How common are the Milky Way and Andromeda?

A recent paper by Licquia, Newmann, and Brinchmann (2015) reported that the Milky Way, if viewed externally, would have an absolute magnitude of Mr5logh=21.00 and a rest frame color of (g-r) = 0.682.    This would place it very near the red sequence or in the green valley, where galaxies are thought to be transitioning to being quiescent systems.  Red spirals generally only make up about 6% of the spiral galaxy population as found by Masters et al. (2010), and following their definition, the Milky Way would qualify as a red spiral. 

Now, the Milky Way makes up part of the local group, whose other major member is the Andromeda Galaxy.  Andromeda is also a very red spiral galaxy at a distance  of 0.78 Mpc away from the Milky Way and approaching the Milky Way at ~300 km/s.   Galaxies like the Milky Way and Andromeda are likely to be incredible rare.  Out of a sample of 130000 galaxies, Mutch, Croton, and Poole (2011) only found 997 (0.77%) that were similar to the Milky Way or Andromeda.

Of those galaxies, how many are close together to each other?  This should be easily answerable using the SDSS CasJobs query.   Nonetheless, I tried to repeat the query from Mutch, Croton, and Poole (2011) of galaxies with spiral structure, stellar masses similar to the Milky Way and Andromeda, and exponential profiles along with adding in a color selection, requiring the galaxies to have a redshift less than 0.1,  and leaving out the face-on requirement. In the end, I found 5375 galaxies fitting the requirement out of an initial sample of 160000 galaxies with z < 0.1.   Compared to the early work, I used a slightly higher redshift cutoff and dropped the face on requirement, which explains the greater number of sources.  Of those 5375, 36 pairs exist where the two galaxies are within 1.5 Mpc of each other and have  a total line of sight velocity difference less than 150 km/s.

So while the local group is unique, it would appear roughly about 1 out of every 2200 giant galaxies would be in a local group analog (give or take a factor of 2).    I have to admit my initial feeling would be that they would be more rare than that.  Then again, this was a quick calculation and their might easily be several factors that I am missing.

Of course, it might be interesting to follow-up a few of these systems in detail to see what might be in store for the local group.    Mutch, Croton, and Poole actually suggest that the Milky Way-Andromeda merger will more likely be a dry merger when it happens and maybe some of these systems can give us some ideas about what is in store for the Milky Way.

For the record, here is the query I used.  If anyone has better ideas, please feel free to suggest it  and I’ll be happy to post an update.

-- Find red spiral galaxies similar to the Milky Way and Andromeda

SELECT count(g.objID) -- g.objID, g.ra, g.dec, t04_spiral_a08_spiral_debiased, sp.z, sm.mstellar_median, fiberMag_g - fiberMag_r, log10(1/expAB_r)
FROM Galaxy as g
join dr10.zoo2MainSpecz as z on g.specobjid=z.specobjid
join SpecObj as sp on g.specobjid=sp.specobjid
join stellarMassPCAWiscBC03 as sm on g.specobjid=sm.specobjid
--join Photoz as pz on g.objID = pz.objID
r < 24 -- r IS NOT deredenned
and sp.z < 0.1
and t04_spiral_a08_spiral_debiased > 0.8 --spiral structure is seen
and sm.mstellar_median > 10.66 and sm.mstellar_median < 11.2 -- only galaxies with stellar masses similar to MW/M31
and fiberMag_g - fiberMag_r > 0.63 -- find all red objects (this should be restframe color, but k-corrections should be small)
and fracDeV_r < 0.5 -- added to remove bulge dominated galaxies

And here’s an image of one of the systems that are very close together, however the on-sky separation of some of these systems can be visually quite large (0.78 Mpc at z=0.05 corresponds to roughly 0.2 degrees).


Credit: SDSS

Some Caveats:  The Mutch, Croton, and Poole estimate included some assumptions about viewing angle, so there probably are more systems where one object is being viewed edge on. Nonetheless, it goes into much more thorough detail then I do here and also has some very interesting thoughts about the long term evolution of the systems.  In addition, I’ve done nothing to control for spectroscopic incompleteness.   Also my SDSS-fu is very rusty so I could have screwed up the above query and I couldn’t figure out/remember an easy way to get to rest frame properties from the SDSS database.   Also a more thorough job could have been done to decide bound/un-bound pairs.

References and Acknowledge:  References are many linked from above, but I also made use of NED, astropy, but primarily the data from SDSS.   H/t to @jbprime for the original tweet.