Why an array of telescopes is used?

[kelvin490]

To increase the resolution of an instrument, smaller wavelength and larger aperture is desirable. It is mentioned in some textbooks that the "effective" diameter of a telescope can be increased by using arrays of smaller telescopes. I just wonder why it is possible because every telescope is separated.

 

[DEvens]

https://en.wikipedia.org/wiki/Phased_array

 

[Andy Resnick]

There's two possible answers to this question:
1) the individual mirrors are all in close proximity and effectively form a single reflecting surface. This is done both for manufacturing reasons (it's not technologically possible to fabricate a 30-meter blank) and to enable optical correction (each of the smaller mirrors can be deformed to correct atmospheric turbulence- this is known as 'adaptive optics').
2) the individual mirrors are separated by an appreciable distance and can either form a sparse array or arms of an interferometric telescope.

The Keck interferometric telescope combines both:

https://en.wikipedia.org/wiki/W._M._Keck_Observatory

If you are thinking of a sparse array, as long as the incoming wavefront is sufficiently spatially coherent, each telescope samples the same wavefront and thus the 'apparent' aperture is much larger than a single telescope, providing the improved angular resolution. To be sure, the received intensity is not the same as a single multi-acre mirror...

This approach is sometimes called 'aperture synthesis'.

 

[Hesch]

every telescope is separated.

Let's call the thing coming out of a telescope a picture. Then it's right that a single picture from a single telescope is "separated" and is very blurred and dominated by noise. But all these pictures from say 30 telescopes within 3 hours are put into a huge computer, that compares these thousands of pictures and combine them into one picture, that is sharp and without noise. So as the pictures are gathered in the computer, they are not separated any longer.

The directions of the telescopes are controlled extremely accurate, so that all the telescopes are exact pointing at the same point.

I don't know about the computer calculations, but maybe a mix of Fouriertransforms, mean value, major votes, etc. are used to determine the color of a pixel.

 

[sophiecentaur]

This thread has included opinions about two distinct types of telescopes and the ideas seem to be a bit jumbled up.
The OP is, I think, about Radiotelescopes.
The telescopes can act together because the em from a distant source can be regarded as being coherent over the region of the array. RF signals from each of the sub-telescopes can be added together 'vectorially' because of the frequencies involved and the diffraction pattern of the whole array can be synthesised to get resolution (but not energy-gathering power) that you could get from an array covering the whole of the area (field) that the reflectors are situated in. This works because the signals can be taken along feed lines which preserve their phase relationship.

Optical wavelengths are so much shorter than RF and it is not possible (afaik) to combine signals in this way. Instead, in an optical telescope with many facets, each facet can be adjusted in position and angle to get the sharpest picture possible at a common focal point for all the facets and eliminate not only the effects of the telescope construction but the effect of the atmosphere on different parts of the overall reflector. A single large reflector could not be made to the required mechanical accuracy.
Edit: Something that I love about this system is that they optimise the positions of the sub reflectors using a nice, bright object which is easily visible and then they will get the best picture possible of a very low visibility object that sits right next to the visible one. That object would not be visible enough to lock onto on its own.

 

[kelvin490]

This thread has included opinions about two distinct types of telescopes and the ideas seem to be a bit jumbled up.
The OP is, I think, about Radiotelescopes.
...

Thanks all for answering. I can understand that every small telescope can be seen as a small tile of a big mirror/telescope. But I still wonder why this kind of arrangement can increase resolution. For a single aperture you need to have big diameter so that the diffraction effect is small (smaller angle for the first dark ring). For an array of telescope each telescope still has relatively small aperture and diffraction would still affect the image.

 

[Drakkith]

But I still wonder why this kind of arrangement can increase resolution. For a single aperture you need to have big diameter so that the diffraction effect is small (smaller angle for the first dark ring). For an array of telescope each telescope still has relatively small aperture and diffraction would still affect the image.

But when you combine those apertures you get the equivalent of a larger telescope that's had its aperture partially blocked off. It would be better to have one big telescope, yes, but this is the next best thing.

 

[kelvin490]

But when you combine those apertures you get the equivalent of a larger telescope that's had its aperture partially blocked off. It would be better to have one big telescope, yes, but this is the next best thing.

I think as they are using arrays of telescopes they certainly get better resolutions, I am just not quite sure about the effect of diffraction on smaller apertures.

If I look through a circular tube with diameter D which consists of many smaller tubes inside, like a bundle of drinking straws, and if the diameter of these drinking straws are small compared to the wavelength of visible light, can we still get the same resolution as looking through the big tube with diameter D but no drinking straws inside?

 

[Drakkith]

If I look through a circular tube with diameter D which consists of many smaller tubes inside, like a bundle of drinking straws, and if the diameter of these drinking straws are small compared to the wavelength of visible light, can we still get the same resolution as looking through the big tube with diameter D but no drinking straws inside?

No, but you can get a better resolution than any individual straw.

 

[sophiecentaur]

I think as they are using arrays of telescopes they certainly get better resolutions, I am just not quite sure about the effect of diffraction on smaller apertures.

If I look through a circular tube with diameter D which consists of many smaller tubes inside, like a bundle of drinking straws, and if the diameter of these drinking straws are small compared to the wavelength of visible light, can we still get the same resolution as looking through the big tube with diameter D but no drinking straws inside?

What goes on is similar to what happens with a directional array of dipoles or directive sub antennae (e.g. four yagi TV antennae arranged in parallel). I often find that the Radio equivalent model is easier to use to work out the optical problem; they are equivalent to each other, with only a change of size and wavelength. A search for "directional multi-element arrays" will produce interesting results for you. The overall pattern is a good approximation to the pattern of an indvidual element 'multiplied by' an 'array factor', which is the pattern of an array of point sources. The widely spaced point sources give you several 'sharp' peaks and nulls and the individual elements provide overall directivity so that only one of the peaks of the basic array is selected. This will produce an array with the directivity of a large aperture but is much cheaper to build. Naturally, the performance is not as good as from a massive antenna but it's good engineering value.
So your smaller aperture elements have diffraction patterns that will affect the overall pattern (possibly by letting in multiple lobes of the wide spaced array of sub reflectors). You can reduce this effect by having sub reflectors closer together (fuzzier beam, of course) or by processing the signals to identify the 'real' picture you are looking at. This is the same thing as happens with interferometry where you count nulls passing through as transmitter and receivers are moved around. It's very sensitive to changes but you may not know exactly where you are without more information.
In the case of your bunch of straws, each straw will have a wide / fuzzy diffraction pattern but the array factor will suppress this and produce a single sharp diffraction peak. In an optical array, the images from all the 'straws' have to arrive at the same point so that they can interfere with each other coherently. So there has to be a single focus somewhere - just as with a normal large spherical reflector. But the straws can be adjusted minutely to produce a better overall image than you would get with a single fixed reflector.

 

[Redbelly98]

The idea of using widely-separate telescopes is about 100 years old, and first employed by Michelson (and Francis G. Pease) to determine the angular size of Betelgeuse in 1920:

https://en.wikipedia.org/wiki/Michelson_stellar_interferometer

Oddly, Michelson had thought of the idea 30 years earlier. Not sure why it took so long to apply it.

 

[sophiecentaur]

The idea of using widely-separate telescopes is about 100 years old, and first employed by Michelson (and Francis G. Pease) to determine the angular size of Betelgeuse in 1920:

https://en.wikipedia.org/wiki/Michelson_stellar_interferometer

Oddly, Michelson had thought of the idea 30 years earlier. Not sure why it took so long to apply it.

Implementation problems, maybe? The modern system must require some seriously up to date control technology.
The RF equivalent has been used for many decades, I believe. Controllable RF phase shift networks are a piece of cake. (relatively, of course)

 

[Redbelly98]

Implementation problems, maybe? The modern system must require some seriously up to date control technology.

Must be something like that. It's easy to lose site of where technology was back then.

 

[Andy Resnick]

Must be something like that. It's easy to lose site of where technology was back then.

It totally was due to lack of available technology- the ability to control relative displacements of telescopes to a fraction of a wavelength means real-time controlling for thermal drift, seismic vibrations, etc. etc. to a factor of 1:10^-8 or so (1 micron over 10 meters). Fabricating the rails at the Palomar test bed required a microscope rig to ensure rail sag was less than a micron.

 

[sophiecentaur]

@Redbelly: Is that a Spotted Woodpecker or something more exotic? (Apologies for the deviation) My guy was at Bristol Zoo and he did not look too happy (he had obviously had an altercation with someone on PF!)

 

[Redbelly98]

It totally was due to lack of available technology- the ability to control relative displacements of telescopes to a fraction of a wavelength means real-time controlling for thermal drift, seismic vibrations, etc. etc. to a factor of 1:10^-8 or so (1 micron over 10 meters). Fabricating the rails at the Palomar test bed required a microscope rig to ensure rail sag was less than a micron.

Ah, thanks for chiming in. Suspected it might be either advances in mechanical tracking or perhaps availability of faster film allowing for shorter exposure time, but wasn't sure.

@Redbelly: Is that a Spotted Woodpecker or something more exotic? (Apologies for the deviation) My guy was at Bristol Zoo and he did not look too happy (he had obviously had an altercation with someone on PF!)

It's a red-bellied woodpecker! They're fairly common where I live (northeastern U.S.A.)

 

[sophiecentaur]

Ah, thanks for chiming in. Suspected it might be either advances in mechanical tracking or perhaps availability of faster film allowing for shorter exposure time, but wasn't sure.It's a red-bellied woodpecker! They're fairly common where I live (northeastern U.S.A.)

Exotic in terms of English wild birds! We have a regular visit by a spotted one to our garden and there are green woodpeckers about. Wild birds are a joy. Better than pets because you don't have to clean out the cage or the carpets.