Schmidt et al. (2017) present the detection of the CIV emission line in multiple images of a quintuply imaged Lyα emitter at z=6.11. The lensed object behind RXJ2248 was first identified by Monna et al. (2014) and multiple components of the system have since then been spectroscopically confirmed via Lyα emission. The rest-frame stack of the GLASS spectra from the multiple components of the lensed system shown in the figure clearly shows the detection of CIV and the confirmation of the Lyα emission, improving the signal-to-noise of the CIV (and Lyα) detections in the individual spectra of the individual components of the lensed system. The stacked spectra indicate an equivalent width of Lyα and CIV of 68±6Å and 24±4Å, respectively. Thanks to the broad wavelength coverage of the near-infrared GLASS spectra, the detections of Lyα and CIV are accompanied by flux limits on other rest-frame UV lines including HeII, OIII and CIII which were not detected in the GLASS spectra.
From updated photometric measurements including deep Hubble Frontier Fields imaging, archival imaging, and Spitzer observations and the corresponding fits of the spectral energy distribution (SED), Schmidt et al. (2017) estimate the galaxy to have a mass of roughly 10^9 Msun and a star formation rate of approximately 10 Msun/yr (both corrected for lensing magnification) in fair agreement with previous estimates. The photometry also reveals a young stellar population with an age of ~50 Myrs and a system with low dust content (E(B-V) ~ 0.05).
Rest-frame UV emission lines are very useful beacons for studying the physical properties of galaxies and provide information beyond what photometry and SED fits provide. Hence, the detection of CIV as well as the limits on the non-detections of the other rest-frame UV lines provide an opportunity to study the physical properties of the z=6.11 Lyα emitter. In particular, Schmidt et al. (2017) compare the rest-frame UV emission line flux ratios obtained from the GLASS spectra, to the predictions from recent photoionization models of AGN and star formatting galaxies from Feltre et al. (2016) and Gutkin et al. (2016). Such a comparison can potentially constraint the ionization parameter (logU) and the gas-phase metallicity of the galaxy (Zgas). The flux limits from the GLASS data show that the ionizing radiation powering the CIV emission is most likely from star formation as opposed to AGN activity, but are not strong enough to give tight constraints on the logU and Zgas.
For more details please refer to Schmidt et al. (2017)
GLASS Paper VII: The Spatial Distribution of Star Formation
Using GLASS data, Vulcani et al. (2016a)
analyzed the spatial distribution of star formation in 76 galaxies in 10 clusters and 85 galaxies in the field at 0.3< z
<0.7. The samples are well matched in stellar mass (10^8-10^11 M⊙) and star formation rate (0.5-50 M⊙/yr). The authors have visually classified galaxies in terms of broad-band morphology, Hα morphology and likely physical process acting on the galaxy. They found that most Hα emitters have a spiral morphology (41±8% in clusters, 51±8% in the field), followed by mergers/interactions (28±8%, 31±7%, respectively) and early-type galaxies (remarkably as high as 29±8 in clusters and 15±6% in the field). A diversity of Hα morphologies have been detected, suggesting a diversity of physical processes. In clusters, 30±8% of the galaxies present a regular morphology, mostly consistent with star formation diffused uniformly across the stellar population (mostly in the disk component, when present). The second most common morphology (28±8%) is asymmetric/jellyfish, consistent with ram pressure stripping or other non-gravitational processes in 18±8% of the cases. Ram pressure stripping appears significantly less prominent in the field (2±2%), where the most common morphology/mechanism appears to be consistent with minor gas rich mergers or clump accretion.
Comparing the position of the peak of the Hα emission to that of the continuum, as traced by the F475W filter, the author found that in both environments, for most of the galaxies the displacement is smaller than 1.5 kpc and the average offset is ~0.5 kpc. The existence of the offset suggests that current star formation is not generally colocated with recent star formation, perhaps as the result of accretion of satellites or gas, or non gravitational interactions such as ram pressure stripping affecting the spatial distribution of the cold gas.
The emerging picture is that Hα emitters are a very heterogeneous population, characterized by a range of morphologies, sizes and SFRs, therefore, a simple explanation can not describe all the observations. Even though some small systematic differences between galaxies in the field and in clusters emerge, both populations present very mixed morphologies and experience a variety of processes.
This work demonstrates that the effects of cluster- specific mechanisms on galaxy evolution are detectable in GLASS unprecedented data. However, they are both subtle and complex. They are subtle in the sense that no dramatic trend is found between the morphology of the current star formation and the environment or other properties of the galaxy. Every trend is weak and there are always exceptions. A full understanding of this complexity requires larger samples and detailed and spatially resolved physical models.
GLASS Paper VIII: The Correlation Between Star Formation and Cluster Environment
Extending the study presented in Vulcani et al. (2016a)
described above, Vulcani et al. (2016b)
used the same GLASS data and the results from Vulcani et al. (2016a)
to characterize the spatial distribution of star formation in 76 galaxies in 10 clusters at 0.3< z <0.7. They correlated the properties of Hα emitters to a number of tracers of the cluster environment to investigate its role in driving galaxy transformations. Hα emitters are found in the clusters out to 0.5 virial radii, the maximum radius covered by GLASS. The peak of the Hα emission is offset with respect to the peak of the UV-continuum. Decomposing this offsets into a radial and tangential component, the radial component points away from the cluster center in 60% of the cases, with 95% confidence. The decompositions agree with cosmological simulations, i.e. the Hα emission offset correlates with galaxy velocity and ram pressure stripping signatures. Our clusters span a wide range of morphologies. Trends between Hα emitters properties and surface mass density distributions and X-ray emissions emerge only for unrelaxed clusters.
The most statistically significant result is that galaxies with asymmetric Hα distribution, interpreted as signatures of recent ram pressure stripping, are preferentially found within 0.3 r500, at higher local density conditions and higher X-ray counts and have a negative radial projected offset, i.e. the peak of the Hα emission is pointing away from the cluster center with respect to the continuum emission.
The lack of strong correlations with the global environment does not allow to identify a unique environmental effect originating from the cluster center. In contrast, correlations between Hα morphology and local number density emerge.
This work concludes that local processes, such as ram pressure, strangulation and galaxy-galaxy interactions, rather than processes taking place on large scale, are the most important, or at least the most easily detectable, drivers of environmental evolution.
For further details on the above studies please refer to:
Using the GLASS spectroscopic and very deep imaging data, Morishita et al. (2016) found that galaxy clusters are dual-role environments, which both accelerate and curtail galaxy evolution without affecting their sizes and structures.
Morishita et al. derived structural parameters of more than 3900 galaxies by using the GLASS grism spectroscopy combined with the imaging from the Hubble Frontier Fields. The sample is unique in terms of the low masses it proves, reaching log(M*/Msun)~7.8, a factor of 10-100 lower than previous studies at the epoch of the Universe observed in this study, about 5 Gyr ago.
The paper studied galaxy size—stellar mass relations in different environments, i.e. the cluster center (the most crowded environment in the universe) and the normal field, to study the impact of the environment on galaxy sizes. The four figures above show the size—stellar mass relations for four populations: blue and red galaxies in clusters and the field.
The origin of low-mass red galaxies is a key topic in galaxy evolution, and the target of this study. By considering structural parameters, color, and the difference in sizes between red and blue galaxies, Morishita et al. found a strong connection between low-mass red galaxies and blue galaxies of similar masses — evidence that blue galaxies are transformed by ram pressure stripping and starvation in cluster environments.
Yet, about 20% of the low-mass red galaxies could not be explained in the same scenario — that population is too old and dense (consistent with more massive log(M*/Msun) > 10 galaxies), though it also seems only to exist in cluster environments.
The authors concluded the cluster environment has two roles on low-mass galaxies evolution; one is killing galaxies (environmental quenching) without significant structural transformations; the other is an acceleration of the evolution phase, which forms high surface density galaxies more than 10 Gyr ago, spanning most masses.
For more information see Morishita et al. (2016).
The catalog used in this study will soon be available through the GLASS website.
In Hoag et al. (2016), the GLASS team used the GLASS spectroscopy combined with the deep imaging from the Hubble Frontier Fields (HFF) to produce a gravitational lens model of the galaxy cluster MACSJ0416.1-2403 (MACS0416), following the approach by Wang et al. (2015) who modeled Abell 2744. The GLASS data allowed the team to measure the spectroscopic redshifts of galaxies that are multiply imaged by the cluster, significantly improving the constraints on the lens model. Figures showing the spectroscopic confirmations of multiply-imaged galaxies with the GLASS data are available on this page. The above figure on the left shows a co-added F105W image from the HFF, Cluster Lensing and Supernova survey with Hubble (CLASH), and the GLASS direct imaging. Overlaid on the figure is the critical curve from the lens model (dark green line) with the multiple images (multi-colored circles) that have been discovered in the cluster field. The Gold circles enclose spectroscopically confirmed multiple images from the GLASS spectroscopy and previous spectroscopic programs. All of the other colored circles lack spectroscopy, so they were vetted more carefully. We only use the Gold and Silver objects to constrain the lens model as they are the most trustworthy multiple image candidates.
From the lens model, the team derived a map of the total surface mass density throughout the cluster. Using deep mid-infrared data from the Spitzer Frontier Fields program, the team obtained a map of the stellar surface mass density in the cluster. The figure on the right shows the projected stellar to total mass ratio (f*) throughout the central ~500 kpc of the cluster. The f* varies significantly with distance from the two BCGs (black dots). The mean projected stellar to total mass ratio is <f*> = 0.009 +/- 0.003 (stat.), using a diet-Salpeter IMF, in agreement with other measurements of <f*> in massive cluster environments.
For more information see Hoag et al. (2016).
In the most recent paper by the GLASS collaboration, Schmidt et al. (2015), we present a sample of 24 objects with emission lines consistent with being Lyα at redshifts around or above 7, at the heart of the epoch of reionization. The figure shows the redshift distribution of these line emitters. Taking advantage of the GLASS 1σ flux limits of 5 x 10-18 erg/s/cm2 (not corrected for lensing magnification) in each of the GLASS spectra, the Lyα emitters were assembled via visual inspection of the GLASS spectra of a sample of more than 150 photometrically selected Lyman break galaxies, in the first 6 completed GLASS clusters.
The discovered fraction of Lyα emitters is consistent with the number of detections, within the uncertainties, expected from the conditionally probability of Lyα emission measured from the ground at z ~ 7. Deep high-resolution ground-based follow-up spectroscopy is needed to confirm the Lyα emitters, as exemplified by the independent Keck-DEIMOS confirmation of one of the sources presented by Huang et al. (2015).
A stack of the most promising Lyα emitters with a mean redshift of 7.2 allowed us to study the spatial extent of the Lyα emission. We found it to be consistent with the spatial extent of the UV continuum. Extended Lyα emission, if present, is below the surface brightness detection threshold in the GLASS Lyα emitter stack.
For more information please see Schmidt et al. (2015)
The GLASS data yield spatially resolved Hα fluxes for all star-forming galaxies in the core (< 1 Mpc) of the clusters. In addition, each cluster is observed at two different position angles. These two orientations allow us to mitigate the impact of contamination from overlapping spectra, and reliably measure for the first time the relative position of the Hα emission with respect to the continuum.
In Vulcani et al. (2015)
we present a pilot study characterizing the spatial distribution of the Hα emission in cluster galaxies beyond the local universe based on WFC3-IR data.
We analyze two of the ten clusters in the GLASS sample. We select MACS0717.5+3745 and MACS1423.8+2404 because they are at similar redshift (z~0.55), so as to minimize evolutionary effects and differences in the sensitivity/selection function, and are in very different dynamical states, so as to span the range of expected environments. We use use foreground and background galaxies as field control sample.
Both in clusters and in the field, Hα is more extended than the rest-frame UV continuum in 60% of the cases, consistent with diffuse star formation and inside out growth. In ∼ 20% of the cases, the Hα emission appears more extended in cluster galaxies than in the field, pointing perhaps to ionized gas being stripped and/or star formation being enhanced at large radii. The peak of the Hα emission and that of the continuum are offset by less than 1 kpc.
The diversity of morphologies and sizes observed in Hα illustrates the complexity of the environmental process that regulate star formation.
Today we released the GLASS survey paper together with the first data product of the survey.
The paper gives a thorough overview of GLASS and its science goals.
We utilize the Hubble Frontier Field cluster MACSJ0717, the first cluster to be observed by GLASS, to illustrate the data quality and the high-level data products we will be releasing for each of the 10 GLASS clusters.
From a visual inspection with GiG and GiGzÂ of each spectrum brighter than H_AB = 23 we measure the redshift when sufficient information is present in the spectra.
Furthermore, we conduct a thorough search for emission lines through all the GLASS WFC3 spectra with the aim of measuring redshifts for sources with continuum fainter than H_AB = 23.
As a product of these efforts, we present a catalog of 139 emission-line based spectroscopic redshifts for extragalactic sources, including three new redshifts of multiple imaged systems.
The GLASS survey paper: http://adsabs.harvard.edu/abs/2015arXiv150900475T
The GLASS data products: https://archive.stsci.edu/prepds/glass/
The GLASS inspection software: https://github.com/kasperschmidt/GLASSinspectionGUIs
Abell 2744 is the first cluster for which both the GLASS spectroscopy and the Hubble Frontier Field (HFF) imaging are complete. InÂ Wang et al. (2015), we perform targeted and blind searches of faint emission lines in the GLASS data and thusÂ compile a sample of 55 galaxies with high-confidence spectroscopic redshift, 5 of which are multiply lensed and 3 of which belong to the same system. We also measure photometric redshifts for this sample fromÂ the HFF 7-filter exposures. A reasonable agreement between these two sorts of redshift estimation is seen, when nebular emission is considered in photometric redshift fitting.
Modeling gravitational potential of galaxy clusters is known as a notorious difficulty, partly because of the inclusion of mis-identified multiple image systems in lens modeling. In this work, we develop a stringent and rigorous procedure to screen off insecure image systems,Â based upon colors derived from HFF, image morphology and spectroscopy from GLASS. This leads to a sample of 25/72 multiple arc systems/images as secure out of a total of 57/179 candidate systems/images. The resulting total mass map is only given by the secure set of arcs.
We also derive a stellar mass map from the Spitzer Frontier Field data and find out thatÂ the stellar to total mass ratio varies substantially across the cluster field, ranging from 0 to 5%, and the light-traces-mass assumption is not valid in this merging cluster.