Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Space Science

A Telescope The Size Of The Earth 66

Posted by timothy from the why-not-even-bigger dept.
Neil Blender writes "From this article: "Astronomers have fashioned an Earth-sized virtual radio telescope that can distinguish celestial features 3,000 times smaller than the those observed by the Hubble Space Telescope. The device, which uses atomic clocks and a custom supercomputer to link together radio dishes on three continents, is the most powerful radio observatory ever, according to scientists." Some parts of the custom supercomputer use linux and IDE RAID."
This discussion has been archived. No new comments can be posted.

A Telescope The Size Of The Earth More

A Telescope The Size Of The Earth

Comments Filter:
  • I wonder how much longer before we will be able to pick out individual geographic features on remote planets? 3000 times better resolution than hubble might actually give us real views of remote plantets.

    I'd love to take a geography class in an astronomy major, discussing the geography of Betelgese-124 ;)

    • Re:Resolution and limits? (Score:2, Informative)

      by Anonymous Coward
      This is a radio telescope and planets aren't radio sources (although civilisations on them may be)
      • Yeah? Well, the hubble is a optical telescope, and planets aren't optical sources. They can reflect light though.
        • Can't we get radio signals that are bounced off the planet's terran from the planet's star, kind of like radar?
          • Such "passive" reflections would be far too weak (plus you'd have to make a variety of assumptions about the "geography" of that solar system since you wouldn't be getting direct perpendicular reflections from the star). You'd need to do something akin to active radar (i.e. we'd have to shoot a high intensity beam of radio waves at the target). Targetting may not be a big issue at that distance since it's likely when the radio waves reached their destination they'd spread to the size of the target solar system or beyond. But, the amount of energy remaining at the target would be low (what, inverse R-squared, or something?) and any amount that got reflected back would be even smaller. Plus, you'd have to account for the fact that you've changed position considerably between the time you originated the signal and the time you received the echo. You'd also only get echos that bounced in the direction the Earth would be at years later - you wouldn't get the echos that bounced directly back.
    • Also, since this is a radio telescope, you're not going to see optical pictures of planets out of them.

      They'll be nice to look at, say, jets coming off neutron stars and such at higher resolution, or things that are major radio sources already. I doubt that they can pick up faint radio sources (powerwise) because the antennas they're using are definitely not earth-sized.

      This is also why some large ground telescopes can see farther than hubble because they have a larger light collecting area.

  • Yes, but (Score:4, Funny)

    by llamalicious ( 448215 ) on Wednesday October 02, 2002 @04:09PM (#4376091) Journal
    Remember, this is a RADIO telescope. Not optical.
    While I don't doubt the value a radio telescope might have for planetary research, I'm willing to bet you're thinking about something akin to being able to see the individual cells on Pathfinder's solar-array on the surface of Mars from a telescope mounted here on Earth.

    Anyone know the _optical_ resolution for maximum "zoom" on Hubble...?

    • Re:Yes, but (Score:2, Informative)

      by ndevice ( 304743 )
      resolution is a function of the aperture size of the telescope. In hubble's case, this is 2.5 meters (according to space.com). I don't remember what this function is, but it has to do with the wavelength of the signal too. Something along the lines of d/lambda. And you get a resolution of angle out of it because you don't know the distance to your target.

      Note that this will be the maximum theoretical resolution, Hubble is probably less. (and that I could be wrong, but I'm pretty sure about the aperture to resolution relation).
      • but then I could check google and find out that hubble has resolution of "less than 0.1 arcseconds"

        http://hubblesite.org/sci.d.tech/nuts_.and._bolt s/ res101.shtml

    • Re:Yes, but (Score:2, Interesting)

      by Tablizer ( 95088 )
      (* thinking about something akin to being able to see the individual cells on Pathfinder's solar-array *)

      Or the Mars Polar Lander debri field. I am curious to know what happend to the bugger, not just speculation.

    • Hubble and radio telescope resolutions. (Score:5, Informative)

      by Christopher Thomas ( 11717 ) on Thursday October 03, 2002 @12:11AM (#4378823)
      While I don't doubt the value a radio telescope might have for planetary research, I'm willing to bet you're thinking about something akin to being able to see the individual cells on Pathfinder's solar-array on the surface of Mars from a telescope mounted here on Earth.

      Anyone know the _optical_ resolution for maximum "zoom" on Hubble...?

      A really good telescope will usually be limited by diffraction effects (the fact that the telescope is of finite size causes light being focused to blur out a bit as it passes through the telescope aperture).

      A back of the envelope calculation suggests that the diffraction-limited resolution of the Hubble (at 2.5m) for 500 nm light is about 0.2 microradians (letting it resolve features ten million kilometres wide at Alpha Centauri, five light-years away [give or take], or letting you read a typewritten letter at 5 km).

      A radio telescope typically operates on wavelengths on the order of a tenth of a metre (as a gross approximation; it's really an order of magnitude in either direction from there, if I understand correctly). The largest radio telescope dish on the planet is about 300m wide, giving a diffraction-limited resolution of about 0.3 milliradians, or about three times sharper than the unaided human eye is at optical wavelengths (the equivalent of reading a typewritten letter at about 10 feet).

      An interferometric radio telescope with an aperture the size of the planet would have a resolution of about 10 nanoradians, letting it resolve features about 0.5 million kilometres wide at Alpha Centauri [slightly smaller than our sun] (the equivalent of reading a typewritten letter at a distance of 100 km, or the title on a paperback book from low Earth orbit).

      If we had a radio telescope with an atomic clock on the moon (about 400,000 km away), we could resolve objects the size of Jupiter in the Alpha Centauri system. If we had a space-based radio telescopes in the Earth-Sun L4 or L5 points (each 150 million km from Earth), we could resolve individual cities on an Earth-like planet.

      This is cheap enough to do that we're probably going to put radio telescopes there within the next couple of decades. Any planet with a magnetosphere within 50-200 light years would be detectable, and we'd have detailed maps of magnetic effects on the surfaces of every star within a thousand light-years.

    • Anyone know the _optical_ resolution for maximum "zoon" on Hubble...?

      There was a relevant picture [nasa.gov] at The Astronomy Picture of the Day [nasa.gov] a few months ago that mentioned this issue:

      With its 2.4 diameter mirror, the smallest object that the Hubble can resolve at the Moon's distance of around 400,000 kilometers is about 80 meters across.
      (and so it can't make out the lunar modules that are there)
  • and then we can have a telescope cluster even BIGGER than the Earth. I believe I've heard of proposals to do just this. Actually, I think it was to have multiple telescopes orbiting Earth's L4 or L5 points, much bigger than the Earth.

    Kind of reminiscent of the question, "How about a Beowulf cluster of these?" only applied to telescopes.
    • Hubble == optical telescope, these are radio telescopes. AFAIK this technique doesn't work on optical observations.
      • I thought I remembered something about doing something similar optically. How do they combine the Kecks output?
        • the light shoots underground to a point underneath Keck II, where they are joined. Keck also is planning a set of 4m outriggers to increase the light collection and hopefully be able to directly image extrasolar planets(!) which is really cool because then they can look at the spectra for organic chemicals (well, probably not Keck, but in principle).
      • AFAIK this technique doesn't work on optical observations.

        It is theoretically possible IIRC, but just much tougher than radio waves. The timing is very important to putting the signals back together properly in the computer. Radio waves are less dense than light-waves, and thus you have more tolerence of timing errors.

        Thus, someday we may be able to do such with light, but for now it is beyond our technology (except at close range).
        • This is correct - the process is based on wave interference and there are two ways to accomplish it. First - you can cause the waves from two sources to actually meet and interfere - as is the case with Keck. The other is to digitize the waves and simulate the interference with computers. To digitize the waves means that you have to sample at a rate at least twice the frequency of the wave, and you also have to get the phase right which means that you have to record the time along with the wave.

          For radio this is difficult enough - radio is measured in the MHz to GHz range - so you have to sample at a GHz frequency and record time to the billionth of a second. For light you are talking wavelengths in the 400nm range - which is a frequency of c/400nm=749 terrahertz. I don't know of anything which samples in quite that range (that would be HIGH!!!). Also, you need to record time in the trillionth of a second range (picoseconds). That probably isn't all that easy either (I don't know much about atomic clocks, but I think that is feasible, though you have to sync the time recording to the data recording, which isn't easy).

          Light is a whole lot tougher to record than radio as the frequency is outside the range of modern electronics to handle.
          • For light you are talking wavelengths in the 400nm range - which is a frequency of c/400nm=749 terrahertz. I don't know of anything which samples in quite that range (that would be HIGH!!!). Also, you need to record time in the trillionth of a second range (picoseconds). That probably isn't all that easy either (I don't know much about atomic clocks, but I think that is feasible, though you have to sync the time recording to the data recording, which isn't easy).

            At two samples per cycle, you need better than femtosecond (10^-15 second) resolution.

            Atomic clocks are stable to one part in 10^14+, but that doesn't mean we can measure time in femtoseconds. Just that drift between two clocks will be less than ten femtoseconds per second.

            You can make a strong circumstantial case for direct waveform sampling of visible light being outright impossible with machines built from normal matter, as the relaxation time of most electron state transitions is longer than the required sampling rate, meaning that there would be no way for any possible device to switch fast enough Various forms of exotic matter have faster state transitions, but making measurements from the surface of a white dwarf or a neutron star is difficult :).

            OTOH, I can believe direct sampling of far-infrared, which would at least give many orders of magnitude better resolution than radio telescopes, with baselines longer than are practical for optical interferometers.
    • Hubble is a multi billion dollar project. multiple ones would never get sent up, also Hubble is so low in orbit that it heats up. To git it to any Lx stable point would take probably a disposable rocket, such as the ones made by Boeing
      • Last I heard, and mentioned by a co-reply by trixillion, there is a followon for the Hubble planned. The really good NASA missions have long exceeded their planned lifespan, so it's entirely possible that there will be operational overlap between the Hubble and the Hubble++. There would be an opportunity to team them, perhaps. I don't know what extra infrastructure is needed to do this.
    • The Next Generation Space Telescope is planned to be placed in L2 orbit.
    • I know that they have optical telescopes that do this now. Keck Observatory come to mind. Here [spie.org] is a link to a page that describes what they are doing and the resolution they are getting. The VLT [eso.org] (Very Large Telescope) is another example of combining the beams of multiple telescopes.
    • Admittedly I didnt read the article, but I wouldnt be suprised if the this "Earth sized" telescope is used for that purpose.

      And interferrometer takes two images at a wide distace apart and then one image is used to cancel the other. The remaining image is a result of the slight angle difference, allowing you to see very dim objects next to bright ones. The same could be used for radio waves.

      In any case this would be tough, but not impossible job, for actually seeing a planet. I remember one scientist literally describing it as trying to see a firefly with a nuclear explosion as a backdrop.

  • Alternative reading (Score:2, Informative)

    by cbv ( 221379 )
    Check the article [spaceflightnow.com] at SpaceFlight Now.
  • Whoaaaaa momma! Look at that telescope! That sucker's huuuuuuge!
  • SETI already uses this approach -- multiple radio telescopes in multiple locations, coordinated analysis of simultaneous observations -- in their systems. They just don't do it in real-time.

    • Actually, I'm wondering why they need to coordinate the analysis in realtime. What's the factor that makes realtime analysis and the bandwidth needed for that necessary?
      Pulling the data together after the fact and then analyzing it shouldn't result in a different conclusion of facts based on the data.
      • That is a wonderful point. SETI doesn't need realtime analysis and observation, and AFAIK they don't have that capability. Lets all just hope some crazy benefactor doesn't get drawn into their scheme and donate money for them to implement something of this nature.
  • This is an interesting use of the term "supercomputer". Traditionally this has been reserved for a machine with tremendous number crunching ability (a CPU(s) characteristic). Here the super reference relates to data rates. Kewl ...

  • Imagine a cluster of these.........oh wait, they *are* a cluster.
  • The beginning of the universe, galaxies 10^100 light years away...yada yada yada.

    All I care about is can it see the pornography on the nearest planet inhabited by intelligent perverts like us?

  • Resolving Golf Balls (Score:4, Funny)

    by Optical Voodoo Man ( 611836 ) on Wednesday October 02, 2002 @07:42PM (#4377641)
    From the article:

    "The resolution achieved by this telescope is the equivalent of sitting in New York and being able to see the dimples on a golf ball in Los Angeles," astronomer Sheperd Doeleman said this week. "

    I'm surprised an astronomer would say that. Most of them know that the earth is round. Seeing the L.A. golf ball would be really tough. I don't believe they can see through dirt, and even Tiger Woods couldn't hit it high enough. I felt the bodies of a thousand astronomers flinching in their graves.

    • Surface of the earth is curved. Space is curved. I don't see the problem. Anyway, this was just an example that is understandable to your ordinary person. Obviously it's not something they could actually do, as you could never see through all the LA smog.

  • contentless article, nothing new (Score:2, Informative)

    by Anonymous Coward
    the technique is several decades old and the article is essentially contentless.

    array of radio telescopes scattered around the whole earch can act like a single telescope if we combine the signal coherently. this can be done by connecting them together in real time (e.g. VLA in New Mexico) or offline (e.g. VLBA). Both VLA and VLBA are run by NRAO [nrao.edu]. if you want to do offline, you need to preserve amplitude, phase and timing information. atomic clocks are used for time stamping and supercomputers are used to combine signal. this simulates a large telescope whose lens is mostly opaque, save for few dots! the technique is really old (originated in late 60s, early 70s) and has been well mastered for more than 20 years. so there is absolutely nothing new in the article. depending upon the scientific relevance, many global telescopes participate in these experiments. in some cases, only US telescopes (VLBA which is scattered from hawaii to virgin islands) participate, which creates somewhat smaller effective telescope size. some experiments have beed done using space based radio telescopes which increases effective size even further (by the time, the space telescope became operational in late 90s, i left astronomy, so don't know the results).

    the supercomputer which combines signal is made up of custom chips and uses custom OS. linux part is quite small and it comes into picture well after the data has been pre-processed (raw data sizes could be few terrabytes a day while processed data is only few gigabytes).
    • One important advance is the use of 2mm wavelength
      radiation (as opposed to VLBI which has a maximum
      of about 2cm). So we've pushed the resolution up by
      a factor of ten (with admitedly poor baseline coverage) and as a result the hardware has to sample
      10x as fast.

      The `new' is in the hardware, not the technique. Lets not forget that the first sub-mm (0.3-1.3mm)
      interferometers are fairly new instruments, indeed
      they are the `last' wavelength to "discover" interferometry.

      ali
  • If you were to build a space-based radio telescope and nudge it over to one of the earth-sun trojan points, you'd have a 1 A.U. baseline. Is this technically feasable? How well does this sort of thing scale?

  • I'm impressed! I just spent a couple hours setting up an Adaptec 1200A IDE RAID 0, only to find out that it runs as fast as the individual drives did, about 40MB/s.
  • isn't there more than 3 continents in this particular Earth or you're talking about another Earth that I don't know? You should name this article "Not-So-Earth-Sized Telescope"
  • Radio observatories have detected fast-moving particle streams gushing from the cores of many such galaxies for millions of light-years.

    For millions of light-years - yes that *is* a long time....

    Just like 10 kg is indeed very fast.

  • MIsleading headline (Score:2, Informative)

    by Hays ( 409837 )
    Properly done interferometry can make it so this telescope array approaches the angular resolution of a telescope that was actually X thousand miles across, but it of course doesn't give it the same light gathering capability (sensitivity). So it won't be able to see anything new, only resolve much better what telescopes could already see. Important, but really only half the benefit of building larger telescopes.

Slashdot Top Deals

Intel CPUs are not defective, they just act that way. -- Henry Spencer

Close