Current Research Interests

Measuring changes in the galaxy population over cosmic time tells us the history of the universe. (McIntosh)

We want to understand how and why galaxies grow. Galaxies are the domain of evolved matter (stars, planets, and the chemistry for life). In the context of the standard theory of cosmology, galaxies are the condensation of gaseous atomic material at the centers of deep gravitation wells of dark matter. As the universe ages, expansion and gravity conspire to accentuate regions of greater and lesser dark matter density, thereby creating a `cosmic web’ of galactic beacons mapping the universe in both time and space.

Galaxies undergoing physical transformations provide clues to the growth processes shaping their morphology, structure, star production, and central black hole activity. Modern galaxy formation theory predicts that galaxies grow through the addition of new stars by two basic channels: (1) convert atomic gas into new stars, or (2) add pre-existing stars made elsewhere. The gravitational merging of dark matter halos is predicted to drive both the accretion of outside gas onto existing galaxies and cause collisions between two galaxies resulting in one massive end product. Despite the success of this simple hierarchical framework in explaining the history of the cosmos on scales much larger than galaxies, the model fails to fully explain the rich diversity observed in the galaxy zoo of billions.

Statistical studies of unusual galaxies hold the key to improving our understanding of the relative importance of baryonic physics and hierarchical assembly governing galaxy evolution. To this end, my team analyzes the properties and incidences of galaxy subpopulations selected from premiere galaxy surveys to provide new tests of current and future galaxy formation theories over most of the history of the universe from `cosmic high noon’ (9-11 billion years ago), when galaxy growth activity peaked, to the present epoch. The nature of these objects ensures their legacy value as primary targets for extending our understanding of galaxy growth modes when tomorrow’s great observatories come on line.

Towards Robustly Constraining the Merger Rate History of Massive Galaxies: Empirical and Theoretical Investigation of Close Pairs and Tidal Features (Ph.D. candidate: Kameswara Bharadwaj Mantha)

Current Publications: [Mantha et al.,] The process in which two or more similar-mass galaxies interact gravitationally and coalesce into one larger-mass galaxy is called as major galaxy-galaxy merging. Major merging is famously invoked by many empirical and theoretical galaxy-evolution studies to comprehend the buildup of spheroidal bulges and formation of massive elliptical galaxies, triggering of extreme star-formation and black-hole activity, and subsequent star-formation quenching. An important step towards quantifying the role of major merging in galaxy evolution is the rate at which galaxies are merging as a function of cosmic time (i.e., major merger rate). Quantifying the major merger rate of massive galaxies has been an active field. Many previous studies measured merger rates at z<1.5, while more recent studies extended to z=3, and sometimes up to z~6. Observational studies often identify major mergers to quantify the merger fraction via close-pair statistics or morphological disturbances. Both these techniques have been shown to be useful in this regard. However, there exists a critical gap in our ability to robustly convert the observed merger fraction to merger rates, owing to poorly understood merger rate conversion factors such as the observability timescale (Tobs). I am leading a coordinated effort to improve galaxy merger rate constraints by calibrating the two frequently-used merger identification techniques using the leading numerical simulations.

Why are Recently Quenched Elliptical (RQE) Galaxies Quenched? (Ph.D. student: Deepak Kumar Deo)

The pathways with which galaxies cease to actively form stars (quenching) is an open question in astrophysics. The present-day universe contains a relatively high quenched fraction and a host of physical processes are predicted to quench galaxies by either removing, heating, or cutting off the supply of cold gas necessary to fuel star formation. The specific processes responsible for quenching, their timescales, and in which cosmological environments they are most efficient, are not well constrained. Recently Quenched Elliptical galaxies (RQEs) are a subset of quenched galaxies residing at the inferred dynamical centers of moderately-low-mass dark matter halos (12 < log[Mhalo/Msun] < 13), but have a notable amount of young starlight (blue emissions) indicative of abrupt star-formation (SF) shutoff. Theory predicts such galaxies should be accreting fresh gas and actively forming stars, making them ideal laboratories to study the quenching mechanism at play. Extending the work of McIntosh et al. (2014; M14), I am working on identifying new RQEs in SDSS DR13 which would add to already discovered 172 RQEs from SDSS DR4 by M14. We are investigating the quenching mechanisms in RQEs selected from SDSS DR4 using a number of methods. We are exploring their near-UV properties to rule out very recent SF (i.e., "frosting" or "rejuvenation"). Moreover, we aim to quantify the cold gas reservoirs of the RQEs by observing their 21cm atomic hydrogen emission. We have carried out sensitivity calculations for MeerKAT 64-dish and uGMRT 30-dish array observations and found <10% of M14 RQEs are bright enough for >3-sigma detections in 10-hour integration. This motivates our ongoing efforts to identify a larger RQE sample using SDSS DR13. Together, we aim to carry out a statistically significant multi-wavelength study to piece together the puzzle: why are RQEs quenched?
I am also developing a simulator which will simulate radio observations from the SKA telescope ( This simulator will be used to plan optimal RQE observations from upcoming SKA telescope and its precursor MeerKAT. Attached plot shows the simulated uv-coverage of a modeled source at (RA, Dec.)= (4,-20) when observed at 9.2 GHz in a bandwidth of 50 MHz from SKA for 15 minutes. This 15 minutes uv-coverage is enough to recover the source brightness and morphology nearly ~ 100%.

Characterizing Residual Structure in Archival Imaging of Massive CANDELS Galaxies (MS Student: Cody Ciaschi)

Its been known that major merging is an important process in the mass build-up of galaxies, but recent theories suggest that non-merging processes might play a more important role in the mass build-up of galaxies at z>1 than previously thought. Using a sample from the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey(CANDELS) we hope to characterize the residual structure of ~10000 galaxies with 1 < z < 3 and log(Msun) > 9.7 that have been modeled using a single sersic profile (van der Wel et al. 2012). By using residual images (image-model), we hope to identify galaxies with tidal features, transient structures created by tidal forces and known products of major merging. Using the results of our classifications, we can produce a tidal based merger rate as a function of redshift, giving us insight into the relative importance of mergers in the mass-buildup of galaxies.

This project is also two-fold. Along with determining the relative importance of mergers in the mass-buildup picture of galaxy evolution we are also providing the scientific community with information about the single sersic fits as well. Many papers about galaxy evolution have been written on data from the catalog produced by van der Wel 2012. Although many of the galaxies' models were correct, some galaxies require more than a single component fit, leading to errors on certain key characteristics of a galaxy such as sersic index or effective radius. By observing the galaxies in our sample, we can look for galaxies that have not been fit well and flag them, allowing other doing analysis on the sample to not include poor fits.

Analyzing Different Residual Substructures Hosted by Massive CANDELS Galaxies (MS Student: Luther Landry)

Constraining the specific physical processes at play in the assembly of massive galaxies during the peak epoch of galaxy development (1<z<3) remains difficult. Hierarchical processes like merging and rapid gas accretion are predicted to transform the appearance of galaxies, and different evolutionary processes may produce unique but faint and transient signatures in the morphological substructure of galaxies. To facilitate the identification and further analysis of plausible hallmark indicators, we visually inspect single Sersic, model-subtracted “residual” images of high-redshift objects. We characterize the H-band (WFC3/F160W) residual images of 10,000 galaxies with log(Mstellar/Msun)>9.7 and spanning 1<z<3 in the five CANDELS fields from van der Wel et al. 2012. We classify these residual images according to two criteria: (1) the quality of the model fit to the galaxy; and (2) the qualitative nature of the residual flux, both objective (e.g., underfit centers, clumpy or patchy substructure), and subjective (e.g., plausible tidal features, spiral arms, disks). We find that 30% of these objects have no residual substructure; 5% have fit quality problems; 31% have minimal residual structure present (e.g., low surface brightness features); and, 34% have strong residual structure present. We detect plausible signs of merging for 3+/-2% of high-mass CANDELS galaxies at 1<z<2, but less at z>2. Future observations with JWST will resolve whether imaging depth is at play here. We plan to publish a value-added catalog of fit-quality and residual-substructure flags to accompany the van der Wel et al. GALFIT-derived data on MAST.

Figure Caption: Central surface mass density (Σ₅₀) vs. specific star-formation rate (log sSFR) in four redshifr bins for a high mass (log M/M_star >= 9.7), high redshift (1 < z < 3) sample of CANDELS galaxies. Overplotted is a selection of objects with H-band GALFIT residual images that have been visually characterized as having asymmetric residual flux at R>R₅₀. Orange diamonds are plausible merger events. Orange squares are plausible interactions. Open diamonds are objects with asymmetric features and no companions. Open squares are any remaining object with an asymmetric companion. Each marker is scaled by its total stellar mass.

Morphologically Disturbed Massive Galaxies: Nature and Evolution During redshifts 0.6-2.5 in the CANDELS Legacy Fields (initiated by Josh Cook, needs lead)

Merging is predicted to be an important process in the early and turbulent assembly of massive galaxies. These violent encounters heavily impact galaxy morphology and structure. As such, the evolution of morphologically disturbed systems may help constrain the relative importance of merging, the answer to which is largely debated especially at higher redshifts. Disagreements between studies however, may be attributed to the various methods used to identify merging galaxies such as visual or quantitative classifications based on different rest-frame wavelengths. Using a new comprehensive catalog of visual rest-frame optical classifications based on HST/WFC3+ACS imaging from the Cosmic Assembly Near-Infrared Deep Extragalactic Legacy Survey (CANDELS), we compare the nature and evolution of merging and highly disturbed galaxy subsamples within the UDS and GOODS-S fields. We limit our sample for completeness to high-mass objects (Mstar > 1e10 Msun) with redshifts between 0.6 < z < 2.5. Most disturbed galaxies are star-forming and two-thirds have masses under 3e10 Msun. We note that one-third appear to be neither interacting nor merging, rather they are isolated and visually disk-like. Under the assumption that many disturbed or unusual morphologies are related to merging, we compare visually-selected subsamples to merger selections based on two popular quantitative methods (Gini-M20 and CAS). We find that all selections produce similar fractions across our redshift range, but the individual galaxies making up the respective fractions are often different. This may indicate that different classification methods are preferentially selecting objects undergoing either different processes such as major merging, minor merging and violent disk instabilities, or different stages of the same process.

The Role of Spheroids in the Evolution of Massive Galaxies During redshifts 0.6-2.5 in the CANDELS UDS and GOODS-S Fields (initiated by Zach Rizer)

Spheroidal galaxies are linked to the observed buildup of massive non-star-forming (quiescent) galaxies over cosmic time. Yet, it remains unclear whether the primary growth channel involves the formation of new bulge-dominated galaxies followed by the quenching of star formation (SF), or the cessation of star production preceded by the transformation from disk-dominated to spheroidal galaxies. Using a new comprehensive catalog of visual classifications based on the HST/WFC3 imaging from the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS), we study the nature and evolution of high-mass (Mstar>1e10 Msun) 'spheroids' (elliptical and bulge-dominated galaxies) over a wide redshift range (0.61, and a possible drop to lower fractions at z<1. We find quantitatively similar results using spheroid samples defined solely or jointly by automatic (Sersic n>2) selection. We find that as the high-mass galaxy population becomes more quenched, it also becomes more dominated by spheroids with very few quiescent disks (<10%) at any redshift. Taken together, these results are consistent with a scenario in which new spheroids were continuously added and subsequently quenched, and inconsistent with an evolutionary process that primarily added newly quenched disks. The actual picture likely includes contributions from multiple channels and requires detailed modeling to better constrain the relative amounts from each.