I overheard a discussion between friends:

If your spaceship was transported to a random location in our galaxy, you could identify nearby stars by their signature, and then triangulate your position.

Another friend replied.

The signature of a star isn't sufficiently unique. You can't identify a star by spectroscopy.

A third friend chimed in

There would be enough unique pulsars to work out your position from.

I came away not convinced on the stellar spectroscopy question.

My question is: Can you identify a star's 'signature' from its spectroscopy?


Realistically, no - there are too many stars with the same spectrum.

The pulsar idea is a good one, though. There is a reason our location relative to some bright pulsars was sent out on the Voyager plaques: we are confident that a civilisation finding one could then find our sun, and us, as the pulsar rates are unique enough.

  • 1
    $\begingroup$ Given that the lifetime of a typical pulsar (as a pulsar) is a million years and that their pulses and periods change on timescales much shorter than that, I think it is not such a good idea at all. $\endgroup$ – ProfRob Dec 5 '17 at 7:03
  • $\begingroup$ Rob - the lifespan of the pulsar isn't the issue. What is the lifespan of the human race...? $\endgroup$ – Rory Alsop Dec 5 '17 at 7:07
  • $\begingroup$ That isn't the issue, how long will it take Voyager to reach just the nearest stars? $\endgroup$ – ProfRob Dec 5 '17 at 7:19
  • 1
    $\begingroup$ A quick suggests >40000 years. The pulsar positions, periods and pulses will have changed by then. Publicity stunt. $\endgroup$ – ProfRob Dec 5 '17 at 7:23
  • 2
    $\begingroup$ Actually the pulsar idea doesn't work at all because most pulsars we see from the Sun would not be visible from a random position in the galaxy because of their tight beaming angles. The pulse shape is also determine by the viewing angle. $\endgroup$ – ProfRob Dec 5 '17 at 7:29

The answer is no, but not only for the reason you suppose.

Each star probably does have a different chemical composition and hence a different spectrum. However, many stars have a very similar chemical composition, and those stars born in a cluster an almost identical composition.

Therefore stellar spectra whilst having the potential to be a unique identifier, are not currently measured with sufficient precision to make this possible.

However, another reason why what you suggest won't work is that by and large, we can only obtain spectra for stars inside a modest radius from the Sun, especially if the stars are near the Galactic plane. One could work in the infrared, but the spectra of bright stars are generally less distinctive in the infrared.

You could go with a database of positions, absolute magnitudes and spectral types and probably work out where you were if it were reasonably close to the Sun. Note this only works if you are transported quickly (in terms of stellar evolutionary lifetimes).

A better bet is to look at the positions of local group galaxies (e.g. the magellanic clouds, Andromeda), which will be in noticeably different positions. Also, globular clusters can play a role, because many of these do have distinctive chemical abundances and multiple populations and are distributed around the Galaxy and out of the Galactic plane.

The pulsars idea doesn't work at all because: (a) The radiation from pulsars is beamed into a small angle, so they would not be visible in the same way from another Galactic location. (b) By the time you had got to another Galactic location (unless faster than light travel is postulated) some pulsars will have turned off, all will have significantly changed their pulses and periods, since the total lifetime of the pulsar phenomenon is of order a million years.

  • $\begingroup$ The clusters piece is a very good point. Hadn't thought of that. $\endgroup$ – Rory Alsop Dec 5 '17 at 7:31

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.