The enhanced VLA will have a major impact in the area of unbiased surveys in the neutral hydrogen line. For continuum observations the VLA has been unsurpassed as a surveying instrument, with a collecting area that is only second to Arecibo, combined with the large field of view. For blind H I surveys a third dimension comes into play-the instantaneous redshift range that can be covered. The new correlator will be more powerful in that respect than any currently operational or planned correlator on a synthesis instrument. Thus the VLA will become the most powerful H I surveying machine by far. Extending the 1.4 GHz band down to 800 MHz would allow the direct study of galaxy evolution by imaging the H I emission of galaxies in . While the highest redshift at which H I has yet been detected in emission is about 0.08, the enhanced VLA could detect thousands of galaxies at redshifts from 0.2 to 0.8 in a 100-hr integration.
The large scale structure of the universe as we know it today, with its large filaments, sheets, walls and voids, is based almost entirely on observations of high-luminosity galaxies, observed at optical wavelengths. Biased galaxy formation predicts a separation between luminosity (bright galaxies) and mass (dark matter). Dwarf galaxies are expected to be more uniformly distributed. Since nearly all investigations of the properties and spatial distribution of galaxies begin with optically (or IRAS) selected galaxy catalogs, any population of gas clouds with very low optical luminosity or surface brightness could have largely escaped detection. Direct searches in the H I line circumvent these optical selection biases.
Examples of questions to be addressed by such surveys are the abundance of gas rich dwarfs and the study of their spatial distribution. Do they follow the optically bright galaxies or are they distributed more uniformly as predicted in some theories of galaxy formation. It will become possible to construct an unbiased H I mass function for the first time. A complete inventory can be made of the H I content in the local universe, the size and mass distribution of H I disks. Galaxies may be found in the zone of avoidance and new light may be shed on the parent population of nearby Ly absorbers. At larger redshifts, the structure of clusters of galaxies and voids may may be determined. At even higher redshifts entirely new ground will be trodden. The H I properties of galaxies at 0.1 will be studied for the first time, and it will be possible to study the upper end of the H I mass function out to 1. This might settle once and for all whether the gas reservoirs around galaxies increase with increasing redshift, and whether entirely new gaseous galaxy types occur at these higher redshifts.
Below we present three examples of such surveys, one concerning the local universe and employing the extreme surface brightness sensitivity of the E configuration, one performing simultaneously an in-depth study of a cluster at a redshift of 0.1, combined with a deep pencil beam survey covering a range from 0 0.2 to be done with the C configuration. The third survey is aimed at studying the evolution of the gas reservoirs around galaxies in the range from 0.2 0.8 with the resolution of the B configuration.
In recent years all sky continuum surveys covering almost the entire electromagnetic spectrum have become available, giving us a good indication of the stellar, dust, hot gas and relativistic electrons distribution in the universe. However for the most abundant element, neutral hydrogen, this information is lacking. In fact, we don't even know the H I content of the local group of galaxies.
The high surface brightness sensitivity of the VLA E configuration, combined with the large instantaneous velocity coverage of the correlator will finally make it feasible to survey H I over the whole sky, simultaneously covering a redshift from 0 to 0.1. This will give a complete inventory of H I masses irrespective of a stellar association to solar masses within the local group, to solar masses at the distance of Virgo Cluster, solar masses in Coma, to solar masses at the survey limit at a distance of 280 Mpc. Note that even in the nearest cluster, Virgo, only a few percent of the total volume has yet been searched for HI. The total search volume of this survey, , is such that about galaxies should be detected with H I mass less than solar masses .
The assumed survey parameters are an integration time of 3 minutes per pointing, giving a column density sensitivity of cm over a beam, using a total bandwidth of 125 MHz and 2048 channels of 12.6 km/s. The entire sky north of deg can then be observed to the 6 mass limits noted above in a (mere) 2.7 yrs.
At higher redshifts, an entire cluster fits into one primary beam. For example at z = 0.1, an Abell diameter is about , so we can study an entire cluster in one pointing, giving the spatial and kinematic distribution of all the gas-rich objects in and around it. The resolution of the C configuration provides some spatial information for individual galaxies. A 36-hr integration would (at z = 0.1), allow a 6 detection of solar masses of H I with 100 km/s line width. A total band of 62.5 MHz is needed, corresponding to 13,000 km/s split up in 2048 channels of 6.3 km/s. The spatial and velocity distribution of the gas-rich galaxies will provide information on the dynamical state of the cluster, whether it is relaxed or still collapsing. The velocity coverage of 13,000 km/s is twice that needed to cover a typical rich cluster, so many foreground and background spirals will also be detected.. The 6 km/s accuracy of the velocity profiles will make it possible to derive useful Tully-Fisher distances to the galaxies and thus to probe deviations from the Hubble flow out to many Mpc from the cluster. These observations will also provide a useful data base to compare with cluster studies at higher redshifts aimed at determining the evolution of the gas content in clusters.
In parallel with such deep integrations a sensitive pencil beam survey can be carried out to sample the entire conical volume within the primary beam between z = 0 - 0.2, using 250 MHz bandwidth and 1024 channels of 250 kHz (50 km/s). H I masses of solar masses will be detectable at the survey limit of z = 0.2, implying that normal galaxies with modest H I content can easily be detected. Note that currently the highest redshift at which H I in galaxies has been detected in emission is about 0.1. Within each search volume of we expect to detect a few hundred galaxies with H I mass less than solar masses .
Recent Hubble Space Telescope results indicate that both field and cluster galaxies are undergoing dramatic evolution at 0.5. A relative increase in the number of spiral-like galaxies is found, representing interacting and merging galaxies in clusters and a population of blue field galaxies unlike anything seen in the local universe. A major question is how does the gas content of galaxies evolve at these redshifts, it is expected to increase with redshift. Extending the frequency coverage to 800-1200 MHz would allow us to probe this critical phase in galaxy evolution in the redshifted H I line. Making the conservative assumption that H I line fluxes at z = 0.5 are similar to the largest seen in the nearby universe, namely solar masses with a velocity width of 300 km/s, a 6 detection can be made in 100 hrs at that redshift. The largest possible bandwidth should be used to probe the entire redshift interval from z = 0.2 to 0.8 simultaneously, if the correlator capacity would permit (500 MHz and 2048 channels). Each field observed would then sample a total volume of , subject to limitations imposed by RFI. An unknown number of H I luminous objects might be detected in the interval 0.5 0.8, but more importantly, assuming only local space densities and H I luminosities we can expect detection of more than galaxies per field in the interval z = 0.2 to 0.5, revolutionizing our understanding of galaxy evolution. Such a survey would be done in the B configuration, matching the beam size to the typical galaxy size at those redshifts.