I have read about ground-based Very Long Baseline Interferometry telescope arrays able to achieve huge resolution at IR/visible wavelengths. There are also space-ground VLBI configurations in operation today that work at radio wavelengths.

  1. Is it feasible to employ many space-based optical/IR telescopes to achieve a high angular resolution visible light image of something the size of Jupiter? (Maybe during a transit or occultation, since it would be faint on its own.)
  2. What would the separation need to be for a space telescope array of that size?
  3. Would it be overly difficult to get precise enough timing, or to synchronize the relative motions of the orbiters or to deconvolve that motion from the sources?
  4. Could it be made of many considerably smaller widely separated mirrors, or would it be better to use a fewer number of large mirrors?

2 Answers 2


Interferometry is dependent upon preserving the phase data from signals from widely separated locations. This can be done physically by relaying the actual EM radiation along waveguides or optical paths, or it can be done synthetically by precisely recording enough phase data from the signals to simulate the same process in a computer. This requires measuring the signal at a very high resolution and precisely measuring the exact 3 dimensional position of the telescope at the scale of the relevant wavelength. This is possible with radio waves but is not technologically possible with optical or IR light.

Consider VLBI using a 1 micron wavelength signal (infrared light). You would need to precisely locate each observatory to sub-micron precision, which is far beyond the capacity of any GPS based system. You would need to record the phase data for the signal at a rate higher than 300 terahertz. We don't have sensors or computers that are capable of doing that. And that doesn't even get into the requirement of storing petabytes of data per second. Perhaps in the future someone will find a more clever way of doing it, but using the same method as radio VLBI is not even remotely feasible for optical wavelengths.

  • $\begingroup$ Not my specialty, but what about LISA? Gravity waves require fantastically precise positioning. $\endgroup$
    – Andrew
    Commented Jul 20, 2011 at 12:53
  • $\begingroup$ @Andrew LISA uses physical interferometry, positioning is determined by the interferometer itself. Potentially you could use a similar system to keep track of the positions of the telescopes but then you run into the same problem again of having to record both positioning information and the signals themselves at 300+ terahertz. But if you're going to maintain an interferometric link you might as well just use the primary signal anyway. $\endgroup$
    – Wedge
    Commented Jul 21, 2011 at 0:10
  • $\begingroup$ Yeah, when I reread your answer, I realized the speed of phase information was really a killer. Gravity waves are only in the kilohertz range. $\endgroup$
    – Andrew
    Commented Jul 21, 2011 at 10:55
  • $\begingroup$ I'm not sure how high they go, but gravity waves go much lower than kilohertz. The waves from the spindown of pulsars are 2x the rotation rate which can get into the 10s of hertz range. Mergers between larger black holes can drop down into the single digits range. Waves dating back to the pre-inflationary era could have been stretched until they're much slower than 1hz. $\endgroup$ Commented Mar 6, 2012 at 19:23

There are several mission proposals for a space based infra red interferometer, Darwin from ESA and TPF from NASA. However both were considered too risky for development and so are unlikely to see the light of day. This does not mean the field is dead and there are a few active research groups.

Separations would need to be of the order of 10's to 100's to 1000's of metres to be of any improvement over ground based instruments.

One current form of Interferometry that is largely insensitive to mechanical defects is Intensity Interferometry, this is an old technique but is also an area of active research.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.