Large Radio Telescopes and the Emergence of the SKA, 1957–1993

Abstract

The SKA is the culmination of radio telescope design development in the decades prior to 1990. This chapter surveys the earlier telescopes and traces the origins of the ideas behind the SKA design.

2.1 Introduction

As Chap. 1 outlined, the SKA will be a radio interferometer with 100 times the collecting area of the Very Large Array (VLA), initially conceived as the instrumental answer to the quest to detect neutral hydrogen spectral-line emission from distant galaxies in the universe. This was the primary goal motivating the people recognised as the originators of the SKA—Govind Swarup in India, Peter Wilkinson in the UK, and Robert Braun, Ger de Bruyn and Jan Noordam in The Netherlands—to explore, independently, over a decade-long period from 1978 to 1990 very large radio telescope concepts capable of satisfying this aim. In addition to these ideas, Yuri Pariiskii and colleagues in the Soviet Union began to develop a 1 km² telescope concept in 1982 with a variety of scientific aims including neutral hydrogen. Later in the chapter, we describe how these pioneering ideas were developed, how they first came to the attention of an international audience in October 1990 at IAU Colloquium 131,¹ and the creation three years later in 1993 of the first formal international working group to take these ideas further.

A radio telescope with a large collecting area (where “large” is arbitrarily defined as >10,000 m²) has a long heritage going back to at least 1957 when the 600-foot (183 m) diameter Sugar Grove antenna was proposed and funding secured through the US Naval Research Laboratory, but never finished. In the 1960s, several large collecting area telescopes were built including the first very large diameter dish, the Arecibo 1000-foot (305 m) fixed spherical antenna, as well as large Mills Cross-type telescopes based on long cylindrical paraboloids, and dipole arrays at low frequencies (<200 MHz) where larger collecting areas are much easier to design and fund compared to instruments operating at higher frequencies (>1 GHz). Large radio telescope construction has continued since then, at a rate of one per decade or so, exemplified by the Very Large Array (VLA, later the Jansky VLA) built in the 1970s in the USA, the Giant Metrewave Radio Telescope (GMRT) built in the late 1980s–1990s in India and, in the 2000–2010s, the Low Frequency Array (LOFAR) in The Netherlands and the European-US-East-Asian Atacama Large Millimetre/submillimetre Array (ALMA) in Chile.

The telescopes built by the time the SKA concept first emerged in 1990 had geometric areas ranging from a few tens of square metres to 150,000 m² (at lower frequencies). Ekers and Wilson (2013) provide a list (in their Table A2) of operational centimetre and metre wavelength telescopes (anno 2012) with diameter greater than 25 m or equivalent area. Figure 2.1 shows a histogram of the distribution of the geometric collecting area of the radio telescopes listed in Ekers and Wilson (2013). Included in Fig. 2.1 are eleven large telescopes with collecting areas >10,000 m² that were built as well as nine not included in Ekers and Wilson (2013) which were proposed but never built. One of the latter group, the Cyclops array, has an entry in two columns in Fig. 2.1 one for the 750,000 m² version and the other for the 7.5 km² version. Cyclops had a major impact on thinking about the SKA concept. Note the entry (in blue on the right-hand side of the figure) for the first working square kilometre array, built in 1961–1962 by Grote Reber to operate at 2.085 MHz (see Sect. 2.2.3).

Fig. 2.1

Histogram of the distribution of geometric collecting area radio telescopes listed in Ekers and Wilson (2013). Metre wavelength telescopes are shown in blue and centimetre wavelength telescopes in orange. Also shown are entries for large (>10,000 m²) radio telescope concepts proposed but not constructed with metre wavelength telescopes in grey, and centimetre wavelength telescopes in yellow

A number of these telescopes influenced the early SKA concepts in direct and indirect ways particularly the “large” (>10,000 m²) telescopes and the somewhat smaller “work-horses” of radio astronomy—the 100 m diameter telescope at Effelsberg (Germany), the 100 m Richard C. Byrd III telescope at Green Bank (USA), the 76 m Lovell telescope at Jodrell Bank (UK), the 64 m Parkes dish (Australia), the RATAN-600 array (USSR), the Westerbork Radio Synthesis Telescope (The Netherlands), the Australia Telescope Compact Array, e-MERLIN (UK), the Very Long Baseline Array (USA), and the European VLBI Network. These provided hands-on experience for many of the engineers who designed the SKA and its principal Precursors and Pathfinders—MeerKAT, ASKAP, LOFAR and MWA.

In the decade leading up to the IAU Colloquium in 1990, there were thought to be several ways to create a collecting area of 1 million m² at frequencies above 300 MHz² including two-dimensional arrays of cylindrical paraboloids and arrays of small or large (up to 100 m diameter) parabolic dishes. Reflector arrays are preferable to a monolithic structure with a collecting area of 1 million m² even if the latter was feasible technically and from a cost point of view. The ability to spread the reflector elements out in an array of much larger dimensions than a monolithic structure provides far higher angular resolution observations. We briefly survey the concepts and constructions of arrays of cylinders and arrays of dishes in the next section to provide the context for the emergence of ideas for what became the SKA project. More comprehensive histories of many of these telescopes can be found elsewhere e.g. Leverington (2016) and Thompson et al. (2017).

One final comment in this introduction—in the early 1990s the range of SKA concepts expanded to include arrays of very large (>300 m) spherical dishes.³ This was stimulated by the existence of many suitable geological depressions in southern China (see Sects. 3.2.6.2, 3.3.3.3, and 7.2.1) similar to the Arecibo location in Puerto Rico. As the design process for the SKA developed further in the 1990s and 2000s and the potential scientific scope expanded, several other completely new, innovative concepts for the SKA that expanded radio astronomy parameter space came into contention for the selection of the SKA “reference design” in late 2005. These were dense aperture arrays, arrays of Large Adaptive Reflectors,⁴ and arrays of Luneberg Lens antennas.⁵ Also in contention in 2005 were low frequency dipole arrays primarily for the detection of the Epoch of Re-ionisation in the early universe (see Sect. 5.5.20). Brief descriptions of the heritage instruments for very large diameter (parabolic and spherical) dish arrays and dipole arrays are given in on-line supplementary material.⁶

2.2 Cylinder and Dish Array Concepts and Constructions: The SKA Instrumental Heritage in 1990

2.2.1 Parabolic Cylinder Arrays

Large collecting area parabolic cylinder arrays were in contention for the SKA antenna element design down select in 2005. Here we describe the extensive heritage of this concept.

2.2.1.1 A Concept Array of Parabolic Cylinders, 1962

Fig. 2.2

Sketch of an array of parabolic cylinders, 900 m × 900 m, with a collecting area of ~0.8 km² (Credit: Bracewell et al. (1962), reproduced with permission of Springer Nature)

Bracewell et al. (1962) published a paper on “Future large telescopes” in which they described an array of parabolic cylinders arranged like a square venetian blind with dimensions of 900 m × 900 m and a collecting area of about 800,000 m² (see Fig. 2.2). The large area cylinder concept was a natural extension of the earlier work by Bernard Mills (see below) but the Nature paper was not a proposal for a specific project and was never funded as such. Bracewell et al. contended this would be a more feasible way than large dishes to achieve a goal set by a US National Science Foundation Panel on Radio Telescopes (Keller, 1961) of 1 arcminute angular resolution at an observing wavelength of 21 cm thereby enabling studies of cosmology and the evolution of galaxies. However much smaller arrays of cylinders arranged like the Bracewell et al. concept were built as the north-south arm of the Northern Cross Telescope, Italy (see below), and as the Canadian Hydrogen Intensity Mapping Experiment (CHIME).⁷

2.2.1.2 Molonglo Cross Radio Telescope, Australia (Operational 1965–2023)

The Molonglo Cross Radio Telescope, a one by two cylinder array in a cross configuration, located at Bungendore, New South Wales (near Canberra), Australia, was proposed by Bernard Mills (University of Sydney) in 1960 and received seed funding the same year from the University of Sydney and from the US National Science Foundation in 1962 and 1964 (McAdam, 2008). At the time these were the largest grants made by the NSF to any science project outside the USA. It came into operation in 1965. The cross concept was conceived by Mills in 1953 while at CSIRO in Australia and earlier versions were built at Potts Hill and Fleurs near Sydney. The Molonglo Cross was formed by two intersecting 778 m × 12 m parabolic cylinders (see Fig. 2.3 left), one a fixed continuous structure north-south that could be steered electronically and separated east and west arms that were steered by mechanical rotation around their long axis. The total collecting area was 19,000 m² and the telescope operated at 408 MHz. For the first 11 years of operation, the primary scientific goals were all-sky surveys and searches for the then newly discovered pulsars. In 1978, it was converted to an Earth-rotation synthesis instrument at 843 MHz called MOST (Molonglo Observatory Synthesis Telescope) using the east and west arms only and a new deep survey of the whole southern sky was made. From 2013–2021 a super-computer backend was installed as part of a further upgrade of instrumentation for a project to detect Fast Radio Bursts. It ceased operation in 2023.

Fig. 2.3

Left: Aerial view of the Molonglo Cross Radio Telescope. The more visible arm in the picture is the 1.6 km east-west arm which intersects the north-south arm, also 1.6 km long, in the centre of the picture. (Courtesy of the University of Sydney Archives). Right: Aerial view of the Northern Cross Radio Telescope at Medicina comprising a single cylinder E-W arm 564 m long in the background, and a 640 m long array of 64 shorter cylinders as the N-S arm in the foreground. (Courtesy of INAF-Institute of Radio Astronomy)

2.2.1.3 Northern Cross Radio Telescope (Operational 1967)

The Northern Cross Radio Telescope, a one by sixty-four cylinder array in a cross configuration, at the Medicina Observatory near Bologna in Italy was constructed from 1960 to 1967⁸ (Fig. 2.3 right). Like the Molonglo Cross, it was a meridian transit telescope but with a different configuration composed of a single cylindrical reflector 564 m × 35 m forming the east-west arm, and an array of 64 cylindrical reflectors each 23.5 m long and 8 m wide forming a north-south arm with dimensions of 640 m × 23.5 m. The total useable collecting area is 27,000 m². Also, like the Molonglo Cross in the southern hemisphere, it observed at 408 MHz and carried out high sensitivity sky surveys yielding large radio source catalogues as well as pulsar research and spectrometric studies.

The Northern Cross has undergone several major upgrades of its instrumentation over the decades; most recently, the N-S arm has been converted into a sensitive Fast Radio Burst detector.

2.2.1.4 Pushchino DKR-1000, Russia (Operational 1964)

The DKR 1000 is a cross-type antenna⁹ is a one by one cylinder array in a T-configuration with cylindrical parabolic E-W and N-S arms each 1000 m long by 40 m wide providing a collecting area of 70,000 m (see Fig. 2.4 left). DKR stands for Wide-band Cross-type Radio telescope (in Russian—Diapazonnyi Krestoobraznyi Radioteleskop). It is located at Pushchino near Moscow and was brought into operation in 1964 in the 30–120 MHz frequency range. It was originally built to carry out counts of radio sources and is now known for its work on pulsar investigations, observations of spectral radio lines corresponding to transitions between levels with high principal quantum numbers, and studies of flux density variations in radio sources.

Fig. 2.4

Left: The East-West arm of DKR-1000 meridian radio telescope at Pushchino Radio Observatory in Russia (Credit: Rustam Dagkesamansky). Right: The single cylinder Ooty Radio Telescope in India (Credit: National Centre for Radio Astrophysics, Tata Institute of Fundamental Research)

2.2.1.5 Ooty Radio Telescope, India (Operational 1970)

Govind Swarup (Tata Institute for Fundamental Research, India) followed up the Bracewell et al. (1962) Nature paper referred to above with the design and construction of the Ooty Radio Telescope (ORT) near Udhagamandalam in India (Swarup et al., 1971), a single North-South cylinder 530 m long, and 30 m wide, with collecting area 16,000 m². Explaining what he wanted to do, Swarup (2006) said “the plan was to construct a large cylindrical radio telescope on a suitably-inclined hill in southern India so as to make its axis parallel to the Earth’s axis, and thus taking advantage of India‘s close proximity to the Equator”. It is located on a hill whose natural slope, 11°, is the same as the latitude of the ORT (see Fig. 2.4 right). This makes it possible to track celestial objects for about 10 h continuously as the Earth rotates only by rotating the antenna mechanically along its long axis. As with the Molonglo Cross, the antenna beam can be steered in the north-south direction electronically. It operates at 327 MHz and is known¹⁰ for its work on radio galaxies, quasars, supernovae, pulsars, the interstellar and interplanetary media. One of the most successful observational programs carried out for many years at Ooty was to determine the angular structures of hundreds of distant radio galaxies and quasars by the technique of lunar occultation.

2.2.1.6 Giant Equatorial Radio Telescope (GERT), Proposal, 1978–1983, India-Africa

In 1978, following several years of successful ORT operation, Swarup led an international African-Indian proposal for a large array of parabolic cylinders comprising one long cylinder +14 shorter cylinder segments with collecting area of ~200,000 m². This was to be located on the equator in Africa where no hill was necessary to enable simple tracking of the radio sources (Swarup, 1981). It was to be composed of one 2 km long × 50 m wide cylinder + fourteen 50 m × 15 m wide cylinder segments spread over an area of 14 × 12 km, operating at 325 and 38 MHz simultaneously. However, by the end of 1983, it was clear that GERT would not be funded and Swarup turned his attention to the GMRT, as described below.

2.2.2 Arrays of Dishes

2.2.2.1 The Benelux Cross Project Proposal, The Netherlands - Belgium, 1958–1964

Once the 25 m Dwingeloo telescope began operation in 1957 in the Netherlands, Jan Oort (Leiden University, The Netherlands) began thinking about a large telescope project that became known as the Benelux Cross (Christiansen & Högbom, 1961). Originally an advanced version of a Mills cross antenna (Sect. 2.2.1.2), it started life with two orthogonal parabolic cylinder arms each 5 km long by 30 m wide operating at 408 MHz and sited close to the border between The Netherlands and Belgium. The collecting area would have been 300,000 m². But prompted by the initial high costs of realising this design and the advent of computers providing the possibility of a smaller dish-based interferometer that could track individual radio sources for long periods of time to compensate for lower instantaneous collecting area, the design changed from cylinders to 100 or more 30 m diameter dishes (collecting area > 75,000 m², see Fig. 2.5 left,¹¹ (Raimond, 1996)). In addition, digital correlators had come onto the scene able to perform the pairwise combinations of the many dishes required and simulate an analogue instrument like a cylinder (W. N. Brouw, priv. comm.). The concept became a cross configuration but with 1.5 km arms and operating at 1.4 GHz to allow observations of the neutral hydrogen line. In 1962, the proposal went to both governments for approval, but political complications led to The Netherlands going it alone in 1964 with a linear E-W array of twelve 25 m dishes over 1.6 km. This was the Westerbork Synthesis Radio Telescope that started operation in 1970.

Fig. 2.5

Artist’s renditions of (left) the Benelux Cross Antenna (credit Herman Kleibrink, courtesy Netherlands Foundation for Research in Astronomy/ASTRON) and (right) the Cyclops array containing 1000 × 100 m diameter dishes (credit: Project Cyclops Report, prepared under Stanford/NASA/Ames Research Centre 1971 Summer Faculty Fellowship Program in Engineering Systems Design)

2.2.2.2 Project Cyclops Study, USA, 1971

In 1960, motivated by a suggestion by Cocconi and Morrison (1959) that interstellar signalling using radio might be feasible, Frank Drake (then at the National Radio Astronomy Observatory, USA) used the 85-foot telescope at Green Bank in the first attempt to detect interstellar radio signals from extra-terrestrial intelligence. His targets were two relatively nearby stars, and the frequency used was 1.4 GHz, the hydrogen line. After a decade of increasingly more sensitive, but unsuccessful, observations by Drake and collaborators, the National Aeronautics and Space Administration (NASA) commissioned a report in 1971 from an external panel led by Bernard (Barney) Oliver¹² (Hewlett-Packard Corp., USA) and John Billingham (NASA Ames Research Center, USA) to investigate what technology would be required to carry out an effective search for what Drake’s experience now showed would be extremely weak radio signals if they existed at all. In their comprehensive report (NASA, 1971), the panel concluded it would be feasible to start with a 3 km diameter array of 100 × 100 m dishes (collecting area 750,000 m²) operating between 1 and 3 GHz as a phased array forming a single beam (a one pixel image).¹³ In the light of the large uncertainty in the average distance between ‘communicative civilisations’ in the galaxy, they argued for an expandable search system, growing perhaps to 1000 × 100 m dishes (see Fig. 2.5, right, collecting area 7.5 km²), and only stopping once an Extra-Terrestrial Intelligence signal had been detected. The interesting concept of building a large telescope which can be incrementally expanded is possible for a radio array and this has influenced the final SKA design. This was the first proposal for a radio telescope with more than a square kilometre of collecting area at cm wavelengths. Estimated overall costs in 1971 were 6–10 billion USD. This cost tag made it impossible to fund, but its legacy influenced the SKA design requirements.

Project Cyclops became well-known in radio astronomy circles, setting a benchmark for “blue-sky” thinking in terms of number of antenna elements, total collecting area and estimated cost. It influenced several key “early players” in the SKA (see “Revisiting the Project Cyclops” by Jill Tarter (SETI Institute, USA) in Ekers et al. (2002)).

2.2.2.3 Very Large Array (VLA), New Mexico, USA (Operational 1980)

In 1961 in parallel with the Largest Feasible Steerable Telescope (LFST) Study,¹⁴ NRAO scientists began to design a radio telescope that could make images with comparable angular resolution to the best optical telescopes. By 1967, NRAO submitted a proposal to the US National Science Foundation (NSF) for the construction of the Very Large Array (VLA, see Fig. 2.6). The original proposal was for 36, later 27, fully steerable 25-m diameter antennas in a “Y” configuration spread over an area 35 km in diameter. The antennas can be moved along rail tracks to form four configurations. The total collecting area was to be 13,000 m², making it the most sensitive interferometer at cm wavelengths. Intense competition for the VLA came from a Caltech proposal for an 8-element array of 130-foot dishes and the 440-foot radome-enclosed antenna proposed by the North-East Radio Observatory Corporation described earlier. Eventually, support of the VLA by the 1970 National Academy Decadal Review of Astronomy led to approval of its construction; it was completed in 1980 (Kellermann et al., 2020, Chap. 8).

Once in operation, it took centre-stage in radio astronomy and has impacted most areas of astronomy and astrophysics and telescope design ever since.

Fig. 2.6

The 27 antennas of the VLA in its most compact configuration (credit: US National Radio Astronomy Observatory)

2.2.2.4 The Radio Schmidt Telescope Proposal, Canada, 1986–1991

The Radio Schmidt Telescope (RST), a dish array in a compact configuration covering 2 × 3 km, was conceived by Peter Dewdney during an extended visit to the VLA in 1986. It was to be a future aperture-synthesis telescope comprising 100 × 12-m antennas (collecting area 11,300 m²) covering an area 2 km E-W by 3 km N-S (see Fig. 2.7), and capable of wide-field imaging of low surface-brightness, complex, extended continuum radio emission and distributed spectral line emission, at frequencies between 408 MHz and 22 GHz (Dewdney & Landecker, 1991a).

Dewdney’s goal was to build a telescope with capabilities that was complementary to the VLA for low surface brightness extended radio emission particularly for the interstellar medium in our own galaxy and for neutral hydrogen (HI) in nearby spiral and irregular galaxies. The telescope would operate as a high-speed wide-field radio camera and with its high sensitivity it could observe HI in other galaxies to a modest redshift (~0.2).

Fig. 2.7

An artist’s impression of the Radio Schmidt Telescope concept

This was the first design study of what, later, became known as the Large Number-Small Diameter (LNSD) dish array concept now adopted for the mid-frequencies for the SKA, as we discuss in Sect. 6.4.

Despite widespread community support including an international workshop in 1989¹⁵ as well as numerous presentations by Dewdney including one at IAU Colloquium 131 in 1990 (see below), it was “tough sledding” in Canada, and the concept did not achieve sufficient traction to be funded.

2.2.2.5 Giant Metre-Wave Radio Telescope (GMRT), Pune, India (Operational 1995)

The GMRT was proposed by Govind Swarup in 1984, funded nationally in 1989, and came into operation in 1995 (Swarup et al., 1991). It has thirty 45-m diameter dishes (total collecting area ~ 50,000 m²) in fixed positions in a “Y” configuration with largest separation of antennas of 25 km (Fig. 2.8). It operates at frequencies from 100 MHz to 1.42 GHz and provides sensitivity at high angular resolution as well as the ability to image radio emission from diffuse extended regions. After the recently completed upgrade (Gupta et al., 2017), the GMRT will remain the most sensitive low frequency interferometer in the world until the first phase of the SKA comes into operation. As with its higher frequency counterpart, the VLA, the GMRT is used for the study of a wide range of astrophysical questions.

Fig. 2.8

Part of the Giant Metre-wave Radio Telescope, near Pune, India. (Credit: National Centre for Radio Astrophysics, Tata Institute of Fundamental Research)

Of particular note are the low-cost antennas based on an imaginative design by Swarup and colleagues called the SMART concept—Stretch Mesh Attached to Rope Trusses (Swarup et al., 1991). This concept returned to prominence a decade later in 2005 as one of the possible telescope designs for the SKA described in Sect. 6.4.4.3.

2.2.3 The First Square Kilometre Array: Grote Reber’s 2 MHz Array in Tasmania, Australia

It is worth remembering in this historical survey that the first extremely large collecting area telescope was designed and constructed near Bothwell in Tasmania in the early 1960s by one of the pioneers of radio astronomy, Grote Reber (George et al., 2015).

Fig. 2.9

An aerial view of the 2 MHz array taken in 1965, showing a significant portion of the array (courtesy Estate of Grote Reber, supplied by Dr. Martin George)

Reber set up a large array of dipoles (see Fig. 2.9), and over the period 1963–1967, mapped the sky at a frequency of 2.085 MHz. The array comprised 192 wire dipole antennas suspended on wooden poles spread at regular intervals over an area of approximately 1 km × 1 km, making it a 1 km² filled aperture array (George et al., 2015).

2.3 Global Collaboration in Radio Astronomy Pre-SKA

Astronomy is a global science with a long tradition of inter-institute and international collaboration, particularly in radio astronomy. One of the main motivations for this was very-long-baseline interferometry (VLBI) , a technique that combines telescopes separated by hundreds to thousands to tens of thousands of km—in the case of an Earth-orbiting element—into a single instrument to achieve the highest possible angular resolution. Initial experiments in 1967 involved pairs of telescopes in the US and Canada (Kellermann et al., 2020) (Chap. 7), and in 1968 in the UK (Lovell, 1970). But this expanded in scope to trans-Atlantic experiments (US-Sweden in 1968, and US-USSR and UK-Puerto Rico¹⁶ in 1969) and, by 1980, multi-station networks, operating part-time, had been established in the USA (1976) and in Europe (1980) making use of the existing telescopes (e.g. Thompson et al. (2017), Chap. 7 in Kellermann et al. (2020), Porcas (2010)). Trans-Atlantic networks became a regular mode of operation from the early 1980s.

Collaboration across institute and country borders lies at the heart of VLBI. In the early stages, agreements, initially at scientist level and later at institute level, were set up that enabled the required coordination to take place before and during observations. This ensured all participating telescopes observed the same point in the sky at the same time with the same band of frequencies. Operational management structures were set up to select proposals for observing time, provide the stations with detailed schedules and operational instructions for the observations, and transport the data from the telescopes to the central data processor. In addition, upgrades to VLBI-specific instrumentation at the telescopes were also carried out in a coordinated manner.

The US Network gave way to the 10-station NRAO Very Long Baseline Array (Fig. 2.10 left) in 1995 which operated mostly in self-contained mode but also in part-time network-mode with the VLA, Arecibo and Effelsberg, as well as with the European VLBI Network (EVN)). The EVN (Fig. 2.10 right) continued to grow beyond Europe to the east and west, encompassing 23 telescopes in 10 countries at the time of writing.

Fig. 2.10

Left: The US Very Long Baseline Array (credit: U.S. National Radio Astronomy Observatory). Right: The European VLBI Network (credit: NASA Earth Observatory, and Paul Boven, Joint Institute for VLBI-ERIC)

Other networks have been established in Australia, China, Russia, and east Asia, as well as global networks for geodetic measurements and for even higher angular resolution at millimetre wavelengths, the “Event Horizon Telescope” (EHT¹⁷). Radio telescopes were launched into earth orbit on two separate occasions, VSOP-HALCA (VLBI Space Observatory Program—Highly Advanced Laboratory for Communications and Astronomy) by Japan in 1997 (Hirabayashi et al., 1998), and RadioAstron by Russia in 2011 (Kardashev et al., 2011) to carry out observations together with the ground arrays to increase the angular resolution. Both these ‘space-VLBI’ missions had their origins in the 1980s and, in each case, were partnerships of the national space agencies and the global radio astronomy community. The HALCA-ground-based array observations were coordinated by the Global VLBI Working Group formed under the auspices of Commission J on Radio Astronomy in the International Union for Radio Science (URSI) (Gurvits et al., in preparation).

As time went on, those involved in VLBI gained an understanding of the different social and work cultures as well as the different scientific points of view in parts of the world other than their own. They also gained a belief that, with clearly stated scientific objectives and mutual benefit for all participants, supra-national projects could be undertaken successfully.

But even before the advent of VLBI, radio astronomers and engineers spent time visiting each other’s institutes, observatories and engineering laboratories, to learn from, and in turn give advice to, colleagues in the general interests of advancing science. Just as importantly, astronomers from anywhere in the world could obtain observing time on any radio telescope by writing proposals in regular open competitions. This “open skies” policy is an embodiment of the collaborative spirit, and is still a common dominator in many, but not all, radio observatories.

2.4 The SKA: First Ideas

The preceding pages make it clear that, in the 40 years after World War II, radio astronomers had not been averse to thinking on big scales, both in terms of individual telescopes and array sizes. With VLBI, they had also made international collaboration part of their scientific way of life. Blue-sky thinking about radio astronomy instrumentation and its impact on astrophysics and cosmology was also part of the culture. Lower frequency telescopes were easier to construct, so large versions of dipole arrays¹⁸ and parabolic cylinders saw the light of day earlier than larger dish telescopes for higher frequency work, with the exception of Arecibo. The arrival of the VLA in 1980 as the largest and most versatile dish array on the planet was a major step forward at higher frequencies, but that did not stifle continued thinking about even larger telescopes. In fact, the early ideas that eventually led to the emergence of the SKA in the early 1990s arose for the most part in the 1980s during informal coffee-time discussions at a number of institutes and observatories around the world, and developed in bursts throughout the decade, finally merging in the 1990–1993 period into the global effort to create what became the SKA.

A number of names can be associated with the crystallisation of the 1 km² collecting area concept in 1990—Govind Swarup in India, Yuri Pariiskii in the USSR, Peter Wilkinson in the UK, and Robert Braun, Ger de Bruyn and Jan Noordam in The Netherlands. Peter Dewdney in Canada led the first proposal for an array design that is now the basis of SKA-mid, but its use for a much larger collecting area than the VLA was not part of Dewdney’s thinking at this time (see Sect. 2.2.2.4). For Wilkinson, Braun and Dewdney, the initial “coffee table” was located at the newly operational VLA site where visiting astronomers from around the world were assured of a receptive audience from local scientists, other visiting scientists, and the inaugural Director (one of the present authors, Ron Ekers). Wilkinson, Braun and Dewdney were at the VLA at different times, and all three came away with ideas on how to improve on what was the world’s most powerful radio telescope.

In the US itself, the VLA was still too new for any energy to be spent thinking about a major upgrade, and with the Very Long Baseline Array under construction and the Atacama Large Millimetre and sub-millimetre Array (ALMA) on the horizon, no funds could be expected for the more minor VLA upgrades that were already under discussion—the short-baseline E-array to provide the low brightness sensitivity of the Radio Schmidt Telescope and the supercomputer processing capability to enable full spectral line observing.

2.4.1 Separate Bubbles of Activity

The early development of ideas on the SKA took place independently in India, the Soviet Union, the UK and The Netherlands, and occurred without any explicit interactions among the main players until Wilkinson (Jodrell bank Observatory, UK) and Noordam (ASTRON, The Netherlands) discussed the large collecting area concept at a coffee break during IAU Colloquium 131 held in Socorro, New Mexico in 1990—see Sect. 2.4.2. This conversation led to Wilkinson presenting a paper on “The Hydrogen Array” (Wilkinson, 1991a, b) in the last session on future visions for radio astronomy, a paper that is generally seen as the “light bulb” moment for the SKA.

Swarup, Pariiskii (Pulkovo Observatory, USSR) and Dewdney also took part in the conference, but it appears that little specific interaction on the large collecting area topic took place with Noordam and Wilkinson. However, Ekers recalls that the talks by Wilkinson and Pariiskii were the trigger for the radio astronomy community to start thinking about a future of a next generation telescope that might even be beyond what a single nation could achieve. This laid the ground for the collective action on a global scale which began to be taken only a few months later, as we describe in Sects. 2.4.2 and 2.4.3.

In the following four sections we describe what did these individuals brought to the table in Socorro in 1990?¹⁹

2.4.1.1 Govind Swarup, India, GERT (1978), GMRT (1984)

Fig. 2.11

Govind Swarup (Credit: National Centre for Radio Astrophysics, Tata Institute of Fundamental Research)

By the time the IAU Colloquium took place, Govind Swarup (see Fig. 2.11) had been a world leader in radio astronomy and in innovative radio telescope design for many years (Swarup, 2021). From the early 1960s, his main focus had been on the use of relatively cheap cylindrical paraboloids as a means of achieving a large collecting area (see Sect. 2.2.1). The Ooty Radio Telescope (ORT, Sect. 2.2.1.5) came into operation in 1971 as the centre-piece of Indian astronomy, and this led in 1978 to the Giant Equatorial Radio Telescope (GERT) proposal (see Sect. 2.2.1.6) for a telescope with at least ten times the collecting area of the ORT.

GERT was to be a project designed and developed by a group of “non-aligned” countries²⁰ led by India and envisaged not only a telescope, but also an associated new international institute (see also Orchiston & Pharkatkar, 2019). One of the key aspects of the GERT science case was the potential provided by a large collecting area for detecting neutral hydrogen in emission in the distant universe at an epoch before the formation of galaxies. This was the first formulation of what became the initial main scientific driver for the SKA.

The proposal was given support by a Resolution of the International Astronomical Union²¹ and was considered for several years by UNESCO and potential funding nations until the end of 1983 when, in exasperation, Swarup radically changed the concept into what became the dish-based Giant Metre-wave Radio Telescope (see Sect. 2.2.2.5), a second nationally funded telescope for India. Swarup (priv. comm. to R. Schilizzi, see also Orchiston and Pharkatkar (2019)) tells the story that he came up with the idea of transforming GERT into the GMRT while seeing in the New Year with a shot of whisky early in the morning of 1 January 1984. The trigger was a Christmas letter from Alec Little (ex-Molonglo Telescope Director and famed instrumentalist) on his investigation of the use of optical fibre for connecting the Australia Telescope antennas. The original idea for the GMRT was 34 cylinders of ~60 m × 50 m, in a 14 km Y-shaped array connected by optical fibre. The cylinders gave way to dishes in 1986 when he conceived the SMART concept for cheap parabolic antennas. As was the case for GERT, detecting neutral hydrogen in emission in the distant universe became a key scientific driver for the GMRT.

In 1987, the GMRT was approved, and the design completed in 1989. But even in this busy period, much larger telescopes were never far from Swarup’s mind. In 1991 he published an article on a concept for a 700,000 m² telescope comprising 160 × 75 m diameter SMART dishes (Swarup, 1991b), and in 1992, he wrote a paper on the use of 1000 GMRT 45 m diameter antennas (1 million m²) for SETI observations (Swarup, 1992). And in 1993 he played a key role, with Ekers, in setting the course for the global collaboration that became the SKA.

However, it was the GMRT success that he came to Socorro in 1990 to describe in his presentation at the IAU Colloquium.

2.4.1.2 Peter Wilkinson, UK, Hydrogen Array, 1985, 1990

Fig. 2.12

Peter Wilkinson (credit: Peter Wilkinson)

Peter Wilkinson (see Fig. 2.12) was a staff member at the Jodrell Bank Observatory (JBO) when he visited Socorro and the VLA for 3 months from July to September 1984 at the end of a one-year sabbatical at NRAO as a visiting scientist. He remembers a day when Leo Blitz from UC Berkeley showed him a state-of-the-art VLA image of neutral hydrogen in the nearby galaxy M51. Wilkinson’s immediate reaction was that it would be even better with ten times the angular resolution to match typical optical images. Discussing the day’s highlights with his wife, Althea Wilkinson, also an astronomer, during the drive back to Socorro later that day, he came to the conclusion that the VLA collecting area needed to be 100 times larger. This would provide images of HI in other galaxies with the same sensitivity per pixel as in current observations, but with ten-times smaller sized pixels matching optical observations.²²

Study Group for the Priorities for Astronomy in the United Kingdom for the Period 1990–2000: The Wilkinson Note, 1985

Wilkinson came back to this conclusion a year later in 1985 in connection with a study of the Priorities for Astronomy in the United Kingdom commissioned by the Royal Astronomical Society (RAS) and the Royal Society (RS) in 1985. The study was chaired by the then Astronomer Royal and JBO Director, Sir Francis Graham-Smith.²³

As input to discussions on radio astronomy priorities at Jodrell Bank, Wilkinson composed a Note on 5 July 1985 on his conclusions on a large collecting area telescope made the previous year while visiting the VLA. The two-page note was handwritten and has never been published before now.²⁴

Two key passages went on to motivate thinking on the SKA in later years: (1) “The evident success of Arecibo in astronomical terms, despite its restricted pointing capability, is a testament to the fact that its collecting area is roughly ten times that of the largest steerable paraboloids. If we could construct a radio telescope with ten times larger collecting area still, we could confidently expect to garner a rich harvest of new, and unexpected, discoveries; and (2) “Its primary goal would be imaging/detecting and hence determining the velocity and column density of atomic neutral hydrogen, which comprises ~90% of the (observable) matter in the universe.”

The original Note also includes annotations by Wilkinson, made during IAU Symposium 119 on Quasars in Bangalore in December 1985, following discussions with Subramaniam Ananthakrishnan from the Tata Institute for Radioastronomy about plans for the GMRT, plans that Wilkinson previously had not heard about.²⁵ In the annotation, Wilkinson notes that the GMRT decision to use dishes rather than cylinders (“dishes not dashes”) made him realise this would be a more flexible and simpler approach for the much larger array he had in mind. This was largely because a circular beam would mean simpler software and greater ease of handling confusing sources as a function of hour angle.

The final report by the RAS-RS Study Group in November 1986 included the Large Radio Flux Collector concept in a section on lower priority projects. The wording of this entry makes it clear that elements of the submissions by Wilkinson and other Jodrell Bank staff had been taken up by the Study Group, including the goal of detecting HI in emission, the possible use of cylindrical paraboloids, and the necessity of international collaboration for a project of this size. However, the idea of the very large collecting area of 20 hectares or more was a step too far, and a figure of ‘several hectares’ was all that was mentioned, equivalent to the GMRT which was mentioned in the report as a similar project.

Following the lack of any support for his proposal, Wilkinson transferred his attention to the higher Jodrell Bank priority of the extension of e-MERLIN interferometer to Cambridge. It was this that he came to Socorro in October 1990 to present at IAU Colloquium 131.

2.4.1.3 Robert Braun, Ger de Bruyn and Jan Noordam, the Netherlands, Large Telescope, 1989–1993

Fig. 2.13

Contemporaneous photographs of Robert Braun (left, credit: Robert Braun), Ger de Bruyn (centre, credit: Jan Noordam) and Jan Noordam (right, credit: Jan Noordam)

Robert Braun (Fig. 2.13 left) was a research associate and assistant scientist at NRAO stationed at the VLA from 1985 to 1989 before moving to ASTRON in The Netherlands as a staff scientist. He went to the VLA as a newly minted PhD graduate from Leiden University with the aim of extending his work on neutral hydrogen in our Galaxy to nearby galaxies. His experience using the Westerbork Synthesis Radio Telescope (WSRT) for galactic HI had shown that a spatial resolution of a few parsecs (~10 light years) was the goal to aim for in nearby galaxies. But the VLA did not have the surface brightness sensitivity to achieve that spatial resolution goal for HI emission, even though that was possible in principle with the interferometer baselines available [R. Braun, priv. Comm.]. This was the same conclusion Wilkinson had arrived at a year earlier, although Braun was unaware of this. Braun discussed this problem with Ekers in his role as VLA Director, as a number of other HI astronomers such as Leo Blitz and Carl Heiles had done, but there was little to be done to upgrade VLA capabilities at that relatively early stage of VLA operations, as noted earlier in this section. This was due in part to computer limitations restricting the number of line channels for HI observations, which meant improvement of image processing capacity had the highest priority in NRAO. However, the lack of pressure for increasing the sensitivity from the user community as a whole was a factor. Most users were very satisfied with the vastly improved continuum sensitivity, resolutions and image quality provided by the VLA, and there was no demand for more collecting area.

When Braun returned to The Netherlands in 1989, he encountered a small group at ASTRON including Ger de Bruyn (Fig. 2.13 centre) on the astronomical side and Jan Noordam (Fig. 2.13 right) on the engineering side who were thinking, mostly during coffee breaks, about a North-South extension to the Westerbork array to improve its imaging quality. As related by Noordam (2012), Braun thought increasing the collecting area of radio telescopes, perhaps as far as a million m², was far more important than the N-S extension. It did not take long for Braun’s idea to take off, underpinned by the results of a computer program for “rudimentary cosmology” written by de Bruyn that calculated the collecting area needed to detect the neutral hydrogen line in emission in distant galaxies assuming plausible masses of HI [R. Braun, priv. comm.]. This showed that 1 km² was needed for galaxies at redshifts of two. Initial thoughts on potential antenna elements for such a telescope focused on an array of a relatively small number of large diameter antenna elements.

Braun made a presentation on the Large Array in a talk on “Imaging the known universe in HI and CO” to a Dutch National Brainstorm on Radio Astronomy on 27 September 1990, just before the IAU Colloquium in Socorro the following month. [No copy of this presentation can be found.] Noordam went to the Socorro conference to present results on polarisation purity using the Westerbork Synthesis Radio Telescope, but also armed with these initial thoughts on large collecting areas in case an opportunity arose to debate them with the assembled experts on radio interferometry in Socorro. In this he was successful.

2.4.1.4 Yuri Pariiskii and Large Telescope Studies in the USSR, 1960–1991

Fig. 2.14

Yuri Pariiskii (photo taken in 1970, Credit: Pariiskii Family Archive)

In the USSR, first discussions of a 1 million m² collecting area telescope were initiated by Semyon Khaikin and Yuri Pariiskii (Fig. 2.14) in 1960 at the Pulkovo Observatory.²⁶ As a first step towards this, design and construction of a novel circular parabolic telescope concept 600 m in diameter and a few thousand m² collecting area, called RATAN-600, began early in the 1960s in the Caucasus under Pariiskii’s leadership. In 1964, a Super-RATAN project was proposed, 20 km across with 2 million m² collecting area (Pariiskii, 1992). This had potential as an early-warning system for Western missiles as well as for radio astronomy (L. I. Gurvits, priv. comm.) which is an interesting, but unsurprising, parallel with the US Sugar Grove 600-foot antenna, described in Kellermann et al. (2020), Chap. 9.²⁷ In the Super-RATAN case, astronomers were involved from the start whereas, for Sugar Grove, astronomers only became involved very late in the process.

In parallel with the Pulkovo activities in the 1960s and 1970s, other groups in the USSR in Pushchino and Kharkov took up the challenge of designing and constructing large, but not 1 km², telescopes at low frequencies, see Braude et al. (2012)—the DKR-1000 (see Sect. 2.2.1.4), BSA and UTR.²⁸

By the early 1980s, these and other facilities had reached their technical limits, and in 1982 Pariiskii used the moment to initiate the formation of a Working Group on the Square Kilometre Telescope (WG-SKT) (Gurvits, 2019) under the auspices of the Radio Astronomy Council of the USSR Academy of Sciences (see Fig. 2.15). Members of the WG were drawn from the main radio astronomy institutes/observatories in the USSR in Pulkovo, Pushchino, Kharkov and Gorky.

Fig. 2.15

Plenary session of the Radio Astronomy Council of the USSR Academy of Sciences in Pushchino, 1981. Front-row nearest the camera are: Nikolai Kardashev (astrophysicist and initiator of the RadioAstron Space VLBI mission) and Yuri Pariiskii. The astrophysicist, Iosif Shklovsky, one of the radio astronomy pioneers in the USSR, is the 3rd from the right in the second row. Both Pariiskii and Kardashev were Shklovky’s students. (credit: Pushchino Radio Astronomy Observatory)

The main science cases for the SKT included studies of the statistics of extragalactic sources (log N–log S), pulsars, and cosmological radio recombination lines. Neutral hydrogen in our own Galaxy and in other galaxies and the search for extra-terrestrial intelligence were also part of the science case. The SKT was one of the national large-scale radio astronomy initiatives, alongside a space-based interferometric array and a 70-m radio telescope for dm-mm radio astronomy on the Suffa plateau in Uzbekistan, both led by Nikolai Kardashev, Pariiskii’s colleague from the Space Research Institute in Moscow and a close friend since their study years at the Moscow University (Gurvits et al., in preparation).

The design of the SKT drew on the experience forged in the earlier telescopes as well as other non-astronomical installations such as low frequency phased arrays for over the horizon radar reception, and mid-frequency Luneburg lenses. (The latter re-emerged in the early 2000s as one of the contenders for the SKA reception element.²⁹) The most important output from the WG-SKT was a Council of Radio Astronomy White Paper in 1986 entitled “Radio Astronomy Aperture System for Metre Wavelengths: Design Study Notes” authored by Valery Bovkun et al. (in Russian) at the Radio Astronomical Institute in Kharkov. This described a 1 km² telescope operating from 30 to 330 MHz whose elements were somewhat similar to the present SKA-Low design and the Giant Ukrainian Radio Telescope (GURT, see Konovalenko et al. (2016)). At the time of writing, the GURT is in operation near Kharkiv as an extension to the UTR-2 telescope.

Fig. 2.16

The figure shows the collecting area of individual radio telescopes as a function of year from 1950 to 2000 on a log-linear scale showing that the collecting area of the largest telescope constructed In 2000 could be expected to be 1 km². (Credit: Fig. 3 from Pariiskii, Y. (1992), Radio astronomy of the next century”. Astronomical and Astrophysical Tronsactions, 1(2). 85–106. Reprinted with permission from the Eurasian Astronomical Society)

In 1990, at the 22nd and the last “All-Union” conference on Radio Telescopes and Interferometers in Yerevan, Armenia, three of the four groups represented in the SKT-WG published a paper on the activities towards the SKT (Bovkun et al., 1990). Pariiskii and his colleagues at Pulkovo were not among the authors despite Pariiskii’s leadership of the WG. According to Gurvits (2019), the Pariiskii group was more concerned with a proposal to upgrade the RATAN-600 telescope at that time. However, Pariiskii did show a diagram at the conference (Fig. 2.16) displaying the exponential increase in collecting area of radio telescopes as a function of year since 1955 and projections into the future. He concluded that it was inevitable that the largest radio telescopes would reach 1 million m² collecting area by 2000 if the trend from the first five decades of radio astronomy continued. Pariiskii also showed this diagram and related projections at the URSI General Assembly in Prague in August 1990 at the same meeting where Ekers gave a General Lecture on “The Invisible Universe” that included a plot of the evolution of radio telescope angular resolution with time and linked such plots to Livingston curves (see Sect. 1.2.3). This was based on analysis done in 1989–1990 as part of the Australian Decadal Review of Astronomy. Finally, Pariiskii showed the plot in a talk in the final session on visions for the future at IAU Colloquium 131 two months later in Socorro, the same session where Peter Wilkinson presented ideas on a 1 km² Hydrogen Array. However, no mention of the SKT project was made in either of Pariiskii’s talks in Prague or Socorro, and his IAU Colloquium talk was not published in the Conference Proceedings, so the rest of the world remained unaware of this long-standing intellectual effort taking place in the USSR.³⁰ Interestingly, VLBI observations between the US and the USSR Crimean antenna began in 1969 in the depths of the Cold War and continued with many other telescopes in global arrays throughout the period under review in this chapter, and beyond, see Kellermann et al. (2020), Chap. 7.

The non-publication of Pariiskii’s paper in the IAU Colloquium Proceedings was most likely for a simpler reason, to avoid the many months of tedious paperwork required for publication in a non-Soviet journal. It was submitted in January 1991 to Astrophysical and Astronomical Transactions: the Journal of the Soviet Astronomical Society (Pariiskii, 1992). Perhaps a more fundamental reason for Pariiskii’s reticence concerning details of ongoing and future radio astronomy plans in the USSR was the prevailing political system which prevented him from talking openly to western colleagues on this topic, and so avoid drawing outside attention to the potential for application of radio astronomy techniques and technology for other (military) objectives (L. I. Gurvits, priv. comm.).

The final activity related to the SKT project was a wide-ranging discussion of its SKT science case led by Pariiskii and Gurvits (then at the Space Research Institute, USSR) at the last “All-Union” Radio Astronomy Conference ever held, in Ashkhabad (Turkmenistan) in September 1991, and plans were laid for a focused SKT conference in 1993. This never came to pass as scientific activity slowed down considerably post-1991 following the dissolution of the Soviet Union, and the SKT did not regain any of the momentum it once had.

2.4.2 Lighting the SKA Torch in October 1990: IAU Colloquium 131 on Radio-Interferometry—Theory, Techniques and Applications

The first decade of VLA operation firmly established it as one of the key instruments for world astronomy. Its state-of-the-art engineering as well as its “open skies” policy of selecting the best proposals for observing time no matter where they originated, contributed to its success. The IAU Colloquium was held to celebrate the first 10 years of VLA operation and to show-case the astronomy results and the technical developments making them possible. A substantial number of papers in the formal program were on new concepts for telescopes and techniques or major upgrades of existing telescopes, including those by Govind Swarup on GMRT (Swarup, 1991b), Peter Dewdney on the Radio Schmidt Telescope (Dewdney & Landecker, 1991b), Peter Wilkinson on Phase 2 of the Jodrell Bank MERLIN interferometer (Wilkinson, 1991b), and Jan Noordam on High Accuracy Polarisation Measurements with the WSRT (Noordam, 1991).

At the meeting, there was a general feeling of great satisfaction with the progress made in radio interferometry over the previous 10 years, both in terms of the science and the development of new software and hardware techniques. There was no shortage of ideas for the new projects for the next decade. As noted above, only Jan Noordam and Yuri Pariiskii came armed with even more innovative ideas than those on the conference program. Noordam (2012) describes the meeting, somewhat tongue-in-cheek, as “suitably self-congratulatory”, but he felt there was a “hovering consensus that, after a dizzy ride, the heyday of radio astronomy was more or less over, and the next great strides would be made in other wavelength areas.” Whether that was a view widely-held at the conference is unlikely in the light of subsequent developments in the discipline, but it galvanised Noordam to float the recent ASTRON ideas on large collecting area telescopes to various people at the meeting. In Peter Wilkinson he found a very positive reception,³¹ over the inevitable cup of coffee. As Wilkinson describes it, being in Socorro again had brought back his earlier thoughts on the need for 100 times the VLA collecting area especially for HI imaging in distant galaxies, but it needed Noordam’s prodding to bring that to the fore. They approached Ekers and the other coordinators of the conference to include a talk by Wilkinson in the final special session on the future of radio astronomy at the end of the meeting. On the basis of Pariiskii’s URSI General Assembly talk 2 months before, Ekers also invited him to make a presentation in the session.

A few handwritten plastic overhead sheets sufficed for Wilkinson to lay out the general idea for “The Hydrogen Array”. As recalled by Bryan Gaensler (2012), then a final-year high-school student in Sydney but 17 years later SKA Project Scientist, Wilkinson’s first words were ‘The time is ripe for planning an array with a collecting area of 1 km²’. In the published paper (Wilkinson, 1991a), Wilkinson sketched possible scientific goals, foremost of these being neutral hydrogen imaging, and various technical considerations. On the scientific goals, he came up with a memorable statement:

The first task is to establish a clear set of goals. To my mind one goal stands out—a volume of the ‘Encylopedia of the Universe’ is written in 21 cm typescript. Unfortunately the printing is rather faint and we need a large ‘lens’ to read the text!

On technical considerations, he suggested 100 × 113 m diameter dishes, using the GMRT SMART design as a starting point. This was reminiscent of Barney Oliver’s Project Cyclops Project Cyclops of which Wilkinson was aware (priv. comm.), although no mention was made in the Proceedings paper.

His final comment was a call to interested parties to contact him so that these issues could be taken further. Only one comment after the talk was recorded in the Conference Proceedings, from Swarup, who said that the major question for such an array was to identify the outstanding science objectives since that would drive the antenna design and cost. It is worth noting that Wilkinson did not refer to his 1985 handwritten note in the Conference Proceedings since it had not been published (P. N. Wilkinson, priv. comm.)

Pariiskii presented the Livingston curve-like plots he had shown in 1990 at the “All-Union” conference on Radio Telescopes and Interferometers in Yerevan, Armenia and 2 months earlier at the URSI General Assembly in Prague (see Fig. 2.16) which were similar to those shown by Ekers in his General Lecture at the same meeting. As mentioned earlier, Pariiskii did not publish a paper in the Conference Proceedings and no questions or comments were recorded.

2.4.3 October 1990–August 1993, Interim Activities

Following IAU Colloquium 131, no direct heed was paid to Wilkinson’s call for collaboration. The time may have been ripe for a square kilometre array, but it certainly did not dominate the discussion at the Socorro meeting or internationally in the immediate years following. There was too much in the way of new and exciting data coming in from the VLA, the Australia Telescope Compact Array and VLBI, as well as other plans for new telescopes or upgrades. However, collective action on the large telescope concept did take place on a smaller scale within a few months of the Socorro meeting including publication of a number of papers.

Wilkinson published the Hydrogen Array paper in the Conference Proceedings but had no further involvement in large telescope ideas for the next 3 years as he continued working on e-MERLIN. Swarup was fully occupied with GMRT but found time to publish his Conference paper as well as papers on the International Telescope for Radio Astronomy (ITRA) and a proposal for an array of 1000 GMRT SMART antennas for the Search for Extra-Terrestrial Intelligence mentioned in Sect. 2.4.1.1. Pariiskii published his paper on “Radio Astronomy of the Next Century” in 1992 (Pariiskii, 1992), while the ASTRON group and others in The Netherlands continued to engage in thinking about antenna design for a large collecting area radio telescope and attempting to enthuse the Dutch astronomical community about the concept via presentations and proposals to national committees and an external visiting committee (the Foreign Advisors). This was the start of a decade-long dominance of the very large telescope landscape by the group at ASTRON.

In September 1991, Noordam, Braun and de Bruyn published an internal NFRA (Netherlands Foundation for Research in Astronomy) Note 585 on their thinking about large telescope concepts.³² This summarised the case for the wider Dutch astronomy community in the following words:

The next big step forward in radio astronomy will require a massive increase in collecting area. Ultimately, 1 km² will be needed for the imaging of individual galaxies in HI at high redshifts.

As an interim step, they proposed the EURO-16 array of 16 telescopes of 100 m diameter in The Netherlands, a maximum baseline of 15–20 km, and operating up to frequencies of 1.4 GHz to enable studies of neutral hydrogen in our own and more distant galaxies. ‘EURO’ stood for Early Universe Radio Observatory and was planned to carry out HI work but also studies of pulsars, variability in stars, supernova remnants and galactic nuclei, as well as non-thermal continuum emission from normal galaxies. Interesting to see in this Note is the considerable innovative thinking that had already gone into possible antenna types for the individual elements: a fixed spherical dish with movable focal point along a circular trajectory to enable source tracking; a fixed active-surface parabolic dish with a movable receiver location on a tiltable pole; a phased array; a scaled-up GMRT-type antenna; and a paraboloid suspended in a spherical shell suspended on a liquid or roller bearing and acting as an omni-directional mount (Fig. 2.17). The last of these concepts is reminiscent of the 600-foot Largest Feasible Steerable Telescope (LFST³³). All but the LFST analogue returned later in the SKA story in one guise or another, and sometimes in combination (see Chap. 6).

Fig. 2.17

Compendium of possible antenna concepts for EURO-16 (Credit: Jan Noordam])

An exchange of letters between de Bruyn, Braun and Noordam and Swarup in August 1991³⁴ shows that the Dutch contingent were eager to engage Swarup in a feasibility study of the EURO-16 concept. He in turn was interested but said he preferred 24 × 80 m SMART design antennas. In the event, the feasibility study was not pursued.

Discussion of the EURO-16 proposal in the Dutch National Brainstorm on Radio Astronomy in August 1991 and by the ASTRON Foreign Advisors in October 1991 led to “honourable mentions” in both cases. The Foreign Advisors Committee, chaired by Ekers, recommended that work should begin as soon as possible on “a new large project” to follow the ongoing upgrade of the WSRT with the new multi-frequency suite of receivers. Other contenders for the new large project were a large infra-red/optical telescope, a large array operating at mm/sub-mm wavelengths and a north-south extension to the Westerbork Telescope to enable useful observations to be made in the southern hemisphere and overlap with the new ESO Very Large Telescopes (VLT) in the optical domain.

Continued discussion of these options was followed up a year later with a proposal by Ed van den Heuvel and Jan van Paradijs (University of Amsterdam, The Netherlands) for a “Square Kilometre Radio Telescope”³⁵ for the study of the Sun and Stars, and in May 1993 by a paper by Robert Braun on ‘A concept for a new generation radio telescope for the frequency interval 150–1500’.³⁶ Braun’s paper discussed the range of science that would be possible and also focused on a technical concept of arrays of small (6–20 m diameter) dishes configured in groups of 200–300 m diameter spread over a region 30 × 50 km. Like the Radio Schmidt telescope proposal (see Sect. 2.2.1.4) this was a forerunner of the LNSD concept later adopted for the SKA.

As described in Sect. 3.2.6.1, in 2004, under the overall leadership of the ASTRON Director, Harvey Butcher, and project leadership of Arnold van Ardenne as Head of Research and Development, ASTRON began the development of the phased array concept for the large radio telescope mentioned in the EURO-16 proposal. A year later in 1995, the Dutch Government made the first substantial grant anywhere in the world (4.5 million guilders) for innovative technology research leading to a square kilometre array. We return to the Dutch story in Sect. 3.2.6.1 on national SKA efforts around the world and their integration into the developing formal international collaboration.

From 1992 onwards, Ekers gave several presentations at national and international meetings on the future of radio astronomy with emphasis on the new technologies that could continue the exponential increase in radio telescope sensitivity over time discussed in Sect. 1.2.3. This involved new technology for signal processing and receiver sensitivity as well as a progressive increase in collecting area. However, it was an intervention by Swarup and Ekers in 1993 that set the future course of the SKA as a global radio astronomy project from the outset.

2.5 SKA Is Born Global, August 1993

In May 1993, the European Space Agency convened a meeting on the Frontiers of Astronomy at the European Space Technology Centre (ESTEC) with an Organising Committee chaired by Malcolm Longair, then Director of the Royal Observatory Edinburgh, and including Ekers as a member. Swarup also attended the meeting. Ekers gave a talk on future developments in radio astronomy including the planned space-borne elements, VSOP-HALCA and RadioAstron, and the benefits of a ground station with an area of a square kilometre. Swarup was keen to discuss his proposal for the International Telescope for Radio Astronomy (ITRA, (Swarup, 1991a) with Ekers and they decided that the time was indeed now ripe to start collective action to realise a large collecting area radio telescope as an international effort.

Formation of a Working Group and finding a home for it in an international entity was the obvious first step. Equally obvious was the choice of the International Union for Radio Science (URSI) as WG home since URSI had a very active Radio Astronomy Commission (J) of which Ekers was Chair, and the WG activities would fit neatly into the Commission’s remit. As senior members of Commission J and URSI as a whole, both Ekers and Swarup felt it would be relatively straightforward to establish the WG at the next General Assembly of URSI in Kyoto, Japan in September 1993 to begin a worldwide effort to develop the scientific goals and technical specifications for a next generation radio observatory.

Neither Ekers nor Swarup wished to initiate the WG since their current institute activities did not allow sufficient time, and in Ekers’ case, he was also conflicted by his current position as Commission J chair and his future membership of URSI’s Long Range Planning Committee following the General Assembly. This led them to invite Robert Braun to prepare the case and chair the WG. At the General Assembly, all went according to plan, and the Large Telescope WG was duly formed with nine members—Robert Braun (NFRA, Netherlands, chair), Ron Ekers (CSIRO ATNF, Australia), Lloyd Higgs (DRAO, Canada), Yuri Pariiskii (SAO, Russia), Wolfgang Reich (MPIfR, Germany), Wu Shengyin (Beijing Astronomical Observatory, China), Govind Swarup (Tata Institute for Radio Astronomy, India), Dick Thompson (NRAO, USA), and Peter Wilkinson (Jodrell Bank Observatory, UK). Francois Viallefond (Meudon Observatory, France) was later added as a member. The WG remit is reproduced in Box 2.1.

Subsequent meetings of the working group provided a forum for discussing the technical research required and for mobilising a broad scientific community to cooperate in achieving this common goal.

August 1993 is regarded as the formal start of the SKA as a global project.

Box 2.1 Remit of the URSI Large Telescope Working Group, 1993

At the URSI General Assembly in Kyoto in September 1993, Commission J resolved to form a Large Telescope Working Group to consider:

• (a) The strong scientific case for a new, internationally accessible radio telescope with one or two orders of magnitude greater sensitivity than that of any existing or planned facility;

• (b) The need for innovative technical developments to realise such a facility at an affordable price;

• (c) The likely need for international collaboration to allow realisation of this facility.

  And resolved to appoint a working group with the following terms of reference:

• (1) To explore the range of scientific problems to be addressed by the instrument.

• (2) To discuss the technical specifications and general design considerations needed to maximise the scientific return of such a facility.

• (3) To identify and, in so far as possible, resolve the major technical challenges to realisation of an affordable radio telescope with the required sensitivity.