The National Radio Astronomy Observatory's Q&A "How To Make Images With a Radio Telescope" says

The most straightforward way to make a radio image with your satellite antenna system, then, is to use an array detector, which I believe would be very difficult to find. Alternatively, one can make an image by pointing an antenna to an array of nearby positions on the sky then creating a contour map of the measured signal strength. This is a rather slow process, though, that only produces an image much later, after you have gathered all of your individual pointing measurements.

Assume now that a lot of people all over a continent (e.g. Europe) would have (fixed, non-moveable) radio dishes pointing somewhere into the sky, with their field of view somewhat overlapping. They have different dishes and different receivers. However, given their exact location on the globe, knowing their exact time-stamp of observation, at least theoretically, I should be able to combine the measured time-series of radio-signals to a radio-image of the sky, correct?

The idea already bothers me a while and I am wondering whether there is some open source software I should be aware of to tackle the problem of caculating such an image, ideally in python (but possibly in any other programming language).

Any input is appreciated!



1 Answer 1


This is an interesting question!

I'll post a partial, preliminary answer and let's see where this takes us.

The NRAO block quote mentions two methods, interferometry and intensity mapping (like how a camera works). There may be hybrid options and other esoteric options, but lets look at these first.

See also this answer to Which techniques are used to convert radio signals received by antenna to images? and since there are so many good answers to Radio telescope targeting those might be worth a good read as well.


This is where all dishes point in the same general direction and you have:

...the exact location(s) on the globe, knowing their exact time-stamp(s) of observation...

and those time stamps must be made to nanosecond relative accuracy; you use the phase information of the radio waves. You either hook them up with coax or fiber optic cable in real time, or write at up to gigabit/second speeds to hard drives and then cart them all to one location and run a very very big and special computer called a correlator to analyze the data.

Intensity mapping

This is just making an intensity map by sampling many spots in the sky and building up a histogram; basically exactly like how a camera works but kilopixels rather than megapixels. Each telescope is an independent sensor, and it contributes to the the big map. You might start with a grid of pixels and as the antennas drift into each box you measure the intensity of radio power, and write that number down in that box. Earth's rotation scans in RA already. so no matter where they are, the dishes should be evenly spaced in declination.

See answers to How did single dish (or single receiver) radio telescopes originally generate images?

Hybrid and other esoteric options

There is intensity interferometry but I don't understand it well enough to write authoritatively on it. It looks at fluctuations in statistical noise, and so you can probably get away with less critical timing; each dish would record the instantaneous power level at perhaps microseconds or even millisecond intervals. I haven't a clue how this works with arrays; I've only heard it used for two receivers, but you could certainly analyze it pair-wise. (see also Narrabri Stellar Intensity Interferometer)

You could also do something fun/interesting with pulsars, but I don't know what exactly.

Intensity plus Doppler frequency can sometimes give you distance if you have a model; this is an early radio map of neutral hydrogen in our galaxy from Why the blank wedges in this very early 21 cm map of the Milky Way? (Oort et al. 1958)

enter image description here click for larger

  • 1
    $\begingroup$ Thanks for your thoughts. I guess in terms of time stamp accuracy I would not be able to go down to nanoseconds, the dished of all the participants should be connected to their standard PCs and data transfer is happening via TCP/IP. Interferometry would be cool, but it would be rather kiloPixel photography. $\endgroup$
    – B--rian
    Commented Mar 26, 2021 at 15:31
  • $\begingroup$ A major challenge I see is how to figure out the individual orientations of the dishes, e.g. each dish's view field. I cannot even assume that the dishes are not vibrating in the wind, I guess. The dishes should be extremely low budget. And yes, data processing will happen central and not in real time. $\endgroup$
    – B--rian
    Commented Mar 26, 2021 at 15:35
  • 1
    $\begingroup$ @B--rian see my comments under Can one build a distributed radio telescope? GPS receivers provide a 1 pulse-per-second timing signal that you can use to correct a local oscillator, but it won't get you to interferometry unless you are working a low frequencies, at which point you don't need a dish you need a lot of wires, and then you have the cut-off of the ionosphere. I just realized that you could consider looking for some kind of transient events, like nearby FRB's or something else, but that's not easy. I'll think about it. $\endgroup$
    – uhoh
    Commented Mar 26, 2021 at 15:40
  • 1
    $\begingroup$ @B--rian I forgot to say it in my post, but any software one might use depends on which observation mode you select. Interferometry is numerically intensive and requires a much more of attention to detail and calibration than intensity mapping, so before someone can recommend software for you, it's necessary to decide in what mode the proposed array will operate and what things it will look at and what level of sensitivity and calibration are necessary. $\endgroup$
    – uhoh
    Commented Mar 26, 2021 at 22:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .