In article , Martin Brown
writes:

On 03/02/2017 23:06, Steve Willner wrote:
In article ,
"Phillip Helbig (undress to reply)" =
writes:
The real question is whether there is any real tension between Planck
and "local" measurements.

I agree that the most likely reason is some small systematic error.

Indeed. The speed of light in vacuum determination as a function of time=

with error bars is a salutary lesson in this regard. I remember the
graph from an introductory relativity textbook in our library.

The gist of the problem was that a very famous experimentalist made an
error in the sign of the correction for imperfect vacuum and everyone
afterwards made exactly the same mistake. It was only when a new method
with even greater precision and radically different systematic errors
came onstream that the problem was uncovered.

OK. I only parsed it correctly on second reading. This is an example
where similar "tension" was resolved when a systematic error was
corrected for. (You don't mean that a variable speed of light in vacuum
could be the cause of the current tension with regard to the Hubble
constant.)

it probably makes sense to wait and see if the tension is real
before trying to come up with something which would make the Hubble
constant depend SLIGHTLY on the redshift range in which it is
measured.

While I also agree that coming up with possible explanations isn't
worth a major effort, I think a bit of speculation might not be a bad
idea. It might suggest independent avenues of research to see
whether there might be evidence for some hypothesis or other.

I'm not enough of an expert to have any real idea what might cause
discrepancies between the local and CMB values of the Hubble
constant. I am _guessing_ a "hiccup" in the expansion history might
do it, but I have no idea what might cause such a hiccup nor how one
might test that. Any other ideas?

If the universe is accelerating with time due to dark energy then the
Hubble constant might be expected to vary slightly with increasing Z.
But I suspect by amounts much smaller than the experimental error bars.

This is a red herring. In general, the Hubble constant is not constant
in time. (It is called the Hubble constant because it is a constant
like "a" in the equation y = ax +b, not because it is constant in time.)
It varies quite dramatically. At the big bang, it is infinite. It
decreases at first, then can increase again, in some universes (such as
hours) asymptotically approaching a constant value. This is well known
and is always taken into account. The tension in measurements we are
talking about is tension in measurements of H_0, which is the current
value. Even if the objects involved are at high redshift, it is still
the current value which is measured. (Even at low redshift, the Hubble
constant is not "directly" measured.)