Radio Renaissance : where’s the dog ?

Radio astronomy is undergoing a renaissance. How did this happen ?

This question is on my mind as I have just returned from sitting on an evaluation comitttee for ASTRON, the Dutch institute for radio astronomy, home of WSRT and LOFAR. Of course it wouldn’t be appropriate for me to say anything about that review, but it won’t hurt to say how exciting it was to actually visit LOFAR in the flesh. When I say flesh I mean mud and wire of course. Its a fascinating mixture of the crude and high-tech. The latter is course all the electronics and fibres and giant computers and software. The crude bit  is the antennae. From pictures I had thought the high-band antennae were strange gleaming tiles, but actually they turned out to be tarpaulin covering polystyrene boxes containing bits of metal that looked like they had been cut out with a Stanley knife. I love it.

Anyhoo. With LOFAR, EVLA, and e-MERLIN happening right now, ASKAP and MeerKAT on the way, and SKA driving its way through the funding agencies, stuff is happening in radio astronomy. A few years back I used to like visiting Jodrell Bank because it was so delightfully 1950s. It was like stepping into an episode of Quatermass and the Pit. Now the chaps at Jodders have upped sticks and moved into a gleaming new building in the centre of Manchester. The radio astronomers themselves used to seem like gentleman amateurs. Now they are all ambitious and thrusting. The SKA project is a tightly controlled PR machine. (I could tell stories but I won’t). Crumbs – how did that all happen ?

Well, they got their act together – scientifically, technically, and politically. But what surprises me is this. Usually what drives scientific changes is the availability of new technology. We all like these days to boast about knowledge transfer, and radio astronomy is nicely embedded in telecommunications technology of course – but exciting things happen when the technology is transferred from industry or the military to science, rather than the other way round. Examples that spring to mind : IR astronomy jumped into the new age when IR arrays fell off the back of the military lorry; optical astronomy was transformed after the CCD was invented; X-ray astronomy got serious when astronomers learned how the guys at Lincoln Labs could make three-axis stabilised spacecraft that could point. What was the new technology we jumped on to make the radio astronomy revolution ?

Put another way… normally we are the flea on the dog. Where’s the dog ?

22 Responses to Radio Renaissance : where’s the dog ?

  1. Where were CCDs used before they were used in astronomy?

    When Tony Tyson was at Bell or Lucent or wherever, he was doing astronomy, despite his employer, so there was some synergy from early on here.

    With regard to advances which have helped radio astronomy: I plead not being an observer. However, maybe high-speed data networks (fibre optics etc) play a role. Of course, that is driven by the internet, which is driven by the WWW, which was invented at CERN.

    Are we really often the flea? Where are the cameras with as many pixels as on the modern orbital observatories?

  2. “A few years back I used to like visiting Jodrell Bank because it was so delightfully 1950s. It was like stepping into an episode of Quatermass and the Pit. Now the chaps at Jodders have upped sticks and moved into a gleaming new building in the centre of Manchester. The radio astronomers themselves used to seem like gentleman amateurs. Now they are all ambitious and thrusting.”

    Both statements are true, but I think we have a case of correlation not implying causation here.

    When I visited both the old Jodrell Bank Observatory (or NRAL to us oldtimers) and the new premises last summer, I was struck by a profound sense of loss. It is really sad. Of course, many (mainly optical) observatories have moved most or all of their staff into town to be with the other physicists (Hamburg is one of the few places where astronomers are still at the observatory, though fewer observations are done there and some buildings are hired out for other purposes), but has the promised synergy really produced anything interesting? An observatory is a place of wonder, which new premises aren’t.

    • Halfway during my PhD they move us from Jodrell Bank into Manchester. The people out in the new building are truly fantastic, but I felt we really lost something in the move. Appart from the wonder of working beneath a giant telescope, it became much hard work with the e-MERLIN engineers and operators (i.e. the people that really understood the equipement).

      I miss Jodrell 😦

  3. Supercomputing – Radio astronomy has always been photon-rich and collecting radio signals is easy. Unfortunately most of the information has to be thrown away as our signal processing tech. isn’t up for the job.

    The ‘new renaissance’ is a direct result of off-the-self components like FPGAs, computer clusters and high-speed networks. Telescopes like LOFAR even market themselves as “software-telescopes”. With a fully digital beamformer you can now even make a new telescope by plugging in few extra components from your favourite industrial partner.

  4. andyxl says:

    Phillip – astronomy was indeed pretty much the first area to capitalise on CCDs, but they didn’t invent them. We jumped on them. Exactly my point. Anthony – half convinced. Supercomputers have been around for ages, as have fibre links. Beamforming has been used by the satellite industry for ages (but at much lower data rates of course). I have the impression that what has changed is a determination by radio astronomers to use these things. But maybe I am wrong – has a distinct new technology been invented which has made radio astronomers say “oooo yes, I’ll have that” ?

    • I’m not sure what it has taken so long, but what really excites me is where this dog is going. The SKA has been [past tense politically intended] a thought experiment on what you could build with the new tech. There is no major JWST-scale hardware that needs developing: it’s all modular and mostly depends on HPC.

  5. Martin says:

    For the EVLA and MERLIN the key thing is actually the fibres — as a by-product of the comms industry it’s possible to do this affordably in a way that it wouldn’t have been a decade ago and with enough bandwidth to make it worth doing. Plus, of course, enough computing oomph to correlate it when you’ve got it, but I think you’re right to say that this is secondary.

    To some extent this is true for LOFAR as well — the idea of fields full of low-tech antennas certainly isn’t new, and it’s not like LOFAR’s correlator is particularly cutting-edge in 2011…

  6. Stuart says:

    I’d say money was the reason why it is now. As time has gone on the technologies have become accessible for large astro projects because they have become affordable (and more capable). Ten years ago using huge amounts of fibre and (equally capable) supercomputers would have been enormously more expensive.

    I think it is true to say that some SKA capabilities may rely on computers to be cheaper/faster by the time those bits are finally built.

    • I think somewhere I read that the SKA would produce a petabyte a second, though I’m not sure at what stage. In any case, that’s a lot of data processing.

      • andyxl says:

        Phillip – LOFAR already produces about 0.1Pbs, but thats raw telemetry at the station. On-station electronics cuts that down to about 100Gps total that goes to the computer in Groningen. After more mangling there, they can only store about 2Pbyte/yr of results…

      • Maybe I can sell them some of my old Exabyte tapes. 🙂

  7. Tony Foley says:

    Mostly FPGAs & graphics processors plus large scale cluster computing

  8. Albert Zijlstra says:

    The SKA is based on a simple idea. Radio astronomy had been stuck at the same sensitivity because apertures weren’t increasing. The SKA is a 10-fold increase in aperture size, which is the kind of improvement that re-invigorates a field. There are other ways to increase sensitivity: bandwidth or aperture arrays, and these are also being developed.

    Radio astronomy was traditionally rather cheap (e-Merlin is so cheap I can’t believe it) , but once people got into the mindset that it is ok to ask for lots of money, the way forward was clear. Of course, this was in the good old days when money still existed. All the other things mentioned – HPC, data transfer rates, optical fibres, correlators, all help but followed after the ambition had grown.

    • ian smail says:

      i also think “radio astronomers” in the late-90’s began to realise that there were interesting questions to ask which didn’t just revolve around FRI/II’s.

      …and that in turn started everyone else getting more interested in VLA/MERLIN/etc.

      • Albert Zijlstra says:

        The original SKA idea was, I believe, based on old-fashioned radio astronomy – business as usual, just bigger. Once the ambitions had grown, people began to realize that this alllowed not just the same science better, but better science.

        Galactic radio astronomy (more my field) has always been diverse. But radio emission is weak – often you could get better sensitivity by moving wavelength (blueshifting). Luckily, radio is cheap per unit collecting area. But imagine what a square kilometre optical telescope (the SKOT) could do! And than imagine what it would cost.

  9. Duncan Lunan says:

    Radio astronomy in the UK got its real start using leftover World War 2 equipment that was developed to detect incoming V2s (see Bernrad Lovell’s autobiography “Astronomer by Chance”). The UK government didn’t want Jodrell Bank at all (Lovell, “The Story of Jodrell Bank”) and has tried several times to close it. Arecibo was built to military specification, ten times larger than they really needed, and there was an attempt to close that recently. But for those of us born in the 1940s, the real puzzle was, why the upsurge in optical astronomy c.1980? Apart from a failed instrument in the USSR, the only major new telescope in our lifetimes was Mt. Palomar and that was a prewar project, finally completed in 1948. Suddenly everybody was so keen for new deep-sky astronomy that Mt. Wilson was threatened with closure so the budget could be re-invested in Chile. Why did Mrs. Thatcher allow commitment to the Isaac Newton Telescope, UKIRT and AAT, at the same time as she was trying to get us out of CERN and the ESA science budget because they didn’t offer a commercial return?

    • Mark McCaughrean says:

      There may be other reasons, but at least one was identified earlier in this thread: CCDs (or electronic/digital detectors in general).

      Building a telescope a factor of two bigger than Palomar was hard, requiring innovative thinking and technologies as subsequently employed in Keck, VLT, Gemini, LBT, etc. Even then, you’d only get a factor of four more photons and go a factor of two deeper in the background limit compared to a 4-5m telescope.

      But with a digital imaging detector, you could suddenly gain a factor of 25-50 in detected photons compared to a photographic plate, yielding gains of 5-7 in sensitivity per unit time in the background limit and (perhaps more tellingly) factors of 25-50 in detector-noise limited medium- to high-resolution spectroscopy.

      OK, CCDs couldn’t compete in those days with photographic plates in terms of field-of-view (even if they can now), but for some science, you just had to go deep / faint, even if only over small fields.

      Of course, then we did figure out how to build 8-10m telescopes as well, so double win. But it’s telling that now that our detectors are large, close to 100% in quantum efficiency, and low in read-noise, it’s generally accepted that the next big ground-based telescopes need to be 3-4 times the diameter of the current generation to ensure major advances, not the traditional factor of two between earlier generations. And even then, they need AO to dig even deeper …

      (OK, I’m ignoring the other approach, namely taking a 4-8m telescope and paving a huge focal plane with detector for surveys, but that also depends on high QE detectors for gains over photographic Schmidt telescopes, say).

      That doesn’t answer your question re: Maggie (there are so many questions there that will likely remain unanswered forever), but I do think that optical (and later IR) astronomy were thoroughly reinvigorated by the digital revolution.

  10. Megan says:

    Reminded me of Jonathan Swift (I think):

    So nat’ralists observe, a flea
    Hath smaller fleas that on him prey,
    And these have smaller fleas that bite ’em,
    And so proceed ad infinitum.

    Who are the fleas on radio astronomy? 🙂

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: