The primary goal of radio source surveys is to generate flux-limited samples. Sources selected at very low frequencies are dominated by unbeamed emission and give the only unbiased view of the parent populations used by “unification” models to account for the diversity of sources seen at high frequencies. Low-frequency surveys favor sources with exceptionally steep spectra. They include radio galaxies at high redshifts, radio halos of nearby galaxies, relic radio sources, diffuse cluster emission, pulsars that may be missed by traditional pulse searches, and a new class of unidentified compact sources. Flux densities from low-frequency surveys extend the spectra of known source populations to frequencies at which free-free and synchrotron absorption become significant and constrain basic source parameters. Finally, telescope fields-of-view scale ∝ λ2, so gridded surveys can be more efficient than directed observations of individual targets. This review covers recent and proposed low-frequency source surveys and their astronomical uses.

The Sydney University Molonglo Sky Survey (SUMSS) is a radio imaging survey at 843 MHz of the whole sky south of declination −30°. With a resolution of 43″ × 43″ cosec |δ| and an rms noise level of ∼ 1 mJy/beam, SUMSS has similar sensitivity and resolution to the northern NRAO VLA Sky Survey (NVSS). Here, we present some results from the first two years of SUMSS and also show what can be done by combining radio data from SUMSS and NVSS with the new generation of large optical redshift surveys (including the 2dF Galaxy Redshift Survey and 6dF Galaxy Survey) now becoming available in the southern hemisphere.

Over many years low-frequency radio surveys have been made in Cambridge that cover much of the northern sky at both 38 and 151 MHz. The characteristics of the earlier surveys − 6C (at 151 MHz, δ above +30°) and 8C (at 38 MHz, δ above +60°), both with a resolution of about 4 arcmin — are reviewed, together with the results from more recent surveys made at 151 MHz, with a resolution of about 1·2 arcmin with the Cambridge Low-Frequency Synthesis Telescope (CLFST). These surveys include the 7C survey, which includes low declinations to complement 6C, selected regions at higher declinations, variability studies, and the 7C(G) survey of much of the northern Galactic plane.

The Mauritius Radio Telescope (MRT) is a Fourier synthesis instrument which has been built to fill the gap in the availability of deep sky surveys at low radio frequencies in the southern hemisphere. It is situated in the north-east of Mauritius at a southern latitude of 20°.14 and an eastern longitude of 57°.73. The aim of the survey with the MRT is to contribute to the database of southern sky sources in the declination range −70° ≤ δ ≤ −10°, covering the entire 24 hours of right ascension, with a resolution of 4' × 4'.6sec(δ + 20.14°) and a point source sensitivity of 200 mJy (3σ level) at 151.5 MHz.

MRT is a T-shaped non-coplanar array consisting of a 2048 m long East-West arm and a 880 m long South arm. In the East-West arm 1024 fixed helices are arranged in 32 groups and in the South arm 16 trolleys, with four helices on each, which move on a rail are used. A 512 channel, 2-bit 3-level complex correlation receiver is used to measure the visibility function. At least 60 days of observing are required for obtaining the visibilities up to the 880 m spacing. The calibrated visibilities are transformed taking care of the non-coplanarity of the array to produce an image of the area of the sky under observation.

This paper will describe the telescope, the observations carried out so far, a few interesting aspects of imaging with this non-coplanar array and present results of a low resolution survey (13' × 18') covering roughly 12 hours of right ascension, and also present an image with a resolution of 4' × 4'.6sec(δ + 20.14°) made using the telescope.

We present WSRT 1.38 GHz observations of the Hubble Deep Field (and flanking fields). 72 hours of data were combined to produce the WSRT's deepest image yet, reaching an r.m.s. noise level of 8 microJy/beam. We detect radio emission from galaxies both in the HDF and HFF which have not been previously detected by recent MERLIN or VLA studies of the field.

I review some recent cosmological studies based on redshift surveys of radio sources selected at low frequencies. The accretion rate onto the central black hole is identified as the basis of a crude physical division of the low-frequency population into two sub-populations: the first comprises ‘Eddington-tuned’ (high accretion rate) quasars and their torus-hidden counterparts; the second comprises ‘starved quasars’ like M87. There exist remarkable similarities between the shapes and evolutionary behaviours of the luminosity functions of radio sources and radio-quiet quasars; all luminous AGN are ∼ 300—times rarer at epochs corresponding to z ∼ 0 than at z ∼ 2.5. I argue that any evidence that quasars were intrinsically rarer at z ∼ 5 than at z ∼ 2.5 is as yet both tentative and indirect. A simple calculation suggests that the radio source population has been over-looked as a potentially important contributor to the entropy budget of the Universe. A recent sub-mm survey of radio sources is used to demonstrate a connection between the events which trigger jets and intense bursts of star formation, and a close link between the histories of star formation and AGN activity is proposed. I discuss the aims and methods of future large redshift surveys of radio sources, emphasising the importance of dovetailing these with the development of robust physical models for radio sources and their epoch-dependent environments.

We find simple parametric models to describe the space density evolution of radio-loud AGN, treating FRI and FRII radio galaxies separately as the two parent populations in our dual-population unified scheme. In this we use low frequency radio data (v < 500 MHz), where radio samples are unbiased by Doppler beaming. Incorporated into this latest analysis is a new determination of the local radio luminosity function at 1.4 GHz from galaxies common to both the 2dFGRS and NVSS surveys.

The peak in the power spectrum at ≈ 130h−1 Mpc at low redshifts provides a standard ruler in comoving space. This scale is shown to be present in the observed quasar distribution at z ∼ 2. This implies strong constraints on the density parameter, Ω0, and weaker constraints on the cosmological constant, λ0. Independently of λ0 (in the range λ0 ∊ [0, 1]), the constraint is 0.1 < Ω0 < 0.45 (68% confidence limit).

Combination of the power spectrum peak method with very independent results from the supernovae Type Ia method yields Ω0 = (0.30 ± 0.11) + (0.57 ± 0.11)(λ0 −0.7), 0.55 < λ0 < 0.95, (68% confidence limits) without assuming Ω0 + λ0 = 1. This supports the almost flat, perturbed Friedmann-Lemaître-Robertson-Walker model, independently of any cosmic microwave background observations.

Estimations of the possible activity periods of galaxies and quasars were made from the UTR-2 catalogue samples by two methods.

The next generation low-frequency radio telescopes may probe cosmological models by means of observations of the cosmic microwave background (CMB). I discuss the prospects for observations of CMB imprints —- recombination lines from the epoch of recombination, μ distortions and angular temperature anisotropies —- at low frequencies. A future low-frequency radio telescope, like the proposed SKA, may be capable of attempting some difficult CMB measurements because of the large collecting area and large element numbers; however, this will require a telescope design that will allow specialized calibration strategies and will give emphasis to the control of spurious responses.

Observation of an interstellar line in absorption against the cosmic 2.7 K background is an unusual phenomenon. The intensity, Iv, of a line generated in an interstellar cloud, with homogeneous excitation conditions, is given by where Iv,bg is the intensity of the continuum against which the line is observed, τv the optical depth of the line, and Sv the source function, which is the Planck's function at the excitation temperature Tex, i.e., Sv = Bv(Tex). Observation of an interstellar line in absorption against the cosmic 2.7 K background, obviously, implies 0 K < Tex < 2.7 K, which requires rather peculiar physical conditions in the molecule, generating the line.

Future radio surveys using facilities like the Giant Metre-wave Radio Telescope or a Square Kilometre Array may detect 21-cm radiation from the Intergalactic Medium (IGM) in either emission or absorption against the Cosmic Microwave Background (CMB) at high redshift. Prior to the reionization of the IGM, the first radiation sources would reveal their presence through their effect on the IGM. The signatures include a sharp absorption feature in the radio sky against the CMB resulting from a rapid rise of a Lyman-alpha continuum background shortly after the birth of the first UV sources in the universe, and 21-cm emission and absorption shells created on the several Mpc scale by the first bright QSOs. The detection of these signatures would map the end of the ‘dark ages', the period of transition from the post-recombination universe to one populated by radiation sources.

The reionization epoch may lie not far beyond the most distant known quasars and galaxies, and it might be detectable in a variety of ways. One possibility discussed here is to search for an all-sky spectral step produced by the redshifted HI that was present prior to reionization.

Radio absorption line observations of neutral hydrogen gas against extended radio sources offers the means to measure sizes and kinematics in intervening galaxies at all redshifts up to the maximum redshift where radio galaxies are detected. Such observations can therefore trace the evolution of galaxies at redshifts z ≳ 2 where the damped Lyman — α statistics indicate that the mass in neutral gas exceeded the mass in stars.

WSRT observations have provided a first inventory of the incidence of HI 21 cm line absorption associated with AGN at redshifts up to z=1.0. There is a large range in line depths, from τ = 0.44 to τ ≤ 0.001, and a substantial variety of line profiles, from Gaussians of less than ten km s−1 to more typically a few hundred km s−1, as well as irregular and multi-peaked absorption, sometimes spanning many hundreds of km s−1. The chance of detecting appreciable HI absorption is greatest in the most compact radio sources, GPSs and CSOs, where it can occur in circumnuclear “disks” or “tori”, as well as in gas enveloping jets and hot spots; inferred densities range at least between 10 cm−3 and 104 cm−3. But HI absorption occurs also in some CSSs, perhaps associated with jet-cloud interaction regions, and in quasars with a large optical reddening. VLBI observations at the unusual UHF frequencies of redshifted HI 21 cm can give unique “sight”lines into the physics and evolution of young radio sources and their inner galactic medium.

The Giant Metrewave Radio Telescope (GMRT) has recently come into operation. This paper reports on observations of the HI line made using the GMRT during its commissioning phase. These include observations of the nearby Low Surface brightness galaxy Cepheus 1, as well as of a number of low and medium redshift damped Ly-α systems. The derived spin temperatures for these systems is typically higher than that characteristic of the gas in the disk of the Galaxy. Higher spin temperatures were known to be characteristic of high redshift damped Ly-α absorbers, and had been earlier attributed to evolutionary effects. Our observations suggest that high spin temperatures are not a consequence of evolutionary effects since the look back time to these low z absorbers is modest. We also report on the short time scale (days) variability of the absorption profile of the z = 0.3127 system in front of B1127-145. This is the second damped Ly-α system to show such variation (the other being the z = 0.524 absorber in front of AO 0235+164).

We summarize results from an extensive search for HI 21cm absorption at moderate redshifts. The survey was limited to systems with Mg II absorption at redshifts in the range 0.2 < z < 1.0, which lie in front of bright radio quasars. Narrow band observations centered at the metalline redshifts were made with the UHF-high system and, when possible, the DZB correlator at the WSRT. Three new absorbers have been confirmed and several candidate absorbers identified. Optical depth limits of a few percent at velocity resolutions of about 6 km s−1 are obtained for most of the non-detections. The 21cm search is complementary to similar damped Lyα surveys, and by combining information from both absorption lines we are able to better understand physical conditions in these high neutral column density absorbers. We find that the 21cm objects represent a cold subset of high neutral column density absorbers, but are not necessarily the highest column density systems.