High Redshift Galaxies

(Jayaram N. Chengalur, Nissim Kanekar, Balpreet Kaur, Yogesh Wadadekar, Rashi Jain, Pralay Biswas, Archishman Khasnovis, Former members: D. J. Saikia, J. N. H. S. Aditya, Omkar S. Bait, Preetish K. Mishra, Apurba Bera, Aditya Chowdhury, Sonalika Purkayastha)

Understanding the nature of high-redshift galaxies, as well as galaxy evolution, is an important research area at NCRA-TIFR. Astronomers here use diverse methods, in both emission and absorption, and over a wide range of observing frequencies, to probe physical conditions in high-redshift galaxies. The techniques used include neutral hydrogen (HI) 21cm absorption studies, HI 21cm emission studies of individual galaxies, "stacking" of HI 21cm emission, millimetre-wave carbon monoxide (CO) emission and absorption studies, ionized carbon (C+) emission studies, hydroxyl (OH) 18cm absorption studies, radio continuum studies, optical and ultraviolet imaging and spectroscopy, etc.. Some of the questions being addressed by NCRA-TIFR astronomers are discussed in more detail below.


Damped Lyman-alpha Absorbers (DLAs)
Damped Lyman-alpha absorbers (DLAs) are absorption-selected galaxies, identified by the presence of strong Lyman-alpha absorption in quasar spectra. Since these galaxies are detected by their absorption signatures, DLA samples contain no bias towards brighter galaxies (unlike emission-selected galaxy samples); understanding the nature of DLAs has hence long been considered an important issue in galaxy evolution. 

For DLAs towards radio-loud quasars, HI 21cm absorption studies allow one to measure the spin temperature of the absorbing gas and the cold gas content of the absorber. NCRA-TIFR astronomers have long used low-frequency radio telescopes like the Giant Metrewave Radio Telescope (GMRT) and the Green Bank Telescope (GBT) to carry out deep searches for HI 21cm absorption in DLAs, obtaining the majority of the detections of HI 21cm absorption in high-z DLAs as well as lower limits to the spin temperature in a large number of these galaxies. They have used optical spectroscopy to measure the metallicity of the absorbers, and have connected the high spin temperatures in DLAs to the paucity of metals in these galaxies. They have also extended HI 21cm absorption studies to lower redshifts, targetting strong MgII absorption systems, allowing the identification of DLAs via their HI 21cm absorption and providing the first understanding of the redshift evolution of the cold gas fraction in normal galaxies.

HI 21cm emission studies of galaxies allow one to directly measure their atomic gas mass, a critical input to describing galaxies and understanding their evolution. Follow-up interferometric mapping studies allow one to measure the velocity field of the galaxies, and infer their dynamical mass. Unfortunately, the weakness of the HI 21cm line has meant that such studies have only been possible at very low redshifts. NCRA-TIFR astronomers have led searches for HI 21cm emission in low-redshift DLAs, finding HI 21cm emission in a number of DLAs at z<0.1. They are now using the GMRT to map the HI 21cm emission in some of these galaxies.

CO emission studies of galaxies allow one to estimate their molecular gas mass, and, with follow-up mapping studies, the velocity field. NCRA-TIFR astronomers have been using such CO emission studies of high-z DLAs with the Atacama Large Millimeter/submillimeter Array (ALMA), the Very Large Array (VLA) and the NOrth European Millimetre Array (NOEMA) to measure the molecular gas masses of the DLA host galaxies. Remarkably, the ALMA and VLA studies have yielded extremely high molecular gas masses for a number of high-metallicity DLAs at both intermediate and high redshifts, raising questions on our understanding of the absorber host galaxies. 

Identifying the host galaxies of high-redshift DLAs has been an important open problem in galaxy evolution for thirty years. The difficulty lies in the fact that one has to detect a faint foreground object (the absorber host) in the presence of a far brighter object (the background quasar). NCRA-TIFR astronomers have used ALMA studies in the redshifted ionized carbon (C+) 158-micron line and the dust continuum as a new tool to identify and study the absorber host galaxies of DLAs at z~4. Initial ALMA studies have shown that the DLA hosts have relatively high star formation rates and C+ line luminosities, and large sizes.


Stacking HI 21cm emission from high-redshift galaxies
Measuring the atomic gas mass of high-z galaxies requires the detection of their HI 21cm line emission. Unfortunately, the weakness of the HI 21cm line has meant that HI 21cm emission has so far only been detected out to relatively low redshifts, z~0.38. NCRA-TIFR astronomers use "stacking" of HI 21cm emission from galaxies with known optical redshifts to extend such HI 21cm emission searches to much higher redshifts, to measure the average gas mass of galaxy samples, as well as the cosmological gas mass density in such galaxies. This is an exciting research area today at NCRA-TIFR, due to the new wideband receivers, new correlator and high sensitivity of the upgraded GMRT


Gas in AGN environments
HI 21cm absorption studies of radio-loud active galactic nuclei (AGNs) allow one to probe the presence of, and physical conditions in, neutral hydrogen in AGN environments. The detection of HI 21cm absorption from such "associated" gas allows one to measure the velocity of the gas relative to the AGN. Such studies thus also provide information on whether the gas is predominantly falling into (and thus, fuelling) the AGN, or is being driven out into the intergalactic medium by the AGN. The wide-band receivers of the upgraded GMRT will allow such associated HI 21cm absorption studies to track the redshift evolution of gas conditions in the AGN environments over a wide variety of AGN types.


Molecules in high-redshift galaxies
NCRA-TIFR astronomers use both molecular absorption and emission studies of high-redshift galaxies to obtain information on physical conditions in these objects. Such studies include searches for OH 18cm megamasers in ultra-luminous infra-red galaxies, blind surveys for CO and HCO+ absorption at high redshifts, searches for OH absorption in red quasars, searches for CO emission in Lyman-alpha emitters and Lyman-break galaxies, searches for molecular oxygen absorption in high-z galaxies, etc.

Recent Results
The HI mass function of star-forming galaxies at z~1
Neutral atomic hydrogen (HI) is the primary fuel for star formation, and thus a key baryonic constituent of galaxies. Understanding the evolution of the HI content of galaxies with cosmological time is thus critical for an understanding of galaxy evolution. A basic descriptor of the HI content of galaxies at any epoch is the "HI mass function" (HIMF), the number density of galaxies of a given HI mass as a function of the HI mass. Unfortunately, the weakness of the HI 21 cm line, the only tracer of the atomic hydrogen content of galaxies, has meant that little is known about the HIMF at cosmological distances. Chowdhury et al. used a recently-introduced approach, based on the stacking of 21cm emission signals from a large number of galaxies at z~1 to measure the average HI mass of the population, to obtain the first estimate of the HIMF of star-forming galaxies at a redshift of 1, approximately 8 billion years ago. These authors obtained the HIMF at z~1 by combining (i) their measurement of the average HI mass of star-forming galaxies at z~1 as a function of the blue-band luminosity, and (ii) a literature estimate of the number density of galaxies at the same epoch, also as a function of the blue-band luminosity. The dependence of the average HI mass of galaxies on the blue-band luminosity was obtained by stacking the 21cm signals from galaxies in different bins of blue-band luminosity, to measure the average HI mass in each luminosity bin. The adjacent figure shows the authors' measurement of the HIMF of star-forming galaxies at z~1 (blue line) and the HIMF in the local Universe (black dashed line); the shaded blue region shows the uncertainty in the estimate of the HIMF at z~1. It is clear from the figure that the number density of galaxies with HI masses greater than 10 billion solar masses is far greater at z~1 than in the local Universe. Indeed, Chowdhury et al. find that such massive galaxies were roughly 4-5 more numerous in the early Universe, 8 billion years ago, than in the Universe today. This paper thus provides the first statistically significant evidence for evolution in the HIMF of galaxies from the epoch of cosmic noon.
The Gas Accretion Rate of Galaxies over z~0-1.3
A galaxy’s evolution is driven by processes that regulate its key baryonic constituents, the neutral atomic gas (HI), the molecular gas (H2), and the stars. These processes can be quantified by three key rates: (i) the net rate of accretion of HI from the circumgalactic medium (CGM) that surrounds the disk of the galaxy, (ii) the formation rate of H2 from HI, and (iii) the star-formation rate (SFR). While the evolution of the SFR density of the Universe with cosmological time has been known for over two decades now, the difficulty of measuring the HI content in distant galaxies has meant that there is so far no estimate of the gas accretion rate or its evolution at z~1. Chowdhury et al. used their recent measurement of the HI content of galaxies at z~1 (8 Gyr ago) from the GMRT- CATz1 survey to estimate the average rates of accretion of HI onto galaxies at two key epochs in galaxy evolution: (i) z ~ 1.3–1.0 (9 to 8 Gyr ago), toward the end of the epoch of peak star formation activity in the Universe, and (ii) z ~ 1–0 (8 Gyr to today), when the star formation activity of the Universe declines by an order of magnitude. The figure shows, for galaxies at the earlier epoch (z~1.3–1.0), as a function of their stellar masses, the time-averaged net gas accretion rate (in green), the H2 formation rate (in orange), and the average star-formation rate (in blue). The figure demonstrates that, for galaxies at the early epoch, between 8 to 9 Gyr ago, the average gas accretion rate is far lower than the average SFR, but the H2 formation rate is comparable to the average SFR. Conversely, the authors find that, at later times (8 Gyr ago to today, not shown in the figure), both the accretion rate and the H2 formation rate are significantly lower than the average SFR. Chowdhury et al. also show that massive galaxies had already acquired most of their present-day baryonic mass 9 Gyr ago. Overall, the results show that the rapid conversion of the existing atomic gas reservoir to molecular gas was sufficient to maintain a high average SFR in galaxies 8-9 Gyr ago, despite the low net gas accretion rate. However, at later times, the combination of the lower net gas accretion rate and the lower H2 formation rate leads to a decline in the fuel available for star formation and results in the observed decrease in the SFR density of the Universe over the last 8 Gyr.
The Gas Accretion Rate of Star-forming Galaxies over the last 4 Gyr
Star-forming galaxies are believed to replenish their atomic gas reservoir, which is consumed in star-formation, through accretion of gas from the circumgalactic medium (CGM). However, there are few observational constraints today on the gas accretion rate in external galaxies. Bera et al. used the recent upgraded GMRT measurement of the scaling relation between the atomic hydrogen (HI) mass and the stellar mass in star-forming galaxies at z~0.35, with the relations between the star-formation rate and stellar mass, and between the molecular gas mass and stellar mass, to determine the evolution of the neutral gas reservoir and the average net gas accretion rate onto the disks of star-forming galaxies over the past 4 Gyr. They found that for galaxies with present day stellar masses exceeding a billion solar mass, both stellar mass and HI mass in the disk have increased, while the molecular gas mass has decreased, since z~0.35. The average gas accretion rate onto the disk over the past 4 Gyr is similar to the average star-formation rate over this period, implying that star-forming galaxies have maintained a stable atomic gas reservoir, despite the consumption of gas in star-formation. The figure shows the estimates of the average star-formation rate (red line), the average net gas accretion rate (black line) and the average net molecular gas formation rate (blue line) of star-forming galaxies over the past 4 Gyr against their present day stellar mass. Bera et al. also estimated an average net gas accretion rate (over the past 4 Gyr) of about 6 solar masses per year for galaxies with the stellar mass of the Milky Way. They concluded that at low redshifts, z<~0.4, the reason for the decline in the cosmic star-formation rate density is likely to be the inefficiency in the conversion of atomic gas to molecular gas, rather than insufficient gas accretion from the CGM.
Atomic hydrogen scaling relations at z~0.35
The atomic hydrogen (HI) properties of star-forming galaxies in the local Universe are known to correlate with other galaxy properties via the “HI scaling relations”. The redshift evolution of these relations serves as an important constraint on models of galaxy evolution. However, until recently, there were no estimates of the HI scaling relations at earlier epochs. Bera et al. used data from a deep upgraded GMRT HI 21 cm survey of the Extended Groth Strip, and the technique of spectral line stacking, to measure the scaling relation between the HI mass and the stellar mass for star-forming galaxies at redshift z ≈ 0.35. Using this measurement, along with the main-sequence relation between the stellar mass and star-formation rate of galaxies, they inferred the HI depletion timescale of star-forming galaxies as a function of their stellar mass, which is shown in the lower panel of the figure. They found that massive star-forming galaxies at z ≈ 0.35 are HI-poor compared to local star-forming galaxies of a similar stellar mass. However, their characteristic HI depletion time is shorter by a factor of ≈ 5 than that of their local analogues, indicating a higher star-formation efficiency at intermediate redshifts. The short characteristic HI depletion timescales, ≲ 3 Gyr, of massive star-forming galaxies at z ≈ 0.35 indicate that they must have acquired a significant amount of neutral gas through accretion from the circumgalactic medium, over the past 4 Gyr, to avoid quenching of their star-formation activity. In the attached figure, the upper panel shows the scaling relation between the HI mass and the stellar mass, while the lower panel shows the scaling relation between the HI depletion timescale and the stellar mass. The blue circles in both panels show the measurements of Bera et al., while the solid blue lines and blue shaded regions show their estimates of the scaling relations. The other three curves in each panel show the corresponding scaling relations (1) in the local Universe (black curve and grey shaded region), from the xGASS survey, (2) at z~0.37, from the MIGHTEE-HI survey (magenta dash-dotted curve and pink shaded region), and (3) at z~1.0 from the CATz1 survey (red dotted curve and orange shaded region).
The HI mass function of star-forming galaxies at z~0.35
The neutral atomic hydrogen (HI) mass function describes the distribution of the HI content of galaxies at any epoch; its evolution provides an important probe of models of galaxy formation and evolution. Unfortunately, the weakness of the HI 21cm line has meant that it has hitherto not been possible to determine the HI mass function of galaxies at cosmological distances. While measuring the HI masses of a large number of galaxies at intermediate redshifts remains challenging today, it is possible to stack the HI 21cm spectra of individual galaxies and measure the average HI mass of the population. Further, stacking the HI 21cm spectra of galaxies as a function of their optical luminosities can be used to obtain the dependence of the average HI mass on the galaxy luminosity. This can then be combined with the optical luminosity function to infer the HI mass function. This interesting approach was used by Bera et al. to obtain the first estimate of the HI mass function at intermediate redshifts: they used Giant Metrewave Radio Telescope HI 21cm spectroscopy of blue star-forming galaxies in the Extended Groth Strip to determine the scaling relation between the average HI mass (M_HI) and the absolute B-band magnitude (M_B) of such galaxies at z~0.35, by stacking the HI 21cm emission signals of galaxy subsamples in different M_B ranges. This M_HI-M_B scaling relation at z~0.35 is shown in blue in the top panel of the figure, with the corresponding relation in the local Universe shown as the dashed red line. They then combined this M_HI-M_B scaling relation with the known B-band luminosity function of star-forming galaxies at these redshifts to determine the HI mass function at z~0.35. They also demonstrated that the use of the correct scatter in the M_HI-M_B relation is critical for an accurate estimate of the mass function; their estimate of the mass function at z~0.35 assumes that the scatter in the relation at this redshift is the same as that in the local Universe. Bera et al. found that the HI mass function has evolved significantly from z~0.35 to today, i.e. over the last four billion years, especially at the high-mass end (this can be seen clearly in the bottom left panel of the figure). High-mass galaxies, with HI masses larger than roughly 10 billion solar masses, are a factor of ~3.4 less prevalent at z~0.35 than at z~0 (as can be seen in the bottom right panel of the figure). Conversely, there are more low-mass galaxies, with HI masses of roughly a billion solar masses, at z~0.35 than in the local Universe. These results suggest that massive star-forming galaxies have acquired a significant amount of neutral atomic gas through mergers or accretion from the circumgalactic medium over the past four billion years.
Atomic Gas Scaling Relations of Star-forming Galaxies at z~1
Gas and stars are the key baryonic constituents of galaxies with neutral atomic hydrogen gas (HI) being the primary fuel for star formation. In the nearby Universe, the HI properties of galaxies have been found to correlate with their various other galaxy properties through the ``HI scaling relations'', essentially relations between the HI mass of the galaxies and the stellar mass, luminosity, size, etc. The scaling relations quantify the connections between gas and stellar properties of galaxies, and thus contain information about the balances between the complex processes underlying galaxy evolution. The existence and the redshift evolution of such scaling relations provide a critical constraint on models of galaxy evolution. While detailed HI 21cm studies of nearby galaxies have yielded accurate determinations of the HI scaling relations in the local Universe, the weakness of the HI 21cm line has meant that it has not so far been possible to determine these relations at cosmological distances. Chowdhury et al. used the Giant Metrewave Radio Telescope (GMRT) Cold-HI AT z~1 (CATz1) survey, a 510 hr HI 21 cm emission survey of galaxies at z = 0.74-1.45, to report the first measurements of the HI scaling relations in star-forming galaxies at z~1, nine billion years ago. The authors divided their sample of ~11,500 galaxies at z~1 into three subsamples with different stellar mass ranges, to measure the average HI masses of galaxies with different average stellar masses. Chowdhury et al. find that the relation between HI and stellar mass at z~1 has the same slope as in the local Universe, but is a factor of ~3.5 higher in normalization. This implies that the average HI masses of galaxies over a wide range of stellar mass are higher by this factor than those of nearby galaxies with similar stellar masses. The authors also measured the relation between the HI depletion timescale (the timescale on which the HI in the galaxy would be entirely converted to stars, at the current star formation rate) and the stellar mass, finding that this relation lies a factor of 2-4 lower than the corresponding relation in the nearby Universe. Chowdhury et al. also find that the efficiency with which HI is converted to stars is much higher for galaxies at z~1 than for those in the nearby Universe. The figure shows [A] the average HI mass and [B] the average HI depletion timescale of galaxies, as a function of the stellar mass, in the nearby Universe (blue points) and at z~1 (red points).
The Giant Metrewave Radio Telescope Cold-HI AT z ~ 1 Survey
Neutral atomic hydrogen (HI) is the primary fuel for star formation in galaxies. An understanding of galaxy evolution thus critically requires measurements of the atomic gas mass of galaxies over cosmological time. Unfortunately, the weakness of the HI 21 cm line, the only tracer of the HI mass of galaxies, has meant that, until very recently, nothing was known about the HI mass of high-redshift galaxies. Chowdhury et al. had used the upgraded Giant Metrewave Radio Telescope (GMRT) in 2020 to obtain the first measurement of the average HI mass of galaxies at z~1, nine billion years ago. The team has now followed this up with a much larger survey, the GMRT Cold-HI AT z~1 (CATz1) survey, a 510 hr upgraded GMRT survey aimed at characterizing HI in galaxies during and just after the epoch of peak star formation activity in the universe (often referred to as ``the epoch of galaxy assembly''), a key epoch in galaxy evolution. In the current paper, Chowdhury et al. describe the design, data analysis, and basic results of the GMRT-CATz1 survey. They combined (``stacked'') the HI 21 cm emission signals of ~11,500 star-forming galaxies at z=0.74-1.45 to obtain a high (7.1-sigma) significance detection of the average HI 21 cm signal from the sample of galaxies. The detected HI 21 cm signal can be clearly seen in the adjacent figure in both [A] the stacked HI 21 cm image, and [B] the stacked HI 21 cm spectrum. The average HI mass of the galaxies inferred from the detected signal is a factor of 1.4 higher than the average stellar mass of the galaxies, and a factor of ~3.5 higher than the HI mass of galaxies with similar stellar masses in the local Universe. However, Chowdhury et al. find that even such a large HI reservoir will be able to sustain the high star-formation rate of these galaxies for only a short duration, ~1.7 Gyr. Chowdhury et al. thus find that although galaxies at z ~ 1 have a high HI mass, their short HI depletion timescale is likely to cause quenching of their star formation activity in the absence of rapid accretion of gas from the environment around the galaxies. The GMRT-CATz1 survey will allow detailed studies of the HI properties of high-redshift galaxies, providing, for the first time, an understanding of atomic gas in galaxies during and just after the epoch of galaxy assembly. A set of companion papers by Chowdhury et al. has already yielded exciting new insights on these issues.
Atomic Gas Dominates the Baryonic Mass of Star-forming Galaxies at z ~ 1.3
Ordinary ``baryonic'' matter in galaxies is mostly in the form of stars and neutral atomic and molecular gas. Over the lifetime of a galaxy, neutral atomic gas gets converted to molecular gas which in turn gets converted to stars. A galaxy's baryonic composition is thus one of its fundamental properties, and an important indicator of its evolutionary stage. The distribution of the baryonic mass of galaxies in the early Universe between stars, atomic and molecular gas, has hence long been an open problem in galaxy evolution. Unfortunately, the weakness of the HI 21 cm line, the only direct tracer of the atomic gas mass of galaxies, has meant that, until very recently, the atomic gas masses of high-redshift galaxies were not known. Chowdhury et al. used their recent detection of the average HI 21 cm emission signals from a large sample of star-forming galaxies at z~1.0 and at z~1.3 to find that high-redshift galaxies, at the epoch of peak star-formation activity in the Universe, have a dramatically different baryonic composition from that of nearby galaxies. The adjacent figure shows the fraction of the average baryonic mass of galaxies in atomic gas (red), molecular gas (blue), and stars (yellow) at z~0, z~1.0, and z~1.3; the samples of galaxies at the three epochs have identical stellar mass distributions. The figure shows that the contribution of stars to the total baryonic mass has increased from approximately 16% at z~1.3 to roughly 60% in the nearby Universe. Conversely, the fraction of mass in molecular gas for such galaxies has declined from about 14% at z~1.3 to only 6% in the nearby Universe. Remarkably, Chowdhury et al. find that atomic gas dominates the baryonic mass of galaxies at z~1.3, with roughly 70% of the total baryonic mass in atomic gas, and only around 16% in stars. Overall, the study provides evidence for strong evolution in the baryonic composition of galaxies over the past 9 billion years, with early galaxies, at the peak of star-formation activity in the Universe, being predominantly made up of neutral gas.
Jansky Very Large Array Detections of CO(1-0) Emission in H I-absorption-selected Galaxies at z>=2
Our current understanding of galaxies and galaxy evolution is based on studies of emission-selected galaxy samples. Such samples contain a luminosity bias, i.e. they are biased towards brighter galaxies, especially at high redshifts. However, it is possible to also identify high-redshift galaxies via their Ly-alpha absorption signature in the spectra of background quasars; such ``HI-absorption-selected galaxies'' do not have the above luminosity bias. In order to understand HI-absorption-selected galaxies and their connections to the emission-selected population, we must first detect them in emission and then characterise their stellar and gas properties. Kaur et al. used the Jansky Very Large Array to search for CO(1-0) emission from three such HI-absorption-selected galaxies, at z~2, that were earlier identified with ALMA using CO(3-2) or CO(4-3) emission. They detected CO(1-0) emission from two HI-selected galaxies, DLA0918+1636g at z=2.5832 and DLA1228-113g at z=2.1933; these are the first detections of CO(1-0) emission in high-z HI-absorption-selected galaxies. Kaur et al. infer molecular gas masses for the two detected galaxies that are ~1.5-2 times lower than earlier estimates based on the mid-J CO lines. Kaur et al. also used the JVLA data to determine the CO spectral line energy distribution in the three galaxies. They find that the J=3 and J=4 levels are thermally excited in DLA0918+1636g, while the data are consistent with thermal excitation of the J=3 level in DLA1228-113g. In the case of the third galaxy, DLA0551-366g, no CO(1-0) emission was detected, yielding lower limits on the excitation of the J=4, J=5, and J=6 levels. Kaur et al. also compared the CO excitation of the HI-selected galaxies with that of the Milky Way, main-sequence galaxies, and sub-mm galaxies at high redshifts. They find that the CO excitation in HI-selected galaxies is similar to that of massive main-sequence galaxies at z>2, but higher than that of main-sequence galaxies at z~1.5. They also compared the ratio of CO(3-2) and CO(1-0) line luminosities (r_31; see attached figure) in different types of galaxies, and find that all galaxies at z>2 have both a higher r_31 value (implying a higher excitation of the J=3 level) and a higher SFR surface density than the main-sequence galaxies at z~1.5. The higher CO excitation in galaxies at z>2 may thus arise due to their higher SFR surface densities, as suggested by earlier theoretical studies.
A Green Pea Starburst arising from a Galaxy-Galaxy merger
Faint star-forming dwarf galaxies have long been believed to be the main contributors of the Lyman-continuum (LyC) photons that ionized the early universe. Green Pea galaxies are low-redshift starburst dwarf galaxies with properties similar to those of the high-redshift galaxies, making them important proxies to understand how ionizing radiation escapes the high-redshift galaxies. Purkayastha et. al. have used the Giant Metrewave Radio Telescope (GMRT) to carry out the first mapping of the spatial distribution of atomic hydrogen (HI) in and around a Green Pea, GP J0213+0056 at z=0.0399. GP J0213+0056 shows both strong HI 21cm emission in single-dish spectroscopy and strong Lyman-alpha emission. This leads to a tension between requiring sufficient neutral hydrogen to fuel the starburst but sufficiently low HI column density to allow the Lyman-alpha emission to escape. The figure shows the GMRT image of the HI 21 cm emission around GP J0213+0056 (blue contours) overlaid on a Subaru HyperSuprem-Cam i-band image. The GMRT images are at resolutions of 16, 12, 9 and 7 arcseconds (panels [A] to [D]). The HI 21cm emission is seen to arise from an extended region around the Green Pea and a companion galaxy (G1), roughly 4.7 kpc from the Green Pea, in a broken ring-like structure. The strongest emission arises from neither the Green Pea nor G1, but from the region around them. The high-resolution images in panel [C] and [D] show that the highest HI coloumn density is seen west of G1, with little emission seen at the location of the Green Pea itself. The HI 21cm images indicate that the starburst in GP J0213+0056 is likely to have been triggered by a major merger with the companion galaxy G1, leading to a disturbed HI spatial and velocity distribution, which in turn allowed Lyman-alpha (and, possibly, Lyman-continuum) emission to leak from the Green Pea. Such mergers, and the resulting holes in the HI distribution, are a natural way to explain the tension between the requirements of cold gas to fuel the starburst and the observed leakage of Lyman-alpha and Lyman-continuum emission in Green Pea galaxies and their high-redshift counterparts.
Insufficient Gas Accretion Caused the Decline in Cosmic Star-Formation Activity 8 Billion Years Ago
The cause of the decline in the cosmic star-formation rate (SFR) density of the Universe after its peak of approximately 8-11 billion years ago (in the redshift range z ~ 1-3) is a key open issue in galaxy evolution. Addressing this requires us to understand the evolution of the gas mass of galaxies, the fuel from which the stars form. The primary fuel for star formation is neutral atomic hydrogen (HI). The HI content of galaxies can be inferred from the strength of their HI 21 cm emission; however, this hyperfine transition is very weak and difficult to detect from individual galaxies at cosmological distances. Chowdhury et al. used the Giant Metrewave Radio Telescope (GMRT) Cold HI AT z~1 (CATz1) survey to report a measurement of the dependence of the average HI mass of a large sample of star-forming galaxies at redshifts z=0.74-1.45 on their average stellar mass and redshift, by stacking the HI 21 cm emission signals of the individual galaxies. They find that galaxies with stellar masses greater than approximately 10 billion solar masses, which dominate the decline in the cosmic SFR density at z<~1, have HI reservoirs that can sustain their SFRs for only a short period, ~0.9 billion years unless the HI is replenished by accretion of gas from the circumgalactic medium. Remarkably, they also measure a steep decline, by a factor of ~3.2, in the average HI mass of star-forming galaxies, over a period of roughly 1 billion years between z~1.3 and z~1.0. Panel~[A] of the figure on the right shows the stacked HI 21 cm emission signal at z~1.3 (orange) and z~1.0 (blue); the average HI 21 cm signal is clearly detected from both subsamples, with the emission signal at z~1.3  being much stronger than that at z~1.0. Panel~[B] of the figure shows the redshift evolution of the average HI mass of star-forming galaxies over the last 9 billion years; the sharp decline in the average HI mass of galaxies from z~1.3 to z~1.0 is again clearly seen. The observed decline in the average HI mass of star-forming galaxies provides direct evidence that the accretion of HI onto massive star-forming galaxies at z~1 is insufficient to replenish their HI reservoirs on the short timescale required to sustain their SFRs. The results of this study indicate that the decline in the cosmic SFR density at z~1 arises due to the decline in the HI mass of the most massive star-forming galaxies, due to insufficient gas accretion from their surroundings.
The Nature of HI-absorption-selected Galaxies at z~4
Galaxy populations selected based on luminosity in deep images are biased towards the brighter systems at high redshifts. This luminosity bias can be avoided by identifying high-redshift galaxies via their strong Lyman-alpha absorption signature in the spectra of background quasars. Unfortunately, it has been difficult to identify and characterize such absorption-selected galaxies using the standard optical techniques as the galaxies are much fainter than the background quasars at optical wavelengths. Neeleman et al. (2017, 2019) used the Atacama Large Millimeter/submillimeter Array to pioneer a new approach to identify absorption-selected galaxies at z~4, via their [CII] 158 micron emission. In this paper, Kaur et al. report Jansky Very Large Array (JVLA) and Hubble Space Telescope Wide Field Camera 3 (HST-WFC3) observations of seven [CII] emitters at z~4, aiming to characterize their molecular gas content and star-formation activity. No CO emission was detected from the seven absorption-selected galaxies, yielding upper limits on their molecular gas mass. Rest-frame near-ultraviolet (NUV) emission was detected from four systems, giving an estimate of the star-formation rate (SFR) unobscured by dust for these galaxies. Comparing the dust-unobscured SFR with the total SFR estimated from the 160-micron dust continuum emission, Kaur et al. find that most absorption-selected galaxies do not contain significant amounts of dust. They find that the molecular gas mass estimates and NUV SFR estimates in HI-selected galaxies at z~4 are consistent with those of main-sequence galaxies with similar [CII] and far-infrared luminosities at similar redshifts. The figure shows the HST-WFC3 rest-frame NUV images (in colour) of the seven absorption-selected galaxies overlaid with the [CII] emission (in red contours).
GMRT Detection of HI 21 cm Emission from Star-forming Galaxies at z ~ 1.3
Neutral atomic hydrogen (HI) is a key constituent of galaxies and is the primary fuel for star formation. Therefore, an understanding of galaxy evolution requires measurements of the HI content of galaxies at different cosmological epochs, to probe how the typical HI mass of galaxies changes with time. Unfortunately, the main tracer of HI in galaxies, the hyperfine spectral line at a wavelength of 21.1 cm, referred to as the ``HI 21cm'' line, is a very weak spectral line. This makes it very difficult to measure the HI mass of high-redshift galaxies with current radio telescopes, which has severely limited our understanding of critical issues in galaxy evolution. For example, the cosmological star-formation rate density of the Universe is observed to peak in the redshift range z~1-3 (approximately 8-11 billion years ago) and to then decline by a factor of ten to its current value in the local Universe. The cause of the decline is an important open question in galaxy evolution. Chowdhury et al. used approximately 400 hrs of GMRT observations to obtain a detection of the average HI 21cm emission signal from ~2800 star-forming galaxies at z~1.3. Panels [A] and [B] of the figure show the average HI 21cm emission spectrum and the average HI 21cm image, respectively; a detection can be clearly seen in both panels. This is the highest redshift at which the HI 21cm signal has so far been detected, coming from galaxies 9 billion years ago. The authors used the detection of the average HI 21cm emission to estimate the average HI mass of star-forming galaxies at z~1.3: they find that the average HI mass of galaxies at this epoch is roughly 2.5 times higher than the average mass in stars. This is very different from galaxies in the local Universe where the HI mass is typically less than half the stellar mass. However, the high-z galaxies also have very high star-formation rates; the authors combine the star-formation rates with the measured average HI mass to find that the atomic gas can fuel the star-formation activity for only around 2 billion years, without replenshment of the gas reservoir. This is much shorter than the timescale on which HI is consumed by galaxies in the local Universe. This indicates that a lack of HI fuel to maintain the high star-formation rate of galaxies at these redshifts is the likely cause of the observed decline in the cosmic star-formation activity at z<1. The new results extend to higher redshifts the group’s earlier detection of the average HI 21cm signal, from galaxies at z~1.0, i.e. roughly 8 billion years ago. Also, the two studies were carried out with different receivers and electronics signal chain: the current result used the original GMRT receivers and electronics, while Chowdhury et al. (2020) used the upgraded GMRT receivers and electronics. The new results are thus an important independent confirmation of the results of the earlier study.
The Atomic Gas Mass of Green Pea Galaxies
Green Pea galaxies are extreme emission-line galaxies at low redshift, with low metallicity and dust content, strong nebular lines, compact or interacting morphology, and intense star formation activity, and which often show leakage of Lyman-continuum radiation. Green Peas are believed to be the best local analogs of the galaxies that drove cosmological reionization at z>6, and offer the exciting possibility of understanding conditions in the high-redshift galaxies by detailed studies of nearby objects. However, while detailed optical and UV imaging and spectroscopic studies have characterized the stellar, nebular and star-formation properties of Green Peas, little was hitherto known about the primary fuel for star-formation in these galaxies, the neutral atomic hydrogen (HI). As such, the cause of the intense starburst activity in the Green Peas was unclear. Kanekar et al. used the Arecibo Telescope and the Green Bank Telescope to carry out a deep search for HI 21cm emission from a large sample of Green peas, obtaining detections of HI 21cm detections and estimates of the HI mass in 19 galaxies, and strong upper limits on the HI mass in 21 systems. These are the first estimates of the atomic gas content of Green Pea galaxies. Kanekar et al. find that the HI-to-stellar mass ratio in Green Peas is consistent with trends identified in star-forming galaxies in the local Universe. However, the median HI depletion timescale in Green Peas is more than ten times lower than that obtained in local star-forming galaxies. This implies that Green Peas consume their atomic gas on very short timescales. Kanekar et al. also find evidence of bimodality in the Green Pea sample, with many Green Peas appearing gas-rich, suggesting recent gas accretion, and others appearing gas-poor, suggesting that all their atomic hydrogen has been eaten by star-formation. The left panel of the figure shows the HI mass of the Green Peas plotted against their optical B-band absolute magnitude (equivalent to their B-band luminosity); the solid line shows the relation between HI mass and B-band magnitude seen in normal galaxies in the local Universe, with the two dashed lines showing the spread around the relation. A number of Green Peas are seen to lie above and below the spread in the relation, indicating that some Green Peas have a higher HI mass than expected (i.e. are gas-rich), while others have a lower HI mass than expected (i.e. are gas-poor). The right panel shows the timescale on which the HI in the Green Peas would be consumed by star formation (green circles) plotted against their stellar mass. The dashed green line shows the median HI depletion timescale in the Green Pea sample, approximately 600 million years. The dashed blue line shows the median HI depletion timescale for the xGASS sample of normal nearby galaxies; this is seen to be a factor of 10 higher than that in the Green Peas. Note that both plots are in logarithmic units: a change by 1 unit corresponds to a factor of 10!
HI 21-centimetre emission from an ensemble of galaxies at an average redshift of one
Baryonic processes in galaxy evolution include the infall of gas onto galaxies to form neutral atomic hydrogen (HI), which is then converted to the molecular state (H2), and, finally, the conversion of H2 to stars. Understanding galaxy evolution thus requires an understanding of the evolution of stars and of neutral atomic and molecular hydrogen. For the stars, the cosmic star-formation rate density is known to peak at redshifts between 1 and 3, and to decline by an order of magnitude over the subsequent 8 billion years; the causes of this decline are not known. For the gas, the weakness of the hyperfine transition of HI at 21 cm wavelength — the main tracer of the HI content of galaxies—means that it has not hitherto been possible to measure the atomic gas mass of galaxies at redshifts higher than about 0.4; this is a critical gap in our understanding of galaxy evolution. Chowdhury et. al. report a measurement of the average HI mass of star-forming galaxies at a redshift of about one, obtained by stacking the HI 21 cm emission signals from 7,653 galaxies over a 1.2 square degree region of the sky. The figure shows [A] the stacked HI 21 cm emission map and [B] the stacked HI 21 cm emission spectrum; the detection of the average 21cm emission signal can be clearly seen in both panels. The measured average HI mass of the sample of galaxies at z~1 is similar to the average stellar mass of the sample but the HI mass can fuel the observed star-formation rates for only 1 to 2 billion years in the absence of fresh gas infall. This suggests that gas accretion onto galaxies at redshifts of less than one may have been insufficient to sustain high star-formation rates in star-forming galaxies. This is likely to be the cause of the decline in the cosmic star-formation rate density at redshifts below one.
High Molecular Gas Masses in Absorption-selected Galaxies at z ~ 2
Kanekar et al. used the Atacama Large Millimeter/submillimeter Array (ALMA) to carry out a search for CO (3-2) or (4-3) emission from the fields of 12 high-metallicity damped Lyman-alpha absorbers (DLAs) at z~1.7-2.6. They detected CO emission from galaxies in the fields of five DLAs, obtaining high molecular gas masses, in the range (13 - 210) billion solar masses. The impact parameters of the CO emitters to the QSO sightline lie in the range 5.6-100 kpc, with the three new CO detections having impact parameters <~ 15 kpc. The highest CO line luminosities and inferred molecular gas masses are associated with the highest-metallicity DLAs, with metallicities within a factor of 2 of the solar metallicity. The high inferred molecular gas masses may be explained by a combination of a stellar mass-metallicity relation and a high molecular gas-to-stars mass ratio in high-redshift galaxies; the DLA galaxies identified by the authors' CO searches have properties consistent with those of emission-selected samples. None of the DLA galaxies detected in CO emission were identified in earlier optical or near-IR searches and vice-versa; DLA galaxies earlier identified in optical/near-IR searches were not detected in CO emission. The high ALMA CO and [CII] 158-micron line detection rate in high-redshift, high-metallicity DLA galaxies has revolutionized the field, allowing the identification of dusty, massive galaxies associated with high-redshift DLAs. The HI-absorption criterion identifying DLAs selects the entire high-redshift galaxy population, including dusty and UV-bright galaxies, in a wide range of environments. The left panel of the figure shows the CO line luminosity (in logarithmic units) plotted against the absorber metallicity; the higher CO line luminosity at [M/H]>= -0.3 dex is clear. The right panel plots metallicity against stellar mass (assumed to be equal to the molecular gas mass), with CO detections shown as filled blue circles and CO non-detections as open blue circles. The filled black squares show the (binned) emission metallicity plotted against the (binned) stellar mass for the UV-selected galaxies of Erb et al. (2006), while the dashed red curve shows the mass-metallicity relation of these galaxies. Three DLA galaxies identified via optical spectroscopy are shown as red stars, with stellar mass estimates from the optical/near-IR photometry.
High-opacity associated HI 21cm absorbers at z~1.2
HI 21cm absorption arising from neutral hydrogen in the environments of Active Galactic Nuclei (AGNs) can be used to probe physical conditions in the AGN’s vicinity and how these conditions change over time. However, despite a large number of searches spanning many decades, only 7 such ``associated'' HI 21cm absorbers have been detected at redshifts greater than one. Chowdhury et al. used the new wide-band GMRT Band-4 receivers to discover two remarkable HI 21cm absorbers at a high redshift, z~1.2, against faint AGNs. The two absorbers were found in an unbiased search for HI 21cm absorption against all radio-continuum sources over a 1.2 square degree region of the sky, the first time that such a search has been carried out with a high sensitivity at high redshifts. The velocity-integrated HI 21cm optical depths of the two systems, shown in the figure, are greater than those of any known associated HI 21cm absorbers, and the two AGNs are very faint in both the radio and the ultraviolet wavebands. The discovery of these two systems is consistent with an earlier hypothesis that the dearth of associated HI 21cm absorbers at high redshifts, z>1, may be due to an observational bias wherein high-redshift AGNs targeted in surveys for associated HI 21cm absorbers are bright in the ultraviolet and radio wavelengths. The high AGN radio and/or ultraviolet luminosity may ionize or excite the HI in its vicinity, reducing the strength of the HI 21cm absorption. The two new HI 21cm absorbers emphasize the need to carry out unbiased HI 21cm absorption surveys and to extend future searches to low luminosity AGNs.
ALMA [CII] 158um imaging of an HI-Selected Major Merger at z~4
Prochaska et al. used the Atacama Large Millimeter/submillimeter Array (ALMA) to obtain high spatial resolution (~ 2 kpc) observations of [CII] 158-micron and dust-continuum emission from a galaxy at z=3.7978 selected by its strong Lyman-alpha absorption (a damped Ly-alpha absorber, DLA) against a background QSO. Their ALMA images reveal a pair of star-forming galaxies separated by approximately 6 kpc (projected) undergoing a major merger. Between these galaxies is a third emission component with highly elevated [CII] 158-micron A emission relative to the dust continuum (by a factor of ~2), which is likely to arise from stripped gas associated with the merger. This merger of two otherwise-normal galaxies is not accompanied by enhanced star-formation, contrary to mergers detected in most luminosity-selected samples. The DLA associated with the merger exhibits extreme kinematics, with a velocity width for the low-ionization metal lines of roughly 470 km/s, that spans the velocity spread revealed in the [CII] 158-micron emission. The authors propose that DLAs with high low-ionization metal line widths are a signpost of major mergers in normal galaxies at high redshifts, and use the distribution of the velocity widths of metal lines in high-z DLAs to provide a rough estimate of the fraction of z>3 galaxies that are undergoing a major merger. Figure: (Left) The zeroth-moment image showing the integrated [CII] 158-micron flux density of the galaxies A and B associated with the absorption-selected galaxy. The axes are labeled in physical units (kpc) at z=3.7978. The lowest contour corresponds to 3-sigma significance, with the contours increasing by a factor of sqrt(2). (right) The first-moment image showing the flux density-weighted velocity field, restricted to regions where the integrated flux density exceeds 2.5-sigma significance. There are two kinematically distinct components which we associate with a pair of merging galaxies, labeled A and B. There is additional emission between these components, referred to as component C, which appears to be gas stripped during the merger.
A cold, massive, rotating disk galaxy 1.5 billion years after the Big Bang
Massive disk galaxies like the Milky Way are expected to form at late times in traditional models of galaxy formation, but recent numerical simulations suggest that such galaxies could form as early as a billion years after the Big Bang through the accretion of cold material and mergers. Observationally, it has been difficult to identify disk galaxies in emission at high redshift in order to discern between competing models of galaxy formation. Neeleman et al. report Atacama Large Millimeter/submillimeter Array (ALMA) imaging, with a resolution of about 1.3 kiloparsecs, of the 158-micrometre emission line from singly ionized carbon ([CII]), the far-infrared dust continuum and the near-ultraviolet continuum emission from a galaxy at a redshift of 4.2603, identified by detecting its absorption of quasar light. The ALMA observations show that the emission arises from gas inside a cold, dusty, rotating disk with a rotational velocity of about 272 kilometres per second. The detection of emission from carbon monoxide in the galaxy yields a molecular mass that is consistent with the estimate from the ionized carbon emission of about 72 billion solar masses. The existence of such a massive, rotationally supported, cold disk galaxy when the Universe was only 1.5 billion years old favours formation through either cold-mode accretion or mergers, although its large rotational velocity and large content of cold gas remain challenging to reproduce with most numerical simulations. Figure: The top row shows the velocity-integrated [CII] flux density for the data (left panel), the constant rotational velocity model (middle panel) and the residual after subtracting the model from the data (right panel). The outer contour is at 3 sigma, where sigma is the standard deviation of the noise in the observations, with contours increasing in powers of sqrt(2). No negative contours at the same levels are observed in the image. The synthesized beam of the observations is shown in the bottom left corner of the leftmost plot. The bottom row shows the mean velocity of the [CII] emission, for the data (left panel), the model (middle) and the residuals (right). Velocities are relative to the systemic velocity of the [CII] emission, corresponding to z = 4.2603.
Atomic hydrogen in star-forming galaxies at intermediate redshifts
Bera et al. used the upgraded Giant Metrewave Radio Telescope to carry out a deep observation of one of the well-known optical deep fields, the Extended Groth Strip, (EGS) covering the frequency range 1000-1370 MHz. This enabled a sensitive search for the hyperfine HI 21cm line from neutral atomic hydrogen (HI) in galaxies in the EGS, in the redshift range z~0.05-0.4. Bera et al. stacked (i.e. averaged) the HI 21cm emission signals from 445 blue star-forming galaxies in the EGS at 0.2<z<0.4 to infer their average HI gas mass, obtaining an average HI mass of (4.93 +/- 0.70) × 10^9 solar masses at a mean redshift of <z>=0.34. This implies a ratio of average gas mass to average stellar mass of ~1.2 for star-forming galaxies at these redshifts, higher than the corresponding value in the local Universe. The author also stacked the rest-frame 1.4 GHz radio continuum emission of the same galaxies, and then used a relation between the 1.4 GHz radio luminosity and the star formation rate (SFR) to obtain a median SFR of (0.54 +/- 0.06) solar masses per year for the galaxies of the sample. If the galaxies continue to form stars at the same rate, their average HI content would be exhausted on a timescale of ~9 Gyr, consistent with values in star-forming galaxies in the local Universe. This suggests that the star-formation efficiency in blue star-forming galaxies has not changed significantly over the last ~4 Gyr. Finally, Bera et al. used the stacked HI 21 cm emission signal to infer the cosmic HI mass density in star-forming galaxies at z=0.2-0.4, obtaining a normalized cosmic HI density of (4.81 +/- 0.75) x 10^−4 at <z>=0.34. This is the first accurate measurement of the cosmic HI density at intermediate redshifts z~0.2-1.8, and indicates no significant evolution in the cosmic HI density from z~0.4 to the present epoch. The top panel of the figure shows the average HI 21cm emission profile of the 445 blue star-forming galaxies whose spectra were stacked together. The bottom panel shows the evolution of cosmic HI density from z~5 to today, with the blue star showing the measurement from the present study.
Ionized carbon 158 micron Emission from z~4 HI Absorption-selected Galaxies
Neeleman, Kanekar et al. have used the Atacama Large Millimeter/sub-millimeter Array (ALMA) to carry out a search for the ionized carbon ([CII]) 158 micron emission line from galaxies associated with four high-metallicity damped Ly-alpha absorbers (DLAs) at z~4. They detected [CII] 158 micron emission from galaxies at the DLA redshift in three fields, with one field showing two [CII] emitters. Combined with previous results, Neeleman et al. have now detected [CII] 158 micron emission from five of six galaxies associated with targeted high-metallicity DLAs at z~4. The galaxies have relatively large impact parameters, ~ 16-45 kpc, [CII] 158 micron line luminosities of 0.04-3 billion solar luminosities, and rest-frame far-infrared properties similar to those of luminous Lyman-break galaxies, with star formation rates of ~7-110 solar masses per year. Comparing the absorption and emission line profiles yields a remarkable agreement between the line centroids, indicating that the DLA traces gas at velocities similar to that of the [CII] 158 micron emission. This disfavours a scenario where the DLA arises from gas in a companion galaxy. These observations highlight ALMA’s unique ability to uncover a high-redshift galaxy population that has largely eluded detection for decades. The figure shows (top panels) the integrated [CII] 158 micron flux density maps over channels containing line emission and (bottom panels) 350 GHz continuum images of the four quasar fields (with the plus sign indicating the quasar position). For the sole [CII] 158 micron non-detection, the [CII] line flux density is integrated over the central 100 km/s around the DLA redshift.
Massive Absorption-selected Galaxies at Intermediate Redshifts
The nature of absorption-selected galaxies and their connection to the general galaxy population have been open issues for more than three decades, with little information available on their gas properties. Kanekar et al. used detections of carbon monoxide (CO) emission with the Atacama Large Millimeter/submillimeter Array to show that five of seven high-metallicity, absorption-selected galaxies at intermediate redshifts, z ~ 0.5–0.8, have extremely large molecular gas masses and high molecular gas fractions relative to stars. Their modest star formation rates then imply long gas depletion timescales, an order of magnitude larger than typical in star-forming galaxies. High-metallicity absorption-selected galaxies at z ~ 0.5–0.8 thus appear distinct from populations of star-forming galaxies at both z ~ 1.3–2.5, during the peak of star formation activity in the Universe, and low redshifts, z < 0.05. Their relatively low SFRs, despite the large molecular gas reservoirs, may indicate a transition in the nature of star formation at intermediate redshifts, z ~ 0.7. The figure shows the five CO detections (in contours), with the position of the background quasar indicated by a star in each panel.
The gas and stellar mass of low-redshift damped Lyman-alpha absorbers
Kanekar et al. report Hubble Space Telescope Cosmic Origins Spectrograph far-ultraviolet and Arecibo Telescope HI 21cm spectroscopy of six damped and sub-damped Lyman-alpha absorbers (DLAs and sub-DLAs, respectively) at z<~0.1, which have yielded estimates of their HI column density, metallicity and atomic gas mass. This significantly increases the number of DLAs with gas mass estimates, allowing the first comparison between the gas masses of DLAs and local galaxies. Including three absorbers from the literature, they obtain HI masses ~(0.24-5.2) billion solar masses, lower than the knee of the local HI mass function. This implies that massive galaxies do not dominate the absorption cross-section for low-z DLAs. Kanekar et al. use Sloan Digital Sky Survey photometry and spectroscopy to identify the likely hosts of four absorbers, obtaining low stellar masses, ~(0.01-0.3) billion solar masses in all cases, consistent with the hosts being dwarf galaxies. They obtain high HI 21cm or CO emission line widths, ~ 100-290 km/s, and high gas fractions, ~5-100, suggesting that the absorber hosts are gas-rich galaxies with low star formation efficiencies. However, the HI 21cm velocity spreads (>~ 100 km/s) appear systematically larger than the velocity spreads in typical dwarf galaxies. The figure shows the Arecibo HI 21cm spectra for the six galaxies of the paper.
The Gas Mass of Star-forming Galaxies at z~1.3
Kanekar et al. used the GMRT 610 MHz receivers to carry out a search for HI 21cm emission from a large sample of massive star-forming galaxies at z~1.18-1.34, lying in sub-fields of the DEEP2 Redshift Survey. The search was carried out by co-adding (``stacking'') the HI 21cm emission spectra of 857 galaxies, after shifting each galaxy’s HI 21cm spectrum to its rest frame. The non-detection of a signal in the stacked HI 21cm spectrum yielded a stringent upper limit of 2.5 microJy on the average HI 21cm flux density of the 857 galaxies, at a velocity resolution of 315 km/s. This implies an upper limit of 20 billion solar masses on the average HI mass of the 857 galaxies, the first direct constraint on the atomic gas mass of galaxies at z>0.5. The upper limit to the ratio of the atomic gas mass to the stellar mass, i.e. the gas fraction, is 0.5, comparable to the cold molecular gas fraction in similar galaxies at these redshifts. Kanekar et al. find that the cosmological mass density of neutral atomic gas in massive star-forming galaxies at z~1.3 is significantly lower than the mass density estimates in both galaxies in the local Universe and damped Lyman-alpha absorbers at z>2. This implies that massive blue star-forming galaxies do not dominate the neutral atomic gas content of the Universe at z~1.3. The figure shows the cosmological mass density in neutral gas plotted as a function of redshift. The open star shows the new GMRT result, for blue star-forming galaxies at z~1.3. See the paper for more details.
Ionized carbon and dust emission from high-redshift galaxies
Gas surrounding high-redshift galaxies has been studied through observations of damped Lyman-alpha absorbers toward background quasars for decades. However, it has proven difficult to identify and characterize the galaxies associated with these absorbers due to the intrinsic faintness (at optical wavelengths) of the foreground galaxies compared with the background quasars. Neeleman et al. used the Atacama Large Millimeter/Submillimeter Array to obtain the first detections of ionized carbon ([CII]) 158-micron line and dust-continuum emission from two galaxies associated with damped Lyman-alpha absorbers at very high redshifts, z~4. The two upper panels of the figure show the dust continuum emission from the galaxies, while the lower panels show the [CII] 158-micron line emission. The results indicate that the host galaxies of the two absorbers are massive, dusty and rapidly star-forming systems. The hosts appear to be embedded in enriched neutral hydrogen gas reservoirs that extend well beyond the star-forming interstellar medium of the galaxies. The figure shows the two detections of ionized carbon (bottom panels) and dust continuum emission (top panels) from the two DLAs at z~4.
The spin temperature of high-redshift damped Lyman-alpha systems
Kanekar et al. report results from a large programme aimed at investigating the temperature of neutral gas in high-redshift damped Lyman-alpha absorbers (DLAs). This involved (1) HI 21cm absorption studies of a large sample of DLAs towards radio-loud quasars, to measure the spin temperature (2) very long baseline interferometric studies to measure the low-frequency quasar core fractions, and (3) optical/ultraviolet spectroscopy to determine DLA metallicities and the velocity widths of low-ionization metal lines. Kanekar et al. found a statistically significant difference between the spin temperature distributions in the high-redshift (z > 2.4) and low-redshift (z < 2.4) DLA samples: the high-z sample contains more DLAs with high spin temperature, >~ 1000 K. The high DLA spin temperatures arise due to low fractions of the cold neutral medium (CNM): only two of 23 DLAs at z > 1.7 have CNM fractions > 20%, comparable to the median value (~ 27%) in the Milky Way. Kanekar et al. robustly confirmed the presence of an anti-correlation between spin temperature and metallicity [Z/H], via a non-parametric Kendall-tau test. The data thus appear to indicate that high-redshift DLAs have significantly larger fractions of the warm phase of neutral hydrogen than is present in the Milky Way and local spiral galaxies, probably because the paucity of metals in the absorbers implies a lack of cooling routes in the absorber host galaxies. The figure shows the spin temperature of the DLAs of the sample plotted versus redshift (left panel) and metallicity (right panel). The left panel shows that there is a higher fraction of DLAs with high spin temperatures at high redshifts, z>1.7. The anti-correlation between spin temperature and metallicity is clearly visible in the right panel.
GMRT detections of HI 21cm absorption in two damped Lyman-alpha absorbers at z~2
Kanekar used the new 250-500 MHz receivers of the upgraded GMRT to detect redshifted HI 21cm absorption in two high column density damped Lyα absorbers (DLAs) at z ~ 2. Both absorbers have high inferred integrated HI 21cm optical depths, and hence low spin temperatures. However, for the z=1.9698 DLA toward TXS1755+578, the difference in HI 21cm and C I profiles and the weakness of the radio core suggest that the HI 21cm absorption arises toward radio components in the jet, and that the optical and radio sightlines are not the same; this precludes an estimate of the DLA spin temperature. For the z=1.9888 DLA toward TXS1850+402, the HI 21cm absorption yields a DLA spin temperature ~ 372 K, lower than typical spin temperature values in high-z DLAs. This low spin temperature and the relatively high metallicity of the z=1.9888 DLA are consistent with the anti-correlation between metallicity and spin temperature that has been found earlier in damped Lyα systems. The figure shows the two new GMRT HI 21cm absorption detections.

BACK