How do radio telescopes work?

Question

If I search online for how radio telescopes work, the found articles talk about how RF is on the spectrum, etc, how the parabolic collector is the aperture which contributes to the sensitivity and reflects the signals into the detector at the focal point. Then they skip to the signal processing circuitry the leads to a rendered image. They also mention that the furthest nodes in a telescope array determine the virtual aperture of the array.

There are a few parts I don't understand: How does the detector work? From my reading, it is merely an RF antenna. I can't think how you'd get but one pixel per "moment" of detection. That the telescope is positioned and begins listening, and as the earth turns one scan-line of image is captured. For high resolution 2D image this would take thousands of passes. This seems impractical as the earth is orbiting and time marches on, otherwise.

Perhaps multiple telescopes each take a scan line and interlace the results ... but especially at such high resolutions, I can't see how you'd engineer the huge machines with such precision. Please explain this to me.

Do I have the 1D scan-line-antenna idea wrong? Is it more like a 2D CCD sensor in a digital camera with the antenna just behind the focus of the dish? If so, I assume the sensor would have to be a 2D array of nanometer scale antennas?

This may be clear in the explanation, but in layman's terns, how are do telescope arrays work together to create the virtual aperture? What is lost, if any between the virtual and a theoretical "real" aperture of equal size?

(I consider this a physics question because understanding the answer to this real-world application will help correct whatever theoretical misconceptions I have about RF, antennas, and radio telescopes)

Thanks!

Answer 1

In the microwave band here are multi-element detectors, but at longer wavelengths the telescope is a single pixel.

Yes it does take a while to build up an image, but radio pictures aren't usually very large - not the millions of pixels of an optical/IR image.

One big advantage of radio telescopes is that you can combine telescopes 1000s of km apart to create an image with the resolution of a single dish that large ( you can just about do this in the optical with 100m separation now)

You know about the interference pattern you get with two slit? If you picture the two telescopes as the two slits and you interfere (electrically) the signals to form the fringe pattern. You can then calculate the shape of the light source - a single point will produce the classic fringe pattern, 2 close points will produce a slightly different pattern etc.

edit: Building an optical 'radio' telescope - the original English solution and the rather more impressive Teutonic result (if you have a lot more $$$)

Answer 2

It depends on what kind of radio telescope you're talking about.

If you're talking about a single dish, with say, a horn feed at the focus, it behaves like a regular 'dish antenna' by focussing radio waves to the focus where it is amplified, processed in some manner (such as downconversion), digitised, and so on. A single dish can also have a feed array at the focus to act analogous to a camera sensor.

If you're talking about a radio telescope array, it can operate in one of two ways.

In the first method, the antennas in the array are pointed at the source, and receive signals that are at different phases due to the path difference associated with the relative locations of the antennas and the source. These phase differences are corrected for, the signals are combined, and then processed as before. This is usually employed when observing a point source (such as a pulsar) with a telescope array.

The second way is as an interferometer. This method is used to create radio images of (usually) extended sources such as radio galaxies. In this method, the phase differences are not removed, but utilised along with the rotation of the Earth to create a Fourier-synthesised image. The measured signal from each antenna is correlated with those from all other antennas to get what is called the 'visibility'. Taking the Fourier Transform of the measured visibilities gives you the brightness distribution in the sky (a.k.a. the image). Following the behaviour of the Fourier Transform, long baselines (antennas that are farther apart) measure small spatial variations (high spatial frequency) in the brightness distribution, while short baselines measure large spatial variations (low spatial frequency). In other words, it neither behaves like a camera with a CCD sensor, nor does it do raster-scanning.

The 'Antennas & Radiometers' chapter of this course should tell you more in detail: http://www.cv.nrao.edu/course/astr534/ERA.shtml

These lectures from the Sixth NRAO/NMIMT Synthesis Imaging Summer School might also be helpful: http://adsabs.harvard.edu/cgi-bin/nph-abs_connect?version=1&warnings=YES&partial_bibcd=YES&sort=BIBCODE&db_key=ALL&bibstem=Synthesis+Imaging+in+Radio+Astronomy+II&year=&volume=&page=&nr_to_return=100&start_nr=1

Answer 3

A single dish aperture can operate a single pixel detector but this pixel projected onto the sky can be quite large compared to optical and infrared telescopes.

The resolution is proportional the wavelength and therefore this can be 1000's of times the optical resolution. Therefore a radio telescope can take its image in a reasonable time simply because they aim at larger targets and take 'worse' images than you might imagine.