Many years ago, I read in one of the popular science magazines about using many small telescopes (say with a 12" diameter) placed in a ring, each half a mile from some central point. That would, IIRC, effectively give a single telescope with a 1 mile diameter.
I don’t remember where I read it (or even if I’m remembering it correctly), but is something like that used today?
I know of a number of radio telescope arrays, which seem to operate under the same general principle, but a little bit of googling fails to find anything of the like for optical telescopes. Doesn’t mean that they don’t exist – it just might be that my google-fu is insufficient. 
I don’t have an answer for you, but I remember reading something similar. Except, I remember them talking about launching many small telescopes into space and having the same effect that you describe.
See Optical interferometry. It’s been routinely done with radio waves for years, but it’s much more difficult to do it with optical instruments because light waves are so much smaller than radio waves.
This? Telescope Array Project - Wikipedia
Apart from making manufacture, transport and repair easier, would there be advantages to that setup? Could it do anything that 1 large telescope couldn’t do?
The advantage is it’s much easier to build two small telescopes and place them 100 feet apart than it is to build a 100 foot wide telescope. But in most respects one large telescope is better than an interferometry setup.
Sweet! Thanks. That is even cooler than what I remembered.
That’s always been what I pictured, too. A vast array of little telescopes orbiting a central point. Or as a structure on the dark side of some tidally locked moon in the outer planets.
There are basically two advantages to having a larger scope. The first is that, for any given aperture size and wavelength, there’s a fundamental limit to the resolution you can achieve, and so making your scope larger will give you better resolution. Interferometry helps with this: Small scopes placed (say) 100 m apart can (in principle) give you the same resolution as a big scope 100 m across. But in practice, for telescopes on the surface of the Earth, this usually doesn’t matter much, because a scope on the surface of the Earth will also have its resolution limited by the atmosphere, and that limit quickly becomes more relevant than the diffraction limit.
The second benefit of a big scope is much simpler: It’s a larger light bucket. The more total area of telescope you have, the more light you’ll collect. An interferometer will usually have much lower area than a single big telescope of the same size, because of all the area between the scopes that isn’t filled with mirror.
Which is why, if I’m not mistaken, it’s of much greater use for radio-wave astronomy, as those photons are not as affected by the atmosphere as visible photons are. So with radio waves, increasing your baseline to increase your resolution can be done to a much greater extent before running into the atmospheric resolution limit.
However, I seem to recall something about “active optics” as something that was coming into vogue at the time of my astronomy course in college (over 15 years ago) that was somehow supposed to figure out how the atmosphere was distorting things and correct for that somehow. It’s possible they’ve developed ways now to get greater resolution out of multiple visible-light collection devices in the same sort of way, but I never really understood how it was supposed to work anyway.
Almost right-- It’s called “adaptive optics”. All of the big-time research telescopes nowadays use it, and it certainly does help a lot, but it still has limits, and you’d much rather not have to deal with the atmosphere at all.
There is active optics and adaptive optics.
The ESO Very Large Telescope uses interferometry to combine light from a number of small telescopes. Very Large Telescope | ESO United Kingdom
Well, that, and the fact that radio waves are much longer than light waves, so radio telescopes of a given aperture have far poorer resolution than optical ones of the same aperture. A smallish radio telescope dish operating at a longish wavelength has awfully poor resolution.
Which might be OK. Amateur radio astronomers might observe Jupiter with a small yagi antenna, and might have a resolution of several degrees, but they can still hear the shhh-shhh-shhh sound of the planet, which is pretty cool for a backyard project.
And it’s a lot easier to do interferometry using the longer waves of radio. To do interferometry with multiple radio telescopes, you just need a good clock at each one, and FedEx the data around from each. You can get an effective aperture the size of the planet, or larger, this way. To do it with visible, though, you need direct light paths between the optics, before you get to the point of any of the detectors.
Indeed. As discussed in another thread, for optical telescopes you either have to have the configuration between the elements exact down to 1/4 of a wavelength of light, or else you would have to be able to compare the light received by each by something like holography using a reference laser.
Interesting article about how there might be a workaround for resolution limits, by invoking the Quantum.
As already described there are significant limits to optical interferometry, esp. for extended objects (galaxies and nebulae) vs point objects like stars.
There are similar limits for adaptive optics, although significant progress is being made. The VLT recently imaged Neptune with better resolution than the Hubble Space Telescope: https://www.eso.org/public/usa/news/eso1824/
Several much larger telescopes are now under construction: Extremely large telescope - Wikipedia
These large telescopes generally use segmented mirror designs but this is different from using several dispersed smaller telescopes. In simple terms they are more like a single monolithic mirror.
The fact these are under construction implies two things:
(1) There’s a general consensus adaptive optics will continue to advance, otherwise there would be less reason to build extremely large telescopes.
(2) Using several smaller telescopes by themselves or in an “outrigger” fashion with a larger telescope has significant limitations – maybe more than adaptive optics – else they would just use several smaller telescopes combined using optical interferometry instead of building huge ones.
Both adaptive optics and optical interferometry would be more achievable at longer wavelengths (such as infra red) than shorter, and they work better for small point objects than extended objects. The longer wavelength reduces the required ultra-precision for manipulating optical paths and delays. Unfortunately the atmosphere blocks most IR, so both techniques can’t use IR: http://coolcosmos.ipac.caltech.edu/cosmic_classroom/ir_tutorial/irwindows.html
I think the above Neptune image was a single VLT telescope using adaptive optics, not multiple telescopes using interferometry. The ESO VLT has the world’s most advanced optical interferometer. If this worked for planetary-size objects the Neptune image would have been even higher resolution.
The Neptune image is a groundbreaking accomplishment of adaptive optics. There is reasonable optimism that in the visible spectrum, new large ground-based telescopes can do much of the work originally envisioned for large space telescopes. Note the upcoming Webb space telescope works in IR from 0.6 to 28 microns so that is needed.
But the fact we’ve never seen a non-point object at even higher resolution implies that optical interferometry is extremely difficult and not yet capable (or even projected) to replace large terrestrial telescopes.
One other limitation to interferometry is that it only gives you high resolution in the directions that you have telescopes spread out in. So if you had only two telescopes (or more of them, but on a straight line), then you would get great resolution in that direction, but only ordinary resolution perpendicular to that line. Of course, for actual scientific purposes (as opposed to just making pretty pictures), that’s often good enough.
Thanks for the detailed explanation. Very helpful.
That’s interesting. I was wondering what a linear configuration would produce.
Will we ever reach a point where we can use ultra-high resolution digital images to replace the precise light paths? Like, if we had an 8K (or hell, even a 16K) image with a big number-bit color-depth, could we use those for the interferometry? Or is that sort of addition and subtraction a different animal altogether?
Well, you can do a sort of interferometry using instrument output (in which case the bit depth of each pixel is much more important than the number of pixels), but it only gives very vague results, not nearly as good as actual interferometry (or an actual big optic). You need the phase information, and visible-light instruments simply don’t record that at all, no matter how many megapixels you have. In principle, you can beat the incoming signal against a stable reference beam before it hits the detector, but that requires that the reference beams at all of your telescopes are extremely well-synchronized. For radio, a high-end but still off-the-shelf clock is good enough for the synchronization, but a clock good enough for visible light is so far beyond the technology horizon that we can’t even envision how it would work.