When we go through multiple astronomy catalogs, we often do Cross-matching. We compare the position parameters $(\alpha, \delta)$ of different sources (galaxies, quasars, etc.) in the equatorial coordinate, where $\alpha$ is the right ascension and $\delta$ is the declination.

However, it is possible that two completely different sources are on almost the same position. We might make make mistakes if: $$ \alpha_1 \approx \alpha_2,\ \delta_1 \approx \delta_2 $$ The question here How can we be sure that we have identified very distant stars correctly? pointed out the redshift features could be very important in this problem.

My question is:

  1. Does the phenomenon often occur when we are doing cross matching?
  2. In specific, how can we use the redshift features in cross matching programs?
  3. Do we have any other way (some special statistics techniques, for example) to solve the problem?


  • 1
    $\begingroup$ I think you should narrow your question down. Of course if you cross-match using only coordinates then there can be spurious correlations, since all positions have uncertainties or correspond to sources where the flux has been gathered from a finite positional area. How you resolve (if you can) this problem just depends on exactly what science you are trying to do. Q2 isn't going to be relevant in many (most) cases unless you have a specific application you want to ask about? $\endgroup$
    – ProfRob
    Dec 21, 2022 at 9:30

1 Answer 1


It's possible that two completely different sources might be located almost on top of each other, which could lead to mistakes in our analysis. One way to help avoid this is by using the redshift of the sources, which is a measure of how far away they are from us. By only considering sources that have similar redshifts, we can be more confident that we're not getting false matches.

Additionally, there are other methods like using physical parameters or statistical tests that can also help us identify and eliminate false coincidences.

  • $\begingroup$ "statistical tests" is rather vague -- do you have a specific, relevant test in mind. (And in fact part 3 of the question is specifically asking this.) $\endgroup$ Dec 22, 2022 at 10:29

You must log in to answer this question.

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