600 have been selected from the DES photometry. The total success rate for redshifts between 0.6 and 1.2 is 71\% and 68\% respectively for a bright and faint, on average more distant, samples including redshifts measured from a single strong emission line. We find a mean redshift of 0.8 and 0.87, with 15 and 13\% of unknown redshifts respectively for the bright and faint samples. In the redshift range 0.6<z<1.2, for the most secure spectroscopic redshifts, the mean redshift for the bright and faint sample is 0.85 and 0.9 respectively. Star contamination is lower than 2\%. We measure a galaxy bias averaged on scales of 1 and 10~Mpc/h of 1.72 \pm 0.1 for the bright sample and of 1.78 \pm 0.12 for the faint sample. The error on the galaxy bias have been obtained propagating the errors in the correlation function to the fitted parameters. This redshift evolution for the galaxy bias is in agreement with theoretical expectations for a galaxy population with MB-5\log h < -21.0. We note that biasing is derived from the galaxy clustering relative to a model for the mass fluctuations. We investigate the quality of the DES photometric redshifts and find that the outlier fraction can be reduced using a comparison between template fitting and neural network, or using a random forest algorithm.

### Mapping the Galaxy Color-Redshift Relation: Optimal Photometric Redshift Calibration Strategies for Cosmology Surveys

Calibrating the photometric redshifts of >10^9 galaxies for upcoming weak lensing cosmology experiments is a major challenge for the astrophysics community. The path to obtaining the required spectroscopic redshifts for training and calibration is daunting, given the anticipated depths of the surveys and the difficulty in obtaining secure redshifts for some faint galaxy populations. Here we present an analysis of the problem based on the self-organizing map, a method of mapping the distribution of data in a high-dimensional space and projecting it onto a lower-dimensional representation. We apply this method to existing photometric data from the COSMOS survey selected to approximate the anticipated Euclid weak lensing sample, enabling us to robustly map the empirical distribution of galaxies in the multidimensional color space defined by the expected Euclid filters. Mapping this multicolor distribution lets us determine where - in galaxy color space - redshifts from current spectroscopic surveys exist and where they are systematically missing. Crucially, the method lets us determine whether a spectroscopic training sample is representative of the full photometric space occupied by the galaxies in a survey. We explore optimal sampling techniques and estimate the additional spectroscopy needed to map out the color-redshift relation, finding that sampling the galaxy distribution in color space in a systematic way can efficiently meet the calibration requirements. While the analysis presented here focuses on the Euclid survey, similar analysis can be applied to other surveys facing the same calibration challenge, such as DES, LSST, and WFIRST.

### Stellar masses and star formation rates of lensed dusty star-forming galaxies from the SPT survey

To understand cosmic mass assembly in the Universe at early epochs, we primarily rely on measurements of stellar mass and star formation rate of distant galaxies. In this paper, we present stellar masses and star formation rates of six high-redshift ($2.8\leq z \leq 5.7$) dusty, star-forming galaxies (DSFGs) that are strongly gravitationally lensed by foreground galaxies. These sources were first discovered by the South Pole Telescope (SPT) at millimeter wavelengths and all have spectroscopic redshifts and robust lens models derived from ALMA observations. We have conducted follow-up observations, obtaining multi-wavelength imaging data, using {\it HST}, {\it Spitzer}, {\it Herschel} and the Atacama Pathfinder EXperiment (APEX). We use the high-resolution {\it HST}/WFC3 images to disentangle the background source from the foreground lens in {\it Spitzer}/IRAC data. The detections and upper limits provide important constraints on the spectral energy distributions (SEDs) for these DSFGs, yielding stellar masses, IR luminosities, and star formation rates (SFRs). The SED fits of six SPT sources show that the intrinsic stellar masses span a range more than one order of magnitude with a median value $\sim$ 5 $\times 10^{10}M_{\Sun}$. The intrinsic IR luminosities range from 4$\times 10^{12}L_{\Sun}$ to 4$\times 10^{13}L_{\Sun}$. They all have prodigious intrinsic star formation rates of 510 to 4800 $M_{\Sun} {\rm yr}^{-1}$. Compared to the star-forming main sequence (MS), these six DSFGs have specific SFRs that all lie above the MS, including two galaxies that are a factor of 10 higher than the MS. Our results suggest that we are witnessing the ongoing strong starburst events which may be driven by major mergers.

### Shadow of a Colossus: A z=2.45 Galaxy Protocluster Detected in 3D Ly-a Forest Tomographic Mapping of the COSMOS Field

Using moderate-resolution optical spectra from 58 background Lyman-break galaxies and quasars at $z\sim 2.3-3$ within a $11.5'\times13.5'$ area of the COSMOS field ($\sim 1200\,\mathrm{deg}^2$ projected area density or $\sim 2.4\,h^{-1}\,\mathrm{Mpc}$ mean transverse separation), we reconstruct a 3D tomographic map of the foreground Ly$\alpha$ forest absorption at $2.2<z<2.5$ with an effective smoothing scale of $\sigma_{3d}\approx3.5\,h^{-1}\,\mathrm{Mpc}$ comoving. Comparing with 61 coeval galaxies with spectroscopic redshifts in the same volume, we find that the galaxy positions are clearly biased towards regions with enhanced IGM absorption in the tomographic map. We find an extended IGM overdensity with deep absorption troughs at $z=2.45$ associated with a recently-discovered galaxy protocluster at the same redshift. Based on simulations matched to our data, we estimate the enclosed dark matter mass within this IGM overdensity to be $M_{\rm dm} (z=2.45) = (9\pm4)\times 10^{13}\,h^{-1}\,\mathrm{M_\odot}$, and argue based on this mass and absorption strength that it will form at least one $z\sim0$ galaxy cluster with $M(z=0) = (3\pm 2) \times 10^{14}\,h^{-1}\mathrm{M_\odot}$, although its elongated nature suggests that it will likely collapse into two separate clusters. We also point out a compact overdensity of six MOSDEF galaxies at $z=2.30$ within a $r\sim 1\,h^{-1}\,\mathrm{Mpc}$ radius and $\Delta z\sim 0.006$, which does not appear to have a large associated IGM overdensity. These results demonstrate the potential of Ly$\alpha$ forest tomography on larger volumes to study galaxy properties as a function of environment, as well as revealing the large-scale IGM overdensities associated with protoclusters and other features of large-scale structure.

### The Lyman Continuum escape fraction of galaxies at z=3.3 in the VUDS-LBC/COSMOS field

The Lyman continuum (LyC) flux escaping from high-z galaxies into the IGM is a fundamental quantity to understand the physical processes involved in the reionization epoch. We have investigated a sample of star-forming galaxies at z~3.3 in order to search for possible detections of LyC photons escaping from galaxy halos. UV deep imaging in the COSMOS field obtained with the prime focus camera LBC at the LBT telescope was used together with a catalog of spectroscopic redshifts obtained by the VIMOS Ultra Deep Survey (VUDS) to build a sample of 45 galaxies at z~3.3 with L>0.5L*. We obtained deep LBC images of galaxies with spectroscopic redshifts in the interval 3.27<z<3.40 both in the R and deep U bands. A sub-sample of 10 galaxies apparently shows escape fractions>28% but a detailed analysis of their properties reveals that, with the exception of two marginal detections (S/N~2) in the U band, all the other 8 galaxies are most likely contaminated by the UV flux of low-z interlopers located close to the high-z targets. The average escape fraction derived from the stacking of the cleaned sample was constrained to fesc_rel<2%. The implied HI photo-ionization rate is a factor two lower than that needed to keep the IGM ionized at z~3, as observed in the Lyman forest of high-z QSO spectra or by the proximity effect. These results support a scenario where high redshift, relatively bright (L>0.5L*) star-forming galaxies alone are unable to sustain the level of ionization observed in the cosmic IGM at z~3. Star-forming galaxies at higher redshift and at fainter luminosities (L<<L*) can be the major contributors to the reionization of the Universe only if their physical properties are subject to rapid changes from z~3 to z~6-10. Alternatively, ionizing sources could be discovered looking for fainter sources among the AGN population at high-z.

### The faint radio source population at 15.7 GHz - II. Multi-wavelength properties [Replacement]

A complete, flux density limited sample of 96 faint ($> 0.5$ mJy) radio sources is selected from the 10C survey at 15.7 GHz in the Lockman Hole. We have matched this sample to a range of multi-wavelength catalogues, including SERVS, SWIRE, UKIDSS and optical data; multi-wavelength counterparts are found for 80 of the 96 sources and spectroscopic redshifts are available for 24 sources. Photometric reshifts are estimated for the sources with multi-wavelength data available; the median redshift of the sample is 0.91 with an interquartile range of 0.84. Radio-to-optical ratios show that at least 94 per cent of the sample are radio loud, indicating that the 10C sample is dominated by radio galaxies. This is in contrast to samples selected at lower frequencies, where radio-quiet AGN and starforming galaxies are present in significant numbers at these flux density levels. All six radio-quiet sources have rising radio spectra, suggesting that they are dominated by AGN emission. These results confirm the conclusions of Paper I that the faint, flat-spectrum sources which are found to dominate the 10C sample below $\sim 1$ mJy are the cores of radio galaxies. The properties of the 10C sample are compared to the SKADS Simulated Skies; a population of low-redshift starforming galaxies predicted by the simulation is not found in the observed sample.

### Multicolor Photometry of the Merging Galaxy Cluster A2319: Dynamics and Star Formation Properties

Asymmetric X-ray emission and powerful cluster-scale radio halo indicate that A2319 is a merging cluster of galaxies. This paper presents our multicolor photometry for A2319 with 15 optical intermediate filters in the Beijing-Arizona-Taiwan-Connecticut (BATC) system. There are 142 galaxies with known spectroscopic redshifts within the viewing field, including 128 member galaxies (called sample I).A large velocity dispersion in the rest frame suggests a merger dynamics in A2319. The contour map of projected density and localized velocity structure confirm the so-called A2319B substructure, at ~ 10' NW to the main concentration A2319A. The spectral energy distributions (SEDs) of more than 30,000 sources are obtained in our BATC photometry down to V ~ 20 mag. With color-color diagrams and photometric redshift technique, 233 galaxies brighter than h=19.0 are newly selected as member candidates. The early-type galaxies are found to follow a tight color-magnitude correlation. Based on sample I and the enlarged sample of member galaxies (called sample II), subcluster A2319B is confirmed. A strong environmental effect on star formation histories is found in the manner that galaxies in the sparse regions have various star formation histories, while galaxies in the dense regions are found to have shorter SFR time scales, older stellar ages, and higher ISM metallicities. For the merging cluster A2319, local surface density is a better environmental indicator rather than the clustercentric distance. Compared with the well-relaxed cluster A2589, a higher fraction of star-forming galaxies is found in A2319, indicating that the galaxy-scale turbulence stimulated by the subcluster merger might have played a role in triggering the star formation activity.

### PRIMUS + DEEP2: Clustering of X-ray, Radio and IR-AGN at z~0.7

We measure the clustering of X-ray, radio, and mid-IR-selected active galactic nuclei (AGN) at 0.2 < z < 1.2 using multi-wavelength imaging and spectroscopic redshifts from the PRIMUS and DEEP2 redshift surveys, covering 7 separate fields spanning ~10 square degrees. Using the cross-correlation of AGN with dense galaxy samples, we measure the clustering scale length and slope, as well as the bias, of AGN selected at different wavelengths. Similar to previous studies, we find that X-ray and radio AGN are more clustered than mid-IR-selected AGN. We further compare the clustering of each AGN sample with matched galaxy samples designed to have the same stellar mass, star formation rate, and redshift distributions as the AGN host galaxies and find no significant differences between their clustering properties. The observed differences in the clustering of AGN selected at different wavelengths can therefore be explained by the clustering differences of their host populations, which have different distributions in both stellar mass and star formation rate. Selection biases inherent in AGN selection, therefore, determine the clustering of observed AGN samples. We further find no significant difference between the clustering of obscured and unobscured AGN, using IRAC or WISE colors or X-ray hardness ratio.

### Planck Intermediate Results. XXXVI. Optical identification and redshifts of Planck SZ sources with telescopes in the Canary Islands Observatories

We present the results of approximately three years of observations of Planck Sunyaev-Zeldovich (SZ) sources with telescopes at the Canary Islands observatories, as part of the general optical follow-up programme undertaken by the Planck collaboration. In total, 78 SZ sources are discussed. Deep imaging observations were obtained for most of those sources; spectroscopic observations in either in long-slit or multi-object modes were obtained for many. We found optical counterparts for 73 of the 78 candidates. This sample includes 53 spectroscopic redshifts determinations, 20 of them obtained with a multi-object spectroscopic mode. The sample contains new redshifts for 27 Planck clusters that were not included in the first Planck SZ source catalogue (PSZ1).

### The Grism Lens-Amplified Survey from Space (GLASS). IV. Mass reconstruction of the lensing cluster Abell 2744 from frontier field imaging and GLASS spectroscopy [Replacement]

We present a strong and weak lensing reconstruction of the massive cluster Abell 2744, the first cluster for which deep Hubble Frontier Field (HFF) images and spectroscopy from the Grism Lens-Amplified Survey from Space (GLASS) are available. By performing a targeted search for emission lines in multiply imaged sources using the GLASS spectra, we obtain 5 high-confidence spectroscopic redshifts and 2 tentative ones. We confirm 1 strongly lensed system by detecting the same emission lines in all 3 multiple images. We also search for additional line emitters blindly and use the full GLASS spectroscopic catalog to test reliability of photometric redshifts for faint line emitters. We see a reasonable agreement between our photometric and spectroscopic redshift measurements, when including nebular emission in photometric redshift estimations. We introduce a stringent procedure to identify only secure multiple image sets based on colors, morphology, and spectroscopy. By combining 7 multiple image systems with secure spectroscopic redshifts (at 5 distinct redshift planes) with 18 multiple image systems with secure photometric redshifts, we reconstruct the gravitational potential of the cluster pixellated on an adaptive grid, using a total of 72 images. The resulting mass map is compared with a stellar mass map obtained from the deep Spitzer Frontier Fields data to study the relative distribution of stars and dark matter in the cluster. We find that the stellar to total mass ratio varies substantially across the cluster field, suggesting that stars do not trace exactly the total mass in this interacting system. The maps of convergence, shear, and magnification are made available in the standard HFF format.

### The Grism Lens-Amplified Survey from Space (GLASS). IV. Mass reconstruction of the lensing cluster Abell 2744 from frontier field imaging and GLASS spectroscopy

We present a strong and weak lensing reconstruction of the massive cluster Abell 2744, the first cluster for which deep \emph{Hubble Frontier Field} (HFF) images and spectroscopy from the \emph{Grism Lens-Amplified Survey from Space} (GLASS) are available. By performing a targeted search for emission lines in multiply imaged sources using GLASS spectra, we obtain 5 secure spectroscopic redshifts and 2 tentative ones. We confirm 1 strongly lensed system by detecting the same emission lines in all 3 multiple images. We also search for additional line emitters blindly and use the full GLASS spectroscopic catalog to test reliability of photometric redshifts for faint line emitters. We see a reasonable agreement between our photometric and spectroscopic redshift measurements, when including nebular emission in photo-z estimations. We introduce a stringent procedure to identify only secure multiple image sets based on colors, morphology, and spectroscopy. By combining 7 multiple image systems with secure spectroscopic redshifts (at 5 distinct redshift planes) with 18 multiple image systems with secure photometric redshifts, we reconstruct the gravitational potential of the cluster pixellated on an adaptive grid, using a total of 72 images. The resulting mass map is compared with a stellar mass map obtained from the deep \emph{Spitzer} Frontier Fields data to study the relative distribution of stars and dark matter in the cluster. We find that the stellar to total mass ratio varies substantially across the cluster, suggesting that stars do not trace exactly the total mass in this interacting cluster. The maps of convergence, shear, and magnification are made available in the standard HFF format.

### Galaxy sizes as a function of environment at intermediate redshift from the ESO Distant Cluster Survey

In order to assess whether the environment has a significant effect on galaxy sizes, we compare the mass--size relations of cluster and field galaxies in the $0.4 < z < 0.8$ redshift range from the ESO Distant Cluster Survey (EDisCS) using HST images. We analyse two mass-selected samples, one defined using photometric redshifts ($10.2 \le \log M_\ast/M_{\odot} \le 12.0$), and a smaller more robust subsample using spectroscopic redshifts ($10.6 \le \log M_\ast/M_{\odot} \le 11.8$). We find no significant difference in the size distributions of cluster and field galaxies of a given morphology. Similarly, we find no significant difference in the size distributions of cluster and field galaxies of similar rest-frame $B-V$ colours. We rule out average size differences larger than $10$--$20$\% in both cases. Consistent conclusions are found with the spectroscopic and photometric samples. These results have important consequences for the physical process(es) responsible for the size evolution of galaxies, and in particular the effect of the environment. The remarkable growth in galaxy size observed from $z\sim2.5$ has been reported to depend on the environment at higher redshifts ($z>1$), with early-type/passive galaxies in higher density environments growing earlier. Such dependence disappears at lower redshifts. Therefore, if the reported difference at higher-$z$ is real, the growth of field galaxies has caught up with that of cluster galaxies by $z\sim1$. Any putative mechanism responsible for galaxy growth has to account for the existence of environmental differences at high redshift and their absence (or weakening) at lower redshifts.

### Anomaly detection for machine learning redshifts applied to SDSS galaxies

We present an analysis of anomaly detection for machine learning redshift estimation. Anomaly detection allows the removal of poor training examples, which can adversely influence redshift estimates. Anomalous training examples may be photometric galaxies with incorrect spectroscopic redshifts, or galaxies with one or more poorly measured photometric quantity. We select 2.5 million 'clean' SDSS DR12 galaxies with reliable spectroscopic redshifts, and 6730 'anomalous' galaxies with spectroscopic redshift measurements which are flagged as unreliable. We contaminate the clean base galaxy sample with galaxies with unreliable redshifts and attempt to recover the contaminating galaxies using the Elliptical Envelope technique. We then train four machine learning architectures for redshift analysis on both the contaminated sample and on the preprocessed 'anomaly-removed' sample and measure redshift statistics on a clean validation sample generated without any preprocessing. We find an improvement on all measured statistics of up to 80% when training on the anomaly removed sample as compared with training on the contaminated sample for each of the machine learning routines explored. We further describe a method to estimate the contamination fraction of a base data sample.

### The galaxy-halo connection from a joint lensing, clustering and abundance analysis in the CFHTLenS/VIPERS field

We present new constraints on the relationship between galaxies and their host dark matter halos, measured from the location of the peak of the stellar-to-halo mass ratio (SHMR), up to the most massive galaxy clusters at redshift $z\sim0.8$ and over a volume of nearly 0.1~Gpc$^3$. We use a unique combination of deep observations in the CFHTLenS/VIPERS field from the near-UV to the near-IR, supplemented by $\sim60\,000$ secure spectroscopic redshifts, analysing galaxy clustering, galaxy-galaxy lensing and the stellar mass function. We interpret our measurements within the halo occupation distribution (HOD) framework, separating the contributions from central and satellite galaxies. We find that the SHMR for the central galaxies peaks at $M_{\rm h, peak} = 1.9^{+0.2}_{-0.1}\times10^{12} M_{\odot}$ with an amplitude of $0.025$, which decreases to $\sim0.001$ for massive halos ($M_{\rm h} > 10^{14} M_{\odot}$). Compared to central galaxies only, the total SHMR (including satellites) is boosted by a factor 10 in the high-mass regime (cluster-size halos), a result consistent with cluster analyses from the literature based on fully independent methods. After properly accounting for differences in modelling, we have compared our results with a large number of results from the literature up to $z=1$: we find good general agreement, independently of the method used, within the typical stellar-mass systematic errors at low to intermediate mass (${M}_{\star} < 10^{11} M_{\odot}$) and the statistical errors above. We have also compared our SHMR results to semi-analytic simulations and found that the SHMR is tilted compared to our measurements in such a way that they over- (under-) predict star formation efficiency in central (satellite) galaxies.

### A WFC3 Grism Emission Line Redshift Catalog in the GOODS-South Field

We combine HST/WFC3 imaging and G141 grism observations from the CANDELS and 3D-HST surveys to produce a catalog of grism spectroscopic redshifts for galaxies in the CANDELS/GOODS-South field. The WFC3/G141 grism spectra cover a wavelength range of 1.1<lambda<1.7 microns with a resolving power of R~130 for point sources, thus providing rest-frame optical spectra for galaxies out to z~3.5. The catalog is selected in the H-band (F160W) and includes both galaxies with and without previously published spectroscopic redshifts. Grism spectra are extracted for all H-band detected galaxies with H<24 and a CANDELS photometric redshift z_phot > 0.6. The resulting spectra are visually inspected to identify emission lines and redshifts are determined using cross-correlation with empirical spectral templates. To establish the accuracy of our redshifts, we compare our results against high-quality spectroscopic redshifts from the literature. Using a sample of 411 control galaxies, this analysis yields a precision of sigma_NMAD=0.0028 for the grism-derived redshifts, which is consistent with the accuracy reported by the 3D-HST team. Our final catalog covers an area of 153 square arcmin and contains 1019 redshifts for galaxies in GOODS-S. Roughly 60% (608/1019) of these redshifts are for galaxies with no previously published spectroscopic redshift. These new redshifts span a range of 0.677 < z < 3.456 and have a median redshift of z=1.282. The catalog contains a total of 234 new redshifts for galaxies at z>1.5. In addition, we present 20 galaxy pair candidates identified for the first time using the grism redshifts in our catalog, including four new galaxy pairs at z~2, nearly doubling the number of such pairs previously identified.

### Data augmentation for machine learning redshifts applied to SDSS galaxies [Replacement]

We present analyses of data augmentation for machine learning redshift estimation. Data augmentation makes a training sample more closely resemble a test sample, if the two base samples differ, in order to improve measured statistics of the test sample. We perform two sets of analyses by selecting 800k (1.7M) SDSS DR8 (DR10) galaxies with spectroscopic redshifts. We construct a base training set by imposing an artificial r band apparent magnitude cut to select only bright galaxies and then augment this base training set by using simulations and by applying the K-correct package to artificially place training set galaxies at a higher redshift. We obtain redshift estimates for the remaining faint galaxy sample, which are not used during training. We find that data augmentation reduces the error on the recovered redshifts by 40% in both sets of analyses, when compared to the difference in error between the ideal case and the non augmented case. The outlier fraction is also reduced by at least 10% and up to 80% using data augmentation. We finally quantify how the recovered redshifts degrade as one probes to deeper magnitudes past the artificial magnitude limit of the bright training sample. We find that at all apparent magnitudes explored, the use of data augmentation with tree based methods provide a estimate of the galaxy redshift with a negligible bias, although the error on the recovered values increases as we probe to deeper magnitudes. These results have applications for surveys which have a spectroscopic training set which forms a biased sample of all photometric galaxies, for example if the spectroscopic detection magnitude limit is shallower than the photometric limit.

### Data augmentation for machine learning redshifts applied to SDSS galaxies

We present analyses of data augmentation for machine learning redshift estimation. Data augmentation makes a training sample more closely resemble a test sample, if the two base samples differ, in order to improve measured statistics of the test sample. We perform two sets of analyses by selecting 800k (1.7M) SDSS DR8 (DR10) galaxies with spectroscopic redshifts. We construct a base training set by imposing an artificial r band apparent magnitude cut to select only bright galaxies and then augment this base training set by using simulations and by applying the K-correct package to artificially place training set galaxies at a higher redshift. We obtain redshift estimates for the remaining faint galaxy sample, which are not used during training. We find that data augmentation reduces the error on the recovered redshifts by 40% in both sets of analyses, when compared to the difference in error between the ideal case and the non augmented case. The outlier fraction is also reduced by at least 10% and up to 80% using data augmentation. We finally quantify how the recovered redshifts degrade as one probes to deeper magnitudes past the artificial magnitude limit of the bright training sample. We find that at all apparent magnitudes explored, the use of data augmentation with tree based methods provide a estimate of the galaxy redshift with a negligible bias, although the error on the recovered values increases as we probe to deeper magnitudes. These results have applications for surveys which have a spectroscopic training set which forms a biased sample of all photometric galaxies, for example if the spectroscopic detection magnitude limit is shallower than the photometric limit.

### Quiescent Compact Galaxies at Intermediate Redshift in the COSMOS field. I. The Number Density

We investigate the evolution of compact galaxy number density over the redshift range $0.2<z<0.8$. Our sample consists of galaxies with secure spectroscopic redshifts observed in the COSMOS field. The compact galaxy number density is constant in the interval $0.2<z<0.8$. Our number density estimates are similar to the estimates at $z>1$ for equivalently selected compact samples. Small variations in the abundance of the COSMOS compact sources as a function of redshift correspond to known structures in the field. The constancy of the compact galaxy number density is robust and does not depend on the compactness threshold or the stellar mass range (for $M_\ast>10^{10}\, M_\odot$). To maintain constant number density any size growth of high-redshift compact systems with decreasing redshift must be balanced by formation of quiescent compact systems at $z<1$.

### Quiescent Compact Galaxies at Intermediate Redshift in the COSMOS Field. The Number Density [Replacement]

We investigate the evolution of compact galaxy number density over the redshift range $0.2<z<0.8$. Our sample consists of galaxies with secure spectroscopic redshifts observed in the COSMOS field. With the large uncertainties, the compact galaxy number density trend with redshift is consistent with a constant value over the interval $0.2<z<0.8$. Our number density estimates are similar to the estimates at $z>1$ for equivalently selected compact samples. Small variations in the abundance of the COSMOS compact sources as a function of redshift correspond to known structures in the field. The constancy of the compact galaxy number density is robust and insensitive to the compactness threshold or the stellar mass range (for $M_\ast>10^{10}\, M_\odot$). To maintain constant number density any size growth of high-redshift compact systems with decreasing redshift must be balanced by formation of quiescent compact systems at $z<1$.

### Cold imprint of supervoids in the Cosmic Microwave Background re-considered with Planck and BOSS DR10 [Replacement]

We analyze publicly available void catalogs of the Baryon Oscillation Spectroscopic Survey Data Release 10 at redshifts $0.4<z<0.7$. The first goal of this paper is to extend the Cosmic Microwave Background stacking analysis of previous spectroscopic void samples at $z<0.4$. In addition, the DR10 void catalog provides the first chance to spectroscopically probe the volume of the Granett et al. (2008) supervoid catalog that constitutes the only set of voids which has shown a significant detection of a cross-correlation signal between void locations and average CMB chill. We found that the positions of voids identified in the spectroscopic DR10 galaxy catalog typically do not coincide with the locations of the Granett et al. supervoids in the overlapping volume, in spite of the presence of large underdense regions of high void-density in DR10. This failure to locate the same structures with spectroscopic redshifts may arise due to systematic differences in the properties of voids detected in photometric and spectroscopic samples. In the stacking measurement, we first find a $\Delta T = - 11.5 \pm 3.7~\mu K$ imprint for 35 of the 50 Granett et al. supervoids available in the DR10 volume. For the DR10 void catalog, lacking a prior on the number of voids to be considered in the stacking analysis, we find that the correlation measurement is fully consistent with no correlation. However, the measurement peaks with amplitude $\Delta T = - 9.8 \pm 4.8~\mu K$ for the a posteriori-selected 44 largest voids of size $R>65~Mpc/h$ that does match in terms of amplitude and number of structures the Granett et al. observation, although at different void positions.

### Dust Attenuation in High Redshift Galaxies -- 'Diamonds in the Sky'

We use observed optical to near infrared spectral energy distributions (SEDs) of 266 galaxies in the COSMOS survey to derive the wavelength dependence of the dust attenuation at high redshift. All of the galaxies have spectroscopic redshifts in the range z = 2 to 6.5. The presence of the CIV absorption feature, indicating that the rest-frame UV-optical SED is dominated by OB stars, is used to select objects for which the intrinsic, unattenuated spectrum has a well-established shape. Comparison of this intrinsic spectrum with the observed broadband photometric SED then permits derivation of the wavelength dependence of the dust attenuation. The derived dust attenuation curve is similar in overall shape to the Calzetti curve for local starburst galaxies. We also see the 2175 \AA~bump feature which is present in the Milky Way and LMC extinction curves but not seen in the Calzetti curve. The bump feature is commonly attributed to graphite or PAHs. No significant dependence is seen with redshift between sub-samples at z = 2 - 4 and z = 4 - 6.5. The 'extinction' curve obtained here provides a firm basis for color and extinction corrections of high redshift galaxy photometry.

### PRIMUS: Galaxy Environment on the Quiescent Fraction Evolution at z < 0.8

We investigate the effects of galaxy environment on the evolution of the quiescent fraction ($f_\mathrm{Q}$) from z =0.8 to 0.0 using spectroscopic redshifts and multi-wavelength imaging data from the PRIsm MUlti-object Survey (PRIMUS) and the Sloan Digitial Sky Survey (SDSS). Our stellar mass limited galaxy sample consists of ~14,000 PRIMUS galaxies within z = 0.2-0.8 and ~64,000 SDSS galaxies within z = 0.05-0.12. We classify the galaxies as quiescent or star-forming based on an evolving specific star formation cut, and as low or high density environments based on fixed cylindrical aperture environment measurements on a volume-limited environment defining population. For quiescent and star-forming galaxies in low or high density environments, we examine the evolution of their stellar mass function (SMF). Then using the SMFs we compute $f_\mathrm{Q}(M_{*})$ and quantify its evolution within our redshift range. We find that the quiescent fraction is higher at higher masses and in denser environments. The quiescent fraction rises with cosmic time for all masses and environments. At a fiducial mass of $10^{10.5}M_\odot$, from z~0.7 to 0.1, the quiescent fraction rises by 15% at the lowest environments and by 25% at the highest environments we measure. These results suggest that for a minority of galaxies their cessation of star formation is due to external influences on them. However, in the recent Universe a substantial fraction of the galaxies that cease forming stars do so due to internal processes.

### PRIMUS: Effect of Galaxy Environment on the Quiescent Fraction Evolution at z < 0.8 [Replacement]

We investigate the effects of galaxy environment on the evolution of the quiescent fraction ($f_\mathrm{Q}$) from z =0.8 to 0.0 using spectroscopic redshifts and multi-wavelength imaging data from the PRIsm MUlti-object Survey (PRIMUS) and the Sloan Digitial Sky Survey (SDSS). Our stellar mass limited galaxy sample consists of ~14,000 PRIMUS galaxies within z = 0.2-0.8 and ~64,000 SDSS galaxies within z = 0.05-0.12. We classify the galaxies as quiescent or star-forming based on an evolving specific star formation cut, and as low or high density environments based on fixed cylindrical aperture environment measurements on a volume-limited environment defining population. For quiescent and star-forming galaxies in low or high density environments, we examine the evolution of their stellar mass function (SMF). Then using the SMFs we compute $f_\mathrm{Q}(M_{*})$ and quantify its evolution within our redshift range. We find that the quiescent fraction is higher at higher masses and in denser environments. The quiescent fraction rises with cosmic time for all masses and environments. At a fiducial mass of $10^{10.5}M_\odot$, from z~0.7 to 0.1, the quiescent fraction rises by 15% at the lowest environments and by 25% at the highest environments we measure. These results suggest that for a minority of galaxies their cessation of star formation is due to external influences on them. However, in the recent Universe a substantial fraction of the galaxies that cease forming stars do so due to internal processes.

### Optical & Sunyaev-Zel'dovich Observations of a New Sample of Distant Rich Galaxy Clusters in the ROSAT All Sky Survey [Replacement]

Finding a sample of the most massive clusters with redshifts $z>0.6$ can provide an interesting consistency check of the $\Lambda$ cold dark matter ($\Lambda$CDM) model. Here we present results from our search for clusters with $0.6\lesssim z\lesssim1.0$ where the initial candidates were selected by cross-correlating the RASS faint and bright source catalogues with red galaxies from the Sloan Digital Sky Survey DR8. Our survey thus covers $\approx10,000\,\rm{deg^2}$, much larger than previous studies of this kind. Deeper follow-up observations in three bands using the William Herschel Telescope and the Large Binocular Telescope were performed to confirm the candidates, resulting in a sample of 44 clusters for which we present richnesses and red sequence redshifts, as well as spectroscopic redshifts for a subset. At least two of the clusters in our sample are comparable in richness to RCS2-$J$232727.7$-$020437, one of the richest systems discovered to date. We also obtained new observations with the Combined Array for Research in Millimeter Astronomy for a subsample of 21 clusters. For 11 of those we detect the Sunyaev-Zel'dovich effect signature. The Sunyaev-Zel'dovich signal allows us to estimate $M_{200}$ and check for tension with the cosmological standard model. We find no tension between our cluster masses and the $\Lambda$CDM model.

### Optical & Sunyaev-Zel'dovich Observations of a New Sample of Distant Rich Galaxy Clusters in the ROSAT All Sky Survey

Finding a sample of the most massive clusters with redshifts $z>0.6$ can provide an interesting consistency check of the $\Lambda$CDM model. Here we present results from our search for clusters with $0.6<z<1.0$ where the initial candidates were selected by cross-correlating the RASS faint and bright source catalogs with red galaxies from the Sloan Digital Sky Survey DR8. Our survey thus covers $\approx10,000\,\rm{deg^2}$, much larger than previous studies. Deeper follow-up observations in three bands using the William Herrschel Telescope and the Large Binocular Telescope were performed to confirm the candidates, resulting in a sample of 44 clusters for which we present richnesses and red sequence redshifts, as well as spectroscopic redshifts for a subset. At least two of the clusters in our sample are comparable in richness to RCS2-$J$232727.7$-$020437, one of the richest systems discovered to date. We also obtained new observations with the Combined Array for Research in Millimeter Astronomy for a subsample of 21 clusters. For eleven of those we detect the Sunyaev-Zel'dovich effect signature. The Sunyaev-Zel'dovich signal allows us to estimate $M_{200}$ and check for tension with the cosmological standard model. We find no tension between our cluster masses and the $\Lambda$CDM model.

### Galaxy triplets in Sloan Digital Sky Survey Data Release 7 - III. Analysis of Configuration and Dynamics

We analyse the spatial configuration and the dynamical properties of a sample of 92 galaxy triplets obtained from the SDSS-DR7 (SDSS-triplets) restricted to have members with spectroscopic redshifts in the range $0.01\le z \le 0.14$ and absolute r-band luminosities brighter than $M_r=-20.5$. The configuration analysis was performed through Agekyan & Anosova map (AA-map). We estimated dynamical parameters, namely the radius of the system, the velocity dispersion, a dimensionless crossing-time and the virial mass. We compared our results with those obtained for a sample of triplets from the catalogue "Isolated Triplets of Galaxies" (K-triplets) and a sample of Compact Groups. We have also studied a mock catalogue in order to compare real and projected configurations, and to estimate the three dimensional dynamical parameters of the triple systems. We found that the SDSS-triplets prefer alignment configurations while K-triplets present an uniform distribution in the AA-map. From the dynamical analysis we conclude that the SDSS-triplets, K-triplets and Compact Groups present a similar behaviour comprising compact systems with low crossing-time values, with velocity dispersions and virial masses similar to those of low mass loose groups. Moreover, we found that observed and simulated triplets present similar dynamical parameters. We also performed an analysis of the dark matter content of galaxy triplets finding that member galaxies of mock triplets belong to the same dark matter halo, showing a dynamical co-evolution of the system. These results suggest that the configuration and dynamics of triple systems favour galaxy interactions and mergers.

### The MUSE 3D view of the Hubble Deep Field South

We observed the Hubble Deep Field South with the new panoramic integral field spectrograph MUSE that we built and just commissioned at the VLT. The data cube resulting from 27 hours of integration covers one arcmin^2 field of view at an unprecedented depth with a 1 sigma emission line surface brightness limit of 1x$10^{-19}$ erg/s/cm$^2$/arcsec$^2$ and contains ~90,000 spectra. We present the combined and calibrated data cube, and we perform a first-pass analysis of the sources detected in the HDF-S imaging. We measured the redshifts of 189 sources up to a magnitude F814W = 29.5, increasing by more than an order of magnitude the number of known spectroscopic redshifts in this field. We also discovered 26 Lya emitting galaxies which are not detected in the HST WFPC2 deep broad band images. The intermediate spectral resolution of 2.3{\AA} allows us to separate resolved asymmetric Lya emitters, [O II] emitters, and C III] emitters and the large instantaneous wavelength range of 4500{\AA} helps to identify single emission lines. We also show how the three dimensional information of MUSE helps to resolve sources which are confused at ground-based image quality. Overall, secure identifications are provided for 83% of the 227 emission line sources detected in the MUSE data cube and for 32% of the 586 sources identified in the HST catalog of Casertano et al 2000. The overall redshift distribution is fairly flat to z=6.3, with a reduction between z=1.5 to 2.9, in the well-known redshift desert. The field of view of MUSE also allowed us to detect 17 groups within the field. We checked that the number counts of [O II] and Ly-a emitters are roughly consistent with predictions from the literature. Using two examples we demonstrate that MUSE is able to provide exquisite spatially resolved spectroscopic information on intermediate redshift galaxies present in the field.

### The evolving SFR-M_star relation and sSFR since z~5 from the VUDS spectroscopic survey

We study the evolution of the star formation rate (SFR) - stellar mass (M_star) relation and specific star formation rate (sSFR) of star forming galaxies (SFGs) since a redshift z~5.5 using 2435 (4531) galaxies with highly reliable (reliable) spectroscopic redshifts in the VIMOS Ultra-Deep Survey (VUDS). It is the first time that these relations can be followed over such a large redshift range from a single homogeneously selected sample of galaxies with spectroscopic redshifts. The log(SFR) - log(M_star) relation for SFGs remains roughly linear all the way up to z=5 but the SFR steadily increases at fixed mass with increasing redshift. We find that for stellar masses M_star>3.2 x 10^9 M_sun the SFR increases by a factor ~13 between z=0.4 and z=2.3. We extend this relation up to z=5, finding an additional increase in SFR by a factor 1.7 from z=2.3 to z=4.8 for masses M_star > 10^10 M_sun. We observe a turn-off in the SFR-M_star relation at the highest mass end up to a redshift z~3.5. We interpret this turn-off as the signature of a strong on-going quenching mechanism and rapid mass growth. The sSFR increases strongly up to z~2 but it grows much less rapidly in 2<z<5. We find that the shape of the sSFR evolution is not well reproduced by cold gas accretion-driven models or the latest hydrodynamical models. Below z~2 these models have a flatter evolution (1+z)^{Phi} with Phi=2-2.25 compared to the data which evolves more rapidly with Phi=2.8+-0.2. Above z~2, the reverse is happening with the data evolving more slowly with Phi=1.2+-0.1. The observed sSFR evolution over a large redshift range 0<z<5 and our finding of a non linear main sequence at high mass both indicate that the evolution of SFR and M_star is not solely driven by gas accretion. The results presented in this paper emphasize the need to invoke a more complex mix of physical processes {abridge}

### Comparing gravitational redshifts of SDSS galaxy clusters with the magnification redshift enhancement of background BOSS galaxies

A clean measurement of the evolution of the galaxy cluster mass function can significantly improve our understanding of cosmology from the rapid growth of cluster masses below z < 0.5. Here we examine the consistency of cluster catalogues selected from the SDSS by applying two independent gravity-based methods using all available spectroscopic redshifts from the DR10 release. First, we detect a gravitational redshift related signal for 20,119 and 13,128 clusters with spectroscopic redshifts contained in the GMBCG and redMaPPer catalogues, respectively, at a level of $\sim - 10$ km s$^{-1}$. This we show is consistent with the magnitude expected using the richness-mass relations provided by the literature and after applying recently clarified relativistic and flux bias corrections. This signal is also consistent with the richest clusters in the larger catalogue of Wen et al. (2012), corresponding to $M_{200m} \gtrsim 2 \times 10^{14}\,\mathrm{M}_\odot\,h^{-1}$, however we find no significant detection of gravitational redshift signal for less riched clusters, which may be related to bulk motions from substructure and spurious cluster detections. Second, we find all three catalogues generate mass-dependent levels of lensing magnification bias, which enhances the mean redshift of flux-selected background galaxies from the BOSS survey. The magnitude of this lensing effect is generally consistent with the corresponding richness-mass relations advocated for the surveys. We conclude that all catalogues comprise a high proportion of reliable clusters, and that the GMBCG and redMaPPer cluster finder algorithms favor more relaxed clusters with a meaningful gravitational redshift signal, as anticipated by the red-sequence colour selection of the GMBCG and redMaPPer samples.

### Feature importance for machine learning redshifts applied to SDSS galaxies [Replacement]

We present an analysis of importance feature selection applied to photometric redshift estimation using the machine learning architecture Decision Trees with the ensemble learning routine Adaboost (hereafter RDF). We select a list of 85 easily measured (or derived) photometric quantities (or `features') and spectroscopic redshifts for almost two million galaxies from the Sloan Digital Sky Survey Data Release 10. After identifying which features have the most predictive power, we use standard artificial Neural Networks (aNN) to show that the addition of these features, in combination with the standard magnitudes and colours, improves the machine learning redshift estimate by 18% and decreases the catastrophic outlier rate by 32%. We further compare the redshift estimate using RDF with those from two different aNNs, and with photometric redshifts available from the SDSS. We find that the RDF requires orders of magnitude less computation time than the aNNs to obtain a machine learning redshift while reducing both the catastrophic outlier rate by up to 43%, and the redshift error by up to 25%. When compared to the SDSS photometric redshifts, the RDF machine learning redshifts both decreases the standard deviation of residuals scaled by 1/(1+z) by 36% from 0.066 to 0.041, and decreases the fraction of catastrophic outliers by 57% from 2.32% to 0.99%.

### Feature importance for machine learning redshifts applied to SDSS galaxies

We present an analysis of importance feature selection applied to photometric redshift estimation using the machine learning architecture Random Decision Forests (RDF) with the ensemble learning routine Adaboost. We select a list of 85 easily measured (or derived) photometric quantities (or 'features') and spectroscopic redshifts for almost two million galaxies from the Sloan Digital Sky Survey Data Release 10. After identifying which features have the most predictive power, we use standard artificial Neural Networks (aNN) to show that the addition of these features, in combination with the standard magnitudes and colours, improves the machine learning redshift estimate by 18% and decreases the catastrophic outlier rate by 32%. We further compare the redshift estimate from RDF using the ensemble learning routine Adaboost with those from two different aNNs, and with photometric redshifts available from the SDSS. We find that the RDF requires orders of magnitude less computation time than the aNNs to obtain a machine learning redshift while reducing both the catastrophic outlier rate by up to 43%, and the redshift error by up to 25%. When compared to the SDSS photometric redshifts, the RDF machine learning redshifts both decreases the standard deviation of residuals scaled by 1/(1+z) by 36% from 0.066 to 0.041, and decreases the fraction of catastrophic outliers by 57% from 2.32% to 0.99%.

### Feature importance for machine learning redshifts applied to SDSS galaxies [Replacement]

We present an analysis of importance feature selection applied to photometric redshift estimation using the machine learning architecture Random Decision Forests (RDF) with the ensemble learning routine Adaboost. We select a list of 85 easily measured (or derived) photometric quantities (or 'features') and spectroscopic redshifts for almost two million galaxies from the Sloan Digital Sky Survey Data Release 10. After identifying which features have the most predictive power, we use standard artificial Neural Networks (aNN) to show that the addition of these features, in combination with the standard magnitudes and colours, improves the machine learning redshift estimate by 18% and decreases the catastrophic outlier rate by 32%. We further compare the redshift estimate from RDF using the ensemble learning routine Adaboost with those from two different aNNs, and with photometric redshifts available from the SDSS. We find that the RDF requires orders of magnitude less computation time than the aNNs to obtain a machine learning redshift while reducing both the catastrophic outlier rate by up to 43%, and the redshift error by up to 25%. When compared to the SDSS photometric redshifts, the RDF machine learning redshifts both decreases the standard deviation of residuals scaled by 1/(1+z) by 36% from 0.066 to 0.041, and decreases the fraction of catastrophic outliers by 57% from 2.32% to 0.99%.

### Feature importance for machine learning redshifts applied to SDSS galaxies [Replacement]

We present an analysis of importance feature selection applied to photometric redshift estimation using the machine learning architecture Random Decision Forests (RDF) with the ensemble learning routine Adaboost. We select a list of 85 easily measured (or derived) photometric quantities (or 'features') and spectroscopic redshifts for almost two million galaxies from the Sloan Digital Sky Survey Data Release 10. After identifying which features have the most predictive power, we use standard artificial Neural Networks (aNN) to show that the addition of these features, in combination with the standard magnitudes and colours, improves the machine learning redshift estimate by 18% and decreases the catastrophic outlier rate by 32%. We further compare the redshift estimate from RDF using the ensemble learning routine Adaboost with those from two different aNNs, and with photometric redshifts available from the SDSS. We find that the RDF requires orders of magnitude less computation time than the aNNs to obtain a machine learning redshift while reducing both the catastrophic outlier rate by up to 43%, and the redshift error by up to 25%. When compared to the SDSS photometric redshifts, the RDF machine learning redshifts both decreases the standard deviation of residuals scaled by 1/(1+z) by 36% from 0.066 to 0.041, and decreases the fraction of catastrophic outliers by 57% from 2.32% to 0.99%.

### Spectroscopic Needs for Training of LSST Photometric Redshifts

This white paper summarizes those conclusions of the Snowmass White Paper "Spectroscopic Needs for Imaging Dark Energy Experiments" (arXiv:1309.5384) which are relevant to the training of LSST photometric redshifts; i.e., the use of spectroscopic redshifts to improve algorithms and reduce photo-z errors. The larger and more complete the available training set is, the smaller the RMS error in photo-z estimates should be, increasing LSST's constraining power. Among the better US-based options for this work are the proposed MANIFEST fiber feed for the Giant Magellan Telescope or (with lower survey speed) the WFOS spectrograph on the Thirty Meter Telescope (TMT). Due to its larger field of view and higher multiplexing, the PFS spectrograph on Subaru would be able to obtain a baseline training sample faster than TMT; comparable performance could be achieved with a highly-multiplexed spectrograph on Gemini with at least a 20 arcmin diameter field of view.

### Milky Way dust extinction measured with QSOs

We investigate reddening by Milky Way dust in the low-extinction regime of $E_{B-V}<0.15$. Using over 50,000 QSOs at $0.5<z<2.5$ from the SDSS DR7 QSO Catalogue we probe the residual SDSS colours after dereddening and correcting for the known spectroscopic redshifts. We find that the extinction vector of Schlafly & Finkbeiner (2011) is a better fit to the data than that used by Schlegel et al. (1998, SFD). There is evidence for a non-linearity in the SFD reddening map, which is similarly present in the V1.2 map of the Planck Collaboration. This non-linearity is similarly seen when galaxies or stars are used as probes of the SFD map.

### Estimating Spectroscopic Redshifts by Using k Nearest Neighbors Regression I. Description of Method and Analysis [Replacement]

Context: In astronomy, new approaches to process and analyze the exponentially increasing amount of data are inevitable. While classical approaches (e.g. template fitting) are fine for objects of well-known classes, alternative techniques have to be developed to determine those that do not fit. Therefore a classification scheme should be based on individual properties instead of fitting to a global model and therefore loose valuable information. An important issue when dealing with large data sets is the outlier detection which at the moment is often treated problem-orientated. Aims: In this paper we present a method to statistically estimate the redshift z based on a similarity approach. This allows us to determine redshifts in spectra in emission as well as in absorption without using any predefined model. Additionally we show how an estimate of the redshift based on single features is possible. As a consequence we are e.g. able to filter objects which show multiple redshift components. We propose to apply this general method to all similar problems in order to identify objects where traditional approaches fail. Methods: The redshift estimation is performed by comparing predefined regions in the spectra and applying a k nearest neighbor regression model for every predefined emission and absorption region, individually. Results: We estimated a redshift for more than 50% of the analyzed 16,000 spectra of our reference and test sample. The redshift estimate yields a precision for every individually tested feature that is comparable with the overall precision of the redshifts of SDSS. In 14 spectra we find a significant shift between emission and absorption or emission and emission lines. The results show already the immense power of this simple machine learning approach for investigating huge databases such as the SDSS.

### Estimating Spectroscopic Redshifts by Using k Nearest Neighbors Regression I. Description of Method and Analysis

Context: In astronomy, new approaches to process and analyze the exponentially increasing amount of data are inevitable. While classical approaches (e.g. template fitting) are fine for objects of well-known classes, alternative techniques have to be developed to determine those that do not fit. Therefore a classification scheme should be based on individual properties instead of fitting to a global model and therefore loose valuable information. An important issue when dealing with large data sets is the outlier detection which at the moment is often treated problem-orientated. Aims: In this paper we present a method to statistically estimate the redshift z based on a similarity approach. This allows us to determine redshifts in spectra in emission as well as in absorption without using any predefined model. Additionally we show how an estimate of the redshift based on single features is possible. As a consequence we are e.g. able to filter objects which show multiple redshift components. We propose to apply this general method to all similar problems in order to identify objects where traditional approaches fail. Methods: The redshift estimation is performed by comparing predefined regions in the spectra and applying a k nearest neighbor regression model for every predefined emission and absorption region, individually. Results: We estimated a redshift for more than 50% of the analyzed 16,000 spectra of our reference and test sample. The redshift estimate yields a precision for every individually tested feature that is comparable with the overall precision of the redshifts of SDSS. In 14 spectra we find a significant shift between emission and absorption or emission and emission lines. The results show already the immense power of this simple machine learning approach for investigating huge databases such as the SDSS.

### Clustering properties of moderate luminosity X-ray selected Type 1 and Type 2 AGN at z~3

We investigate, for the first time at z~3, the clustering properties of 189 Type 1 and 157 Type 2 X-ray active galactic nuclei (AGN) of moderate luminosity (log<Lbol> = 45.3 erg/s), with photometric or spectroscopic redshifts in the range 2.2<z<6.8. These samples are based on Chandra and XMM-Newton data in COSMOS. We find that Type 1 and Type 2 COSMOS AGN at z=3 inhabit DMHs with typical mass of logMh = 12.84+0.10/-0.11 and 11.73+0.39/-0.45 Msun/h, respectively. This result requires a drop in the halo masses of Type 1 and 2 COSMOS AGN at z~3 compared to z<2 XMM COSMOS AGN with similar luminosities. Additionally, we infer that unobscured COSMOS AGN at z~3 reside in 10 times more massive halos compared to obscured COSMOS AGN, at 2.6sigma level. This result extends to z~3 that found in COSMOS at z<2, and rules out the picture in which obscuration is purely an orientation effect. A model which assumes that the AGN activity is triggered by major mergers is quite successful in predicting both the low halo mass of COSMOS AGN and the typical mass of luminous SDSS quasars at z~3, with the latter inhabiting more massive halos respect to moderate luminosity AGN. Alternatively we can argue, at least for Type 1 COSMOS AGN, that they are possibly representative of an early phase of fast (i.e. Eddington limited) BH growth induced by cosmic cold flows or disk instabilities. Given the moderate luminosity, these new fast growing BHs have masses of e7-8 Msun at z~3 which might evolve into e8.5-9 Msun mass BHs at z=0. Following our clustering measurements, we argue that this fast BH growth at z~3 in AGN with moderate luminosity occurs in DMHs with typical mass of 6 times e12 Msun/h.

### Early-type galaxies at intermediate redshift observed with HST WFC3: perspectives on recent star-formation

We present an analysis of the stellar populations of 102 visually-selected early-type galaxies (ETGs) with spectroscopic redshifts (0.3<z<1.5) from observations in the Early Release Science program with the Wide Field Camera 3 (WFC3) on {\it Hubble Space Telescope} (HST). We fit one- and two-component synthetic stellar models to the ETGs UV-optical-near-IR spectral energy distributions and find a large fraction (~40%) are likely to have experienced a minor (f$\lesssim$10% of stellar mass) burst of recent (t$\lesssim$1 Gyr) star-formation. The measured ages and mass fraction of the young stellar populations do not strongly trend with measurements of galaxy morphology. We note that massive (log(M[$M_{\odot}$])>10.5) recently star-forming ETGs appear to have larger sizes. Furthermore, high-mass, quiescent ETGs identified with likely companions populate a distinct region in the size-mass parameter space, in comparison with the distribution of massive ETGs with evidence of RSF. We conclude that both mechanisms of the quenching of star-formation in disk-like ETGs and (gas-rich, minor) merger activity contribute to the formation of young stars and the size-mass evolution of intermediate redshift ETGs. The number of ETGs for which we have both HST WFC3 panchromatic (especially UV) imaging and spectroscopically-confirmed redshifts is relatively small, therefore a conclusion on the relative roles of both of these mechanisms remains an open question.

### Galaxy And Mass Assembly (GAMA): Curation and reanalysis of 17.5k redshifts in the G10/COSMOS region

We discuss the construction of the Galaxy And Mass Assembly (GAMA) 10h region (G10) using publicly available data in the Cosmic Evolution Survey region (COSMOS) in order to extend the GAMA survey to z~1 in a single ~1deg$^2$. In order to obtain the maximum number of high precision spectroscopic redshifts we re-reduce all archival zCOSMOS-bright data and use the GAMA automatic cross-correlation redshift fitting code autoz. We combine autoz redshifts with all other available redshift information (zCOSMOS-bright 10k, PRIMUS, VVDS, SDSS and photometric redshifts) to calculate robust best-fit redshifts for all galaxies and visually inspect all 1D and 2D spectra to confirm automatically assigned redshifts. In total, we obtain 17,466 robust redshifts in the full COSMOS region. We then define the G10 region to be the central ~1deg$^2$ of COSMOS, which has relatively high spectroscopic completeness, and encompasses the CHILES VLA region. We define a combined r < 23.0 mag & i < 22.0 mag G10 sample (selected to have the highest bijective overlap) with which to perform future analysis. The G10 sample contains 10,247 sources with reliable high precision VLT-VIMOS spectra - with a median redshift of 0.55 and ~53% completeness to all non-stellar r < 23 mag & i < 22.0 mag sources, we define this to be the G10-HR sample. We also produce a full spectroscopic sample (G10-ALL) which contains a further 2,504 r < 23 mag & i < 22.0 mag galaxies with lower precision PRIMUS spectroscopy - sufficient for all GAMA-type analyses other than group finding. In total the G10- ALL sample contains 12,751 galaxies with reliable redshifts and is ~66% complete to r < 23 mag & i < 22.0 mag. All tables and spectra are released through the G10 cutout tool at: http://ict.icrar.org/cutout/G10.

### The temperature dependence of the far-infrared-radio correlation in the Herschel-ATLAS

We use 10,387 galaxies from the Herschel Astrophysical TeraHertz Large Area Survey (H-ATLAS) to probe the far-infrared radio correlation (FIRC) of star forming galaxies as a function of redshift, wavelength, and effective dust temperature. All of the sources in our 250 {\mu}m-selected sample have spectroscopic redshifts, as well as 1.4 GHz flux density estimates measured from the Faint Images of the Radio Sky at Twenty centimetres (FIRST) survey. This enables us to study not only individual sources, but also the average properties of the 250 {\mu}m selected population using median stacking techniques. We find that individual sources detected at $\geq 5\sigma$ in both the H-ATLAS and FIRST data have logarithmic flux ratios (i.e. FIRC $q_\lambda$ parameters) consistent with previous studies of the FIRC. In contrast, the stacked values show larger $q_\lambda$, suggesting excess far-IR flux density/luminosity in 250{\mu}m selected sources above what has been seen in previous analyses. In addition, we find evidence that 250 {\mu}m sources with warm dust SEDs have a larger 1.4 GHz luminosity than the cooler sources in our sample. Though we find no evidence for redshift evolution of the monochromatic FIRC, our analysis reveals significant temperature dependence. Whilst the FIRC is reasonably constant with temperature at 100 {\mu}m, we find increasing inverse correlation with temperature as we probe longer PACS and SPIRE wavelengths. These results may have important implications for the use of monochromatic dust luminosity as a star formation rate indicator in star-forming galaxies, and in the future, for using radio data to determine galaxy star formation rates.

### The FMOS-COSMOS survey of star-forming galaxies at z~1.6 III. Survey design, performance, and sample characteristics [Replacement]

We present a spectroscopic survey of galaxies in the COSMOS field using the Fiber Multi-Object Spectrograph (FMOS), a near-infrared instrument on the Subaru Telescope. Our survey is specifically designed to detect the Halpha emission line that falls within the H-band (1.6-1.8 um) spectroscopic window from star-forming galaxies with 1.4 < z < 1.7 and M_stellar>~10^10 Msolar. With the high multiplex capability of FMOS, it is now feasible to construct samples of over one thousand galaxies having spectroscopic redshifts at epochs that were previously challenging. The high-resolution mode (R~2600) effectively separates Halpha and [NII]6585 thus enabling studies of the gas-phase metallicity and photoionization state of the interstellar medium. The primary aim of our program is to establish how star formation depends on stellar mass and environment, both recognized as drivers of galaxy evolution at lower redshifts. In addition to the main galaxy sample, our target selection places priority on those detected in the far-infrared by Herschel/PACS to assess the level of obscured star formation and investigate, in detail, outliers from the star formation rate - stellar mass relation. Galaxies with Halpha detections are followed up with FMOS observations at shorter wavelengths using the J-long (1.11-1.35 um) grating to detect Hbeta and [OIII]5008 that provides an assessment of extinction required to measure star formation rates not hampered by dust, and an indication of embedded Active Galactic Nuclei. With 460 redshifts measured from 1153 spectra, we assess the performance of the instrument with respect to achieving our goals, discuss inherent biases in the sample, and detail the emission-line properties. Our higher-level data products, including catalogs and spectra, are available to the community.

### The FMOS-COSMOS survey of star-forming galaxies at z~1.6 III. Survey design, performance, and sample characteristics

We present a spectroscopic survey of galaxies in the COSMOS field using the Fiber Multi-Object Spectrograph (FMOS), a near-infrared instrument on the Subaru Telescope. Our survey is specifically designed to detect the Halpha emission line that falls within the H-band (1.6-1.8 micron) spectroscopic window from star-forming galaxies with M_stellar>10^10 Msolar and 1.4 < z < 1.7. With the high multiplex capabilities of FMOS, it is now feasible to construct samples of over one thousand galaxies having spectroscopic redshifts at epochs that were previously challenging. The high-resolution mode (R~2600) is implemented to effectively separate Halpha and [NII] emission lines thus enabling studies of gas-phase metallicity and photoionization conditions of the interstellar medium. The broad goals of our program are concerned with how star formation depends on stellar mass and environment, both recognized as drivers of galaxy evolution at lower redshifts. In addition to the main galaxy sample, our target selection places priority on those detected in the far-infrared by Herschel/PACS to assess the level of obscured star formation and investigate, in detail, outliers from the star formation rate - stellar mass relation. Galaxies with Halpha detections are followed up with FMOS observations at shorter wavelengths using the J-long (1.11-1.35 micron) grating to detect Hbeta and [OIII]5007 that provides an assessment of extinction required to measure star formation rates not hampered by dust and an indication of embedded Active Galactic Nuclei. With the first 1215 spectra that yield 401 redshifts, we assess the performance of the instrument with respect to achieving our goals, discuss inherent biases in the sample, and detail the emission-line properties. Our data are presented in the form of a catalog for use by the community.

### The XMM-LSS survey: the Class 1 cluster sample over the extended 11 deg$^2$ and its spatial distribution

This paper presents 52 X-ray bright galaxy clusters selected within the 11 deg$^2$ XMM-LSS survey. 51 of them have spectroscopic redshifts ($0.05<z<1.06$), one is identified at $z_{\rm phot}=1.9$, and all together make the high-purity "Class 1" (C1) cluster sample of the XMM-LSS, the highest density sample of X-ray selected clusters with a monitored selection function. Their X-ray fluxes, averaged gas temperatures (median $T_X=2$ keV), luminosities (median $L_{X,500}=5\times10^{43}$ ergs/s) and total mass estimates (median $5\times10^{13} h^{-1} M_{\odot}$) are measured, adapting to the specific signal-to-noise regime of XMM-LSS observations. The redshift distribution of clusters shows a deficit of sources when compared to the cosmological expectations, regardless of whether WMAP-9 or Planck-2013 CMB parameters are assumed. This lack of sources is particularly noticeable at $0.4 \lesssim z \lesssim 0.9$. However, after quantifying uncertainties due to small number statistics and sample variance we are not able to put firm (i.e. $>3 \sigma$) constraints on the presence of a large void in the cluster distribution. We work out alternative hypotheses and demonstrate that a negative redshift evolution in the normalization of the $L_{X}-T_X$ relation (with respect to a self-similar evolution) is a plausible explanation for the observed deficit. We confirm this evolutionary trend by directly studying how C1 clusters populate the $L_{X}-T_X-z$ space, properly accounting for selection biases. We point out that a systematically evolving, unresolved, central component in clusters and groups (AGN contamination or cool core) can impact the classification as extended sources and be partly responsible for the observed redshift distribution.[abridged]

### Clustering-based Redshift Estimation: Comparison to Spectroscopic Redshifts

We investigate the potential and accuracy of clustering-based redshift estimation using the method proposed by M\'enard et al. (2013). This technique enables the inference of redshift distributions from measurements of the spatial clustering of arbitrary sources, using a set of reference objects for which redshifts are known. We apply it to a sample of spectroscopic galaxies from the Sloan Digital Sky Survey and show that, after carefully controlling the sampling efficiency over the sky, we can estimate redshift distributions with high accuracy. Probing the full colour space of the SDSS galaxies, we show that we can recover the corresponding mean redshifts with an accuracy ranging from $\delta$z=0.001 to 0.01. We indicate that this mapping can be used to infer the redshift probability distribution of a single galaxy. We show how the lack of information on the galaxy bias limits the accuracy of the inference and show comparisons between clustering redshifts and photometric redshifts for this dataset. This analysis demonstrates, using real data, that clustering-based redshift inference provides a powerful data-driven technique to explore the redshift distribution of arbitrary datasets, without any prior knowledge on the spectral energy distribution of the sources.

### Searching for highly obscured AGN in the XMM-Newton serendipitous source catalog

The majority of active galactic nuclei (AGN) are obscured by large amounts of absorbing material that makes them invisible at many wavelengths. X-rays, given their penetrating power, provide the most secure way for finding these AGN. The XMM-Newton serendipitous source catalog is the largest catalog of X-ray sources ever produced; it contains about half a million detections. These sources are mostly AGN. We have derived X-ray spectral fits for very many 3XMM-DR4 sources ($\gtrsim$ 114 000 observations, corresponding to $\sim$ 77 000 unique sources), which contain more than 50 source photons per detector. Here, we use a subsample of $\simeq$ 1000 AGN in the footprint of the SDSS area (covering 120 deg$^2$) with available spectroscopic redshifts. We searched for highly obscured AGN by applying an automated selection technique based on X-ray spectral analysis that is capable of efficiently selecting AGN. The selection is based on the presence of either a) flat rest-frame spectra; b) flat observed spectra; c) an absorption turnover, indicative of a high rest-frame column density; or d) an Fe K$\alpha$ line with an equivalent width > 500 eV. We found 81 highly obscured candidate sources. Subsequent detailed manual spectral fits revealed that 28 of them are heavily absorbed by column densities higher than 10$^{23}$ cm$^{-2}$. Of these 28 AGN, 15 are candidate Compton-thick AGN on the basis of either a high column density, consistent within the 90% confidence level with N$_{\rm H}$ $>$10$^{24}$ cm$^{-2}$, or a large equivalent width (>500 eV) of the Fe K$\alpha$ line. Another six are associated with near-Compton-thick AGN with column densities of $\sim$ 5$\times$10$^{23}$ cm$^{-2}$. A combination of selection criteria a) and c) for low-quality spectra, and a) and d) for medium- to high-quality spectra, pinpoint highly absorbed AGN with an efficiency of 80%.

### Optical Confirmation and Redshift Estimation of the Planck Cluster Candidates overlapping the Pan-STARRS Survey

We report results of a study of Planck Sunyaev-Zel'dovich effect (SZE) selected galaxy cluster candidates using the Panoramic Survey Telescope & Rapid Response System (Pan-STARRS) imaging data. We first examine 150 Planck confirmed galaxy clusters with spectroscopic redshifts to test our algorithm for identifying optical counterparts and measuring their redshifts; our redshifts have a typical accuracy of $\sigma_{z/(1+z)} \sim 0.022$ for this sample. We then examine an additional 237 Planck galaxy cluster candidates that have no redshift in the source catalogue. Of these 237 unconfirmed cluster candidates we are able to confirm 60 galaxy clusters and measure their redshifts. A further 83 candidates are so heavily contaminated by stars due to their location near the Galactic plane that we do not attempt to identify counterparts. For the remaining 94 candidates we find no optical counterpart but use the depth of the Pan-STARRS1 data to estimate a redshift lower limit $z_{\text{lim}(10^{15})}$ beyond which we would not have expected to detect enough galaxies for confirmation. Scaling from the already published Planck sample, we expect that $\sim$12 of these unconfirmed candidates may be real clusters.

### Optical Confirmation and Redshift Estimation of the Planck Cluster Candidates overlapping the Pan-STARRS Survey [Replacement]

We report results of a study of Planck Sunyaev-Zel'dovich effect (SZE) selected galaxy cluster candidates using the Panoramic Survey Telescope & Rapid Response System (Pan-STARRS) imaging data. We first examine 150 Planck confirmed galaxy clusters with spectroscopic redshifts to test our algorithm for identifying optical counterparts and measuring their redshifts; our redshifts have a typical accuracy of $\sigma_{z/(1+z)} \sim 0.022$ for this sample. Using 60 random sky locations, we estimate that our chance of contamination through a random superposition is ~ 3 per cent. We then examine an additional 237 Planck galaxy cluster candidates that have no redshift in the source catalogue. Of these 237 unconfirmed cluster candidates we are able to confirm 60 galaxy clusters and measure their redshifts. A further 83 candidates are so heavily contaminated by stars due to their location near the Galactic plane that we do not attempt to identify counterparts. For the remaining 94 candidates we find no optical counterpart but use the depth of the Pan-STARRS1 data to estimate a redshift lower limit $z_{\text{lim}(10^{15})}$ beyond which we would not have expected to detect enough galaxies for confirmation. Scaling from the already published Planck sample, we expect that $\sim$12 of these unconfirmed candidates may be real clusters.

### Interacting LAEs at z = 5.1. Episodic star formation in a group of LAEs at z= 5.07

We are undertaking a search for high-redshift low luminosity Lyman Alpha sources in the SHARDS survey. Among the pre-selected Lyman Alpha sources 2 candidates were spotted, located 3.19 arcsec apart, and tentatively at the same redshift. Here we report on the spectroscopic confirmation with GTC of the Lyman Alpha emission from this pair of galaxies at a confirmed spectroscopic redshifts of z=5.07. Furthermore, one of the sources is interacting/merging with another close companion that looks distorted. Based on the analysis of the spectroscopy and additional photometric data, we infer that most of the stellar mass of these objects was assembled in a burst of star formation 100 Myr ago. A more recent burst (2 Myr old) is necessary to account for the measured Lyman Alpha flux. We claim that these two galaxies are good examples of Lyman Alpha sources undergoing episodic star formation. Besides, these sources very likely constitute a group of interacting Lyman Alpha emitters (LAEs).

### Photometric redshift analysis in the Dark Energy Survey Science Verification data

We present results from a study of the photometric redshift performance of the Dark Energy Survey (DES), using the early data from a Science Verification (SV) period of observations in late 2012 and early 2013 that provided science-quality images for almost 200 sq.~deg.~at the nominal depth of the survey. We assess the photometric redshift performance using about 15000 galaxies with spectroscopic redshifts available from other surveys. These galaxies are used, in different configurations, as a calibration sample, and photo-$z$'s are obtained and studied using most of the existing photo-$z$ codes. A weighting method in a multi-dimensional color-magnitude space is applied to the spectroscopic sample in order to evaluate the photo-$z$ performance with sets that mimic the full DES photometric sample, which is on average significantly deeper than the calibration sample due to the limited depth of spectroscopic surveys. Empirical photo-$z$ methods using, for instance, Artificial Neural Networks or Random Forests, yield the best performance in the tests, achieving core photo-$z$ resolutions $\sigma_{68} \sim 0.08$. Moreover, the results from most of the codes, including template fitting methods, comfortably meet the DES requirements on photo-$z$ performance, therefore, providing an excellent precedent for future DES data sets.