There are prospects for dramatic steps forward in radio and millimeter-wave astronomy within the decade, thanks to a new generation of large to gigantic interferometers as well as refurbishment of old interferometers. Interferometric observations gain over single-dish observations not just through resolution but through improved sensitivity, because correlation of the signals from the antennas can distinguish signal from noise and background. Long integrations become possible without the limitation of systematic errors. However, observing with interferometers requires careful set-up of the antenna array (and its parameters in software) to image sources correctly by measuring their flux densities on the appropriate angular scales. There is also the dreaded problem of `missing flux' from (lack of) low-order harmonics in the spatial transform, corresponding to structure on the larger scales (Section 2). In addition, the amount of post-processing required is large in comparison to single-dish measurements, to correct for the many instrumental and atmospheric issues, to convert the Fourier components into brightness images of radio sources, and (because of incomplete sampling in the Fourier plane) to apply algorithms to maximize image fidelity and dynamic range. Interferometers offer an additional advantage: on the increasingly noise-polluted surface of our planet, the processing can excise radio-frequency interference (RFI), which, even at the remote sites of future large arrays, would otherwise seriously compromise observations.
The small beam size at millimeter wavelengths makes large-area deep surveys extremely difficult because of the time penalty. New scanning techniques will need to be developed to perform such surveys. The next generation of interferometers, thanks to larger collecting areas, broader bandwidths and faster scanning capabilities are expected to produce deeper surveys of large sky regions, both in total intensity and in polarization. Receiver advances have resulted in huge bandwidths ( 10 per cent) and very low equivalent noise temperatures over these bandwidths. To realize these gains in sensitivity with interferometers requires development in correlator speed and in processing power. Moreover instead of single-pixel feeds, the development of focal plane arrays (FPAs, or phased-array feeds) looks to realize the long-standing dream of making near-full use of the information brought to the focal plane (e.g. APERTIF, Verheijen et al. 2008). There are further implications for correlator- and processor-power requirements.
The focus of current effort in the radio-astronomy community is towards the realization of the Square Kilometer Array (SKA), the largest and most sensitive radio telescope ever. The SKA stands to be one of the iconic scientific instruments of the 21st century. It will consist of an array of thousands of dishes, each 10-15 m in diameter, as well as a complementary aperture array - a large number of small, fixed antenna elements plus receiver chains arranged in a regular or random pattern on the ground. Between these two technologies, a frequency range of 100 MHz to 25 GHz will be covered. The collecting area will add up to approximately one million square meters, with baselines ranging from ~ 15 m to more than 3000 km. The SKA will require super-fast data transport networks (of order the current total internet capacity) and computing power far beyond current capabilities. Indeed the concept is only feasible if Moore's Law (the packing density of processing elements approximately doubling every two years) continues to hold; and this in itself requires revolutions in processor technologies. Site-testing has narrowed the choice to remote regions either in Western Australia or in South Africa. The telescope is expected to be fully operational after 2020, but a 10% SKA may be operating as early as 2015. Many different technological solutions will be selected and integrated into the final instrument: they will represent the results of developing the so-called SKA pathfinders (ASKAP, MeerKAT, ATA, LOFAR, e-MERLIN, EVLA; www-page descriptions are readily available for each).
These pathfinders carry the shorter-term excitement; they themselves represent leaps forward in observational capability. Most of them will be operational by 2015. They will realize dramatic improvements in survey speed and sensitivity; ASKAP for example, is expected to produce a survey similar to the NVSS (Condon et al. 1998) in sky coverage, but bettering it by a factor of 50 in sensitivity and of 5 in angular resolution (Johnston et al. 2008). These new tools will allow us to distinguish the star-forming population from the AGN population at low flux-density levels and to investigate source populations at extreme redshifts.
The SKA itself will impact every area of astronomy and cosmology, from detection and mapping of planetary systems, study of individual stars, star clusters, pulsars, the structure of our Galaxy in both baryons and in magnetic field, through to normal galaxies, AGNs, proto-galaxies, and large-scale structure of the Universe. The compelling science (e.g. Carilli & Rawlings 2004 and updates on the SKA website) to be realized makes irresistible reading.
LOFAR (Rottgering et al. 2006), will open up the frequency window at the low end of the radio spectrum, below 240 MHz. LOFAR will survey the sky to unprecedented depths at low frequencies and will therefore be sensitive to the relatively rare radio sources that have very steep spectra, extreme luminosities and redshifts (Section 1). A unique area of investigation will be the search for redshifted 21cm line emission from the epoch of reionisation.
At the other end of the radio band, the Atacama Large Millimeter Array (ALMA) is eagerly anticipated by the mm continuum and molecular-line community. Being built on a high (5000m), dry plain in the Atacama desert, northern Chile, it is an international project, a giant array of 50 12-m submillimetre quality antennas, with baselines of several kilometres. An additional compact array (ACA) of 12 7-m and 4 12-m antennas is also foreseen. ALMA will be equipped with mm and sub-mm receivers covering ultimately all the atmospheric windows at 5000m altitude in ten spectral bands, from 31 to 950 GHz. The array will be operational by 2012 with a subset of the high-priority receivers. The steep rise of the dust emission spectrum at mm and sub-mm wavelengths implies that the K-correction compensates, at z 0.1, for the dimming due to increasing distance (Blain & Longair 1993), making the observed mm flux of dusty galaxies of given bolometric luminosity only weakly dependent on redshift up to z 10. This makes ALMA the ideal instrument for investigating the origins of galaxies in the early universe, with confusion made negligible by the high spatial resolution. Using far-IR emission lines and CO rotational emission, ALMA will reveal the astrophysics of early phases of galaxy formation and provide the redshift of large numbers of obscured star-forming galaxies up to very large distances. This will enable us to establish the star-forming history of the universe, without the uncertainties caused by dust extinction in optical studies.
Technological advances have resulted in upgrades of existing telescopes that revolutionize performance. For examples:
Finally, several new survey instruments for the Sunyaev-Zeldovich effect have either started operations or will shortly do so (see Section 8.1).