≡ Menu

A Telescope Eight Times the Diameter of Earth

If you’re looking for detailed imagery of a distant astronomical object, VLBI (Very Long Baseline Interferometry) can deliver the goods. As witness the image below, which the National Radio Astronomy Observatory (NRAO) is calling “the highest resolution astronomical image ever made.” Here we see radio emission from a jet of particles moving close to the speed of light. The particles are being accelerated by a supermassive black hole at the core of the galaxy BL Lacertae, a highly variable ‘active galaxy’ some 900 million light years from the Earth.


Image (click to enlarge): Signals from 15 ground-based radio telescopes, combined with data from the RadioAstron orbiting satellite, produced the highest resolution astronomical image ever made. Credit: Gomez, et al., Bill Saxton, NRAO/AUI/NSF.

What fascinates me about this work is the technique. Very Long Baseline Interferometry works by collecting a signal at multiple radio telescopes, the distance between them being calculated from the different arrival times of the radio signal at the different installations. The calculations are made using a local atomic clock, with the data correlated with data from other antennae in the array to produce an image.

The upshot is that VLBI methods allow us to work with an emulated ‘telescope’ that has a size equal to the maximum separation between the constituent telescopes in the array. Conventional interferometry does much the same thing, but VLBI does not demand the physical coaxial cable or optical fiber connections of the former, allowing for greater distances between the telescopes.

In this case, we have fifteen ground-based radio telescopes combining their data with signals from the Spektr-R satellite of the RadioAstron mission, a Russian spacecraft with a 10-meter radio telescope. Launched in July of 2011, Spektr-R can work with large Earth-based radio telescopes to form interferometric baselines that extend up to 350,000 kilometers. In this case, working with radio telescopes in Europe and nine antennae from the Very Long Baseline Array in the US, the image produced was one with the resolving power of a telescope 100,000 kilometers wide, roughly eight times the diameter of the Earth. From the paper:

RadioAstron provides the first true full-polarization capabilities for space VLBI observations on baselines longer than the Earth’s diameter, opening the possibility to achieve unprecedentedly high angular resolution in astronomical imaging. In this paper we present the first polarimetric space VLBI observations at 22 GHz, obtained as part of our RadioAstron KSP [Key Science Program] designed to probe the innermost regions of AGN and their magnetic fields in the vicinity of the central black hole.

BLLacOort_nrao (1)

Credit: Artwork by Jon Lomberg.

Above we see the jet from BL Lacertae fitted within a diagram of our local interstellar neighborhood. Notice that if we take the Oort Cloud as the outer perimeter of our Solar System, the jet just about fits within the system or, if we imagine one end of the jet at the outer edge of the Oort, the jet would extend almost to the Alpha Centauri stars. The detail we can see here is roughly equivalent to seeing a 50-cent coin on the Moon. The image elongation results from the greater resolving power in the direction of the satellite as compared to the ground telescopes, with resolution in the orthogonal direction being exaggerated to compensate for the effect.

As for BL Lacertae itself, it was originally thought to be a variable star within the Milky Way after its discovery in 1929, but was later identified as a bright radio source. Subsequent redshift measurements showed a recession velocity of 21,000 kilometers per second, implying a distance of 900 million light years. BL Lacertae is considered an AGN, or active galactic nucleus galaxy, the latter being a designation for a galaxy with a compact region at its center that has a much higher than normal luminosity over at least part of the electromagnetic spectrum.

The paper is Gómez et al., “Probing the innermost regions of AGN jets and their magnetic fields with RadioAstron. I. Imaging BL Lacertae at 21 microarcsecond resolution,” Astrophysical Journal Vol. 817, No. 2 (26 January 2016). Preprint available.


Comments on this entry are closed.

  • Gregory Benford February 2, 2016, 14:07

    Wow, this idea surfaced in the 1980s, back when I was using observing methods at the VLA on jets. I’m gratified that it has been worked out, complete with the timing-distance calibrations. Now we can peer into scales that can reveal intricate structures in other galaxies!
    Maybe I should go back into modeling galactic jets…

    • Jose L. Gomez February 5, 2016, 14:55

      Hi Greg. I’m glad you find our results interesting. I’ve always enjoyed reading your papers and science fiction novels, and looks like we need more of your coherent synchrotron modeling to explain the large brightness temperature we found in BL Lac! JL

  • Jon Lomberg February 2, 2016, 14:42

    Hey, that’s my Oort Cloud artwork, RadioAstron. How about an image credit, or even asking permission first?

    • Paul Gilster February 2, 2016, 16:12

      No credit given to you on the NRAO site, from which this comes. But I’ve just added the credit to the image in the post above — thanks for letting me know, Jon!

    • Jim Strom February 2, 2016, 16:20

      And a nice image at that.

      • Jason The Great February 4, 2016, 1:37

        Great site Paul. Your attention to detail in the site alongside a focus of maintaining a positive philosophy to the rigors of long-term research / goals is inspiring and of good form. Keep up the great work

  • Ashley Baldwin February 2, 2016, 16:31

    Utterly wonderful . But O ! if only we could do this with optical wavelength telescopes. The CHARA optical array has just six 1m unit telescopes and has already imaged the surface of Betelgeuse and other Stars. It’s maximum baseline is just 330m . The number of baselines is important as they effectively “fill in” the gaps left over by the fact that despite having similar resolution the interferometer doesn’t collect lift across this “aperture ” . The more baselines , the more this gap is filled in to help create a “dilute aperture ” with detail determined by baseline number. So called “aperture synthesis ” . The equation for baseline calculation is
    Number of baseless = ((number of telescopes ) squared – number of telescopes )/2.
    So for CHARA this seems to be 15. By cleverly utilising the Earths rotation this allows the number to be increased significantly though. The resolution of a Interferometer is 1.02wavelength /aperture . So for nano meter optical light divided by an aperture of 330m this gives incredible precision.( bearing in mind that radio wavelengths are measured in mm , 1000000times bigger ) . The new Navy Interferometer will end up with 10 unit telescopes and a 600m plus aperture so will be even more potent . The ESO VLTI has 8 units up to 600m apart and is also very productive .The more baselines the better , so that “dilute aperture” ( as opposed to a filled in telescope or maximum baseline width) is filled in and the greater the detail of any image.
    The image itself is created by reversing one of the cleverest equations in astronomy, the Fourier Transform. This allows everything to be converted into an equation which can then be reversed by a computer to form an image and that’s exactly what interferometers do. Radio interferometry ( reaching it’s peak with the Square Kilometer array ) has a huge advantage of being able to convert its incoming light ( radio waves ) into energy fields which can then be saved and mixed later at leisure . Incoming light from each telescope of the interferometer must reach the “beam combiner” of the array in perfect time to constructively interfere ( maximum wavelengths in synchrony ) and add to each other in effect. Doing this by computer for radio interferometry is relatively easy. Optical wavelength light cannot be converted to an electrical field and thus has to be mixed ” real time” with a series of optical delays ensuring that light from each unit telescope arrives at or near exactly the correct phase to constructively interfere. This is hard , precision work requiring long optical trains of mirrors in vacuum filled tunnels as fibre optic cable isn’t yet good enough quality to perform the task . Each mirror, however reflective , will always loose some light with the result that optical interferomerters only really work on bright targets , like Betelegeuse. Worse still intermeometers don’t have adaptive optics to compensate for atmospheric distortion as each unit telescope would require them and they are expensive . Ironically the first interferometer in Cambridge,UK, was an optical version. If this problem could be overcome Labeyrie has shown that several hundred 3m space scopes with a maximum baseline of 150km could probably resolve a few pixels image of nearby exoplanets ! So that’s the potential of interferometry , the imaging technique of the (near?) future. ( though not able to convert the optical eaves to an equivalent energy field , it us now possible to create one as near as possible and allow for the difference thus allowing the radio syle approach to come nearer )
    Meantime, radio interferometry has already achieved get things through the VLBI, eMerlin and ALMA with more to come through the fabulous Sqaure kilometer array of next decade with thousands of unit radio telescope dishes in Australia and South Africa . Though even it needs a computer with processing power not yet invented so one can see what stands in the way of its optical equivalent . Come on Moores Law!

  • Ashley Baldwin February 2, 2016, 16:57

    I had heard about this a while back but great to see the reality of it. Now it’s the baseline that is the limit, not the Sky. Why stop here either ? Throw in some uncomplicated but tough deep space radio antennae and you can start having baselines that are a fraction of an AU. For imaging ( thanks to the wonderful mathematical “magic” that is the Fourier Transform) the less dilute the aperture ( this latter determined by the longest baseline) the better , so the more baselines the to fill it the better ( imagine a circle representing the aperture filled with evenly spread lines each capturing part of the final image ) . Shorter baselines allow “a zoom ” function. The ALMA array does this by moving its unit telescopes about with cranes on the high ATACAMA plain in the Andes, with shorter baselines allowing more detail. So ironically more smaller telescopes offer a greater return in interferometry than less but bigger telescopes . It’s baseline number that counts . A hundred unit telescopes gives – ((100) squared -100)/2 = 4950 baselines. Ideally all unequal or “non redundant “. That’s a lot of baselines but the more the merrier.

    • Mike Fidler February 3, 2016, 9:39

      I have been studying the Lockheed Martin SPIDER optical system and wondering about its capability to null out the exoplanets primary star, since it is an optical interferometer.
      The area between microwave and infrared is the transition between radio waves and optical, could that transition be used for both radio and optical and could more information be obtained? The red dwarfs radiate in the infrared to microwave range, could that be used to take advantage of exoplanet searches around them?
      I know I have been harping on this SPIDER system but it looks to have great potential, and DARPA work an optical phased arrays and laser communication would definitely help with data transmission rates from the distant realms of Planet 9. Beam focusing at great distances could even help with power and propulsion on such distant probes.

  • Marc Longoria February 2, 2016, 19:29

    So it wouldn’t make sense to point this at KIC 8462852, right?

  • Michael February 3, 2016, 2:44

    The use of the atomic clock to effectively time stamp the data is a novel trick and the same concept could be used with a set of telescopes on the way to the solar gravity focus point. It would allow us to start viewing a target sooner with great resolution all the way to the SGFP and as time goes on and we get closer to that focus point more and more light would be collected creating an astounding image.

    I suppose this timestamp idea could also be applied to set of telescopes in orbit around the Luna L2 point, orbits around the L2 point can be low energy path ways and much easier to achieve. Also the technique could use the sun as an additional focus along a narrow plane.


    • Mark Zambelli February 4, 2016, 12:21

      A huge baseline could be built using all of Earths Lagrange points. Add in L3 over on the opposite side of our orbit and we could probably get a 2 AU baseline.

      • Michael February 5, 2016, 8:01

        Although connecting all the Lagrange point would create a huge baseline the communication between them may be an issue. Maybe once an observation is made the information is packaged and sent back to Earth for combination via a mobile satellite (using low energy pathways) or at least it gets closer for better information transfer rates.

  • Ashley Baldwin February 3, 2016, 13:30

    Another option is a VLBI radios telescope network on the “dark side” of the moon that’s cuts out all the disruptive background interference from Earth.

    • Michael February 3, 2016, 15:14

      Yes a good idea, may be use both the Moons farside and the shadowed L2 point, that would give us a huge baseline to work with and as you wrote a very interference free region.

  • Alex Tolley February 3, 2016, 14:37

    Is the timestamp granular enough for optical interferometry?

    My general take, whether radio or optical, is that these techniques will mature relatively rapidly, given us data on exoplanets (and life) that we can acquire quickly, without physical probe travel times and c time lags for data return. They won’t give us the close ups and physical samplings that we will eventually want, but, like Kepler, we will amass a lot of data that may answer the question of whether life is ubiquitous or not.

    As always, any technology we develop, we have to assume ETIs can and have built too. Does being “radio silent” really keep us unobserved by predatory species?

  • Eniac February 4, 2016, 0:02

    This a very useful method. Alas, it works only for very intense objects, because of the mismatch between aperture and light gathering power. It is no accident that all the high resolution VLBA images are of black holes, and the optical interference images of nearby stars, i.e very intense light sources.

  • Patient Observer February 6, 2016, 0:56

    As always, a quality article with interesting/useful information. Only addition to the many fine comments is to commend the following Russian institutions involved in developing, launching and operating this inspiring piece of technology. Per


    “The RadioAstron is an international space VLBI project led by the Astro Space Center of Lebedev Physical Institute in Moscow, Russia.
    The payload – Space Radio Telescope, is based on spacecraft Spektr-R, that have been designed by the Lavochkin Association.”

    The success of this mission bodes well for future international collaboration as long as the politicians can stay out of the way of good science.