In my question Why are quasars so far away that they couldn't be optically resolved in the 1950's? I included the following short paragraph, but then added strikethrough to the second sentence and added a parenthetical based on @DavidTonhofer's comment.

So today instead of "quasi-stellar" we might simply say "unresolved". The light is star-like because the light from galaxies has a strong stellar component. (this second sentence is argued against in comments)

I then read further and found this short but informative historical account in University of Manchester's Jodrell Bank Centre for Astrophysics page The MKI and the discovery of Quasars:

Initially the radio sources were observed with the telescopes relativly close together. In this case the fringe pattern is very broad and greater than the typical angular size of radio sources so virtually all are observed to give fringes. However as the telescopes are moved apart, the fringe spaceing becomes finer and the fringes from many of the radio sources first reduced in amplitude and then disappeared. The separation of the telescopes when the fringes from each source reduced in amplitude (and the wavelength of the radio waves observed) allowed the angular size of the radio sources to be found.

To the surprise of the astronomers quite a number were found to have exceedingly small angular sizes - so small in fact that their images on a photographic plate would look like the images of stars. Much work then went into finding the precise location of these objects and finally it was possible to obtain their spectra. They were nothing like the spectra of stars or galaxies and remained a puzzle for some time. Finally it was realised that the reason that the spectra were so different was that the objects were so far away that their light was greatly red shifted. They were at very great distances away from us and so it was not surprising that their images were so small!

Because their images looked liked those of stars they were called Quasi-Stellar-Objects or QUASARS for short.

Question: In the late 1950's how were astrometric positions of these radio sources made so precisely that they could be assigned unambiguously to dim, star-like spots on photographic emulsions?

The 25ft Transportable Radio Telescope used with Jodrell Bank The 25ft Transportable Radio Telescope


In many cases in astronomy, there’s a difference in finding the position of something precisely vs. seeing detailed structure in that object. In the former case, what you need to do is to find the centroid of the image. So even with relatively coarse spatial resolution, you can often get very precise positions because you can centroid the images very well. (EDIT: However, this may not apply to early radio observations because single-dish radio telescopes don’t form images.)

Specifically for 3C273, the first quasar identified, Wikipedia says that its position was refined with lunar occultation. This technique involves observing a source as the Moon approaches it and determining the precise moment when it disappears behind the edge of the Moon. Here you need to measure time precisely and to know the Moon's path very well.

Also see this answer to Has radio astronomy ever been done on objects that appear very close to the Moon? Is this avoided?

The first radio source to have a star-like optical counterpart associated with it was 3C84, and that was done by measuring the radio position precisely using interferometry, i.e. combining signals from multiple telescopes.

So I think that this was not in fact possible with single-dish radio telescopes in the 1950s, and required more advanced techniques that came a bit later.

  • $\begingroup$ interferometry provides resolution but to get astrometric positions don't you need some significant amount of calibration or referencing to other sources? $\endgroup$ – uhoh Jun 27 '20 at 14:06
  • $\begingroup$ Why wouldn't you know where you point your dish to? You could always Co - align a small optical telescope for calibration $\endgroup$ – planetmaker Jun 27 '20 at 17:26
  • 1
    $\begingroup$ @uhoh Correct - this paper discusses the calibration of the Owens Valley radio interferometer for measuring precise positions in that era. It's pretty technical, but the bottom line is that they used radio sources with known positions, either those with obvious optical counterparts (like radio galaxies) or objects whose radio positions had already been measured precisely, for example with lunar occultation or other interferometers. $\endgroup$ – ELNJ Jun 28 '20 at 18:03
  • $\begingroup$ @planetmaker You can't co-align an optical telescope with a 90-foot radio dish with anywhere near the needed level of precision; with a big dish there isn't any obvious large mechanical part pointing in precisely the center of the dish to align with. Even piggybacking two optical telescopes together requires significant tweaking and calibration to get them co-aligned at the arcsecond level. If you look at the picture above, you can see that it would be hard to attach a separate telescope and ensure that it is pointing in precisely the same direction. $\endgroup$ – ELNJ Jun 28 '20 at 18:08
  • $\begingroup$ Thanks! I love reading old scientific paper. Authors really took the time and space to explain carefully and in plain language. In the paper in your comment using baselines of 60, 120 and 480 meters on a north-south line at 960 MHz they improved declination determinations of catalogued radio sources using "lobe rotation" meaning rather than just time the flat interference fringes (on the scale of minutes) they slowly ramped the relative phase between the two signals to add an offset to the beat frequency and used the "rotation" period to compare to the recorded zero-crossings. $\endgroup$ – uhoh Jun 28 '20 at 23:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.