On Fri, Feb 03, 2017 at 03:39:54PM -0800, C. Berry wrote:
> Ah, if we're talking about dark energy then it makes a lot more
> sense. But none of the writeups I've seen have ever said that.

Now that I've read the original paper on which the popular articles
were based, I can see that they aren't talking about the lambda
term/dark energy.  They are looking at variations in movements of
galaxies, compared with what one would have expected if the density
had been uniform throughout (the Lambda-CDM model).

In that type of comparison, regions of greater density behave like an
attractor and lower density regions behave like a repeller.  The
scales they're looking at are rather large(*), so it is very likely
that the gravitational effect of their "dipole repeller" region is
directly repulsive, but that's not relevant to the analysis they're
doing as they're only looking at differences not absolutes.

The conclusion of their paper is that attraction from the known
Shapley concentration of galaxy clusters is not sufficient to explain
the observed variations in velocities.  It accounts for about half of
the motion, with some unobserved factor (probably a region of lower
density) having a *relatively* repulsive effect in the other direction
accounting for another half.

(*) They use the unit "megaparsec" (a million parsecs) to denote some
of the tiniest distances associated with their analysis.  The region
they consider is hundreds of megaparsecs across, and they use
cosmological expansion velocities (up to about 10% of c) to denote
most distances.

- Tim