[astro-ph/0603647] The Hubble Constant: A Summary of the HST Program for the Luminosity Calibration of Type Ia Supernovae by Means of Cepheids

Authors:  A. Sandage (1), G.A. Tammann (2), A. Saha (3), B. Reindl (2), F.D. Macchetto (4), N. Panagia (4) ((1) Obs. Carnegie Inst. Washington, (2) Astr. Inst. Univ. Basel, (3) NOAO, (4) STScI)
Abstract:  This is the fifth and final summary paper of our 15 year program using the Hubble Space Telescope (HST) to determine the Hubble constant using Type Ia supernovae, calibrated with Cepheid variables in nearby galaxies that hosted them. Several developments not contemplated at the start of the program in 1990 have made it necessary to put the summary on H_0 on a broader basis than originally thought, making four preparatory papers necessary. The new Cepheid distances of the subset of 10 galaxies, which were hosts of normal SNeIa, give weighted mean luminosities in B, V, and I at maximum light of -19.49, -19.46, and -19.22, respectively. These calibrate the adopted SNeIa Hubble diagram from Paper III to give a global value of H_0 = 62.3 +/- 1.3 (random) +/- 5.0 (systematic). Local values of H_0 between 4.4 and 30 Mpc from Cepheids, SNeIa, 21cm-line widths, and the tip of the red-giant branch (TRGB) all agree within 5% of our global value. This agreement of H_0 on all scales from 4 - 200 Mpc finds its most obvious explanation in the smoothing effect of vacuum energy on the otherwise lumpy gravitational field due to the non-uniform distribution of the local galaxies. The physical methods of time delay of gravitational lenses and the Sunyaev-Zeldovich effect are consistent (but with large errors) with our global value. The present result is also not in contradiction with existing analyses of CMB data, because they either lead to wide error margins of H_0 or depend on the choice of unwarrented priors that couple the value of H_0 with a number of otherwise free parameters in the CMB acoustic waves. Our value of H_0 is 14% smaller than the value of H_0 found by Freedman et al. (2001) because our independent Cepheid distances to the six SNeIa-calibrating galaxies used in that analysis average 0.35mag larger than those used earlier.
[PDF]  [PS]  [BibTex]  [Bookmark]

Discussion related to specific recent arXiv papers
Post Reply
Syksy Rasanen
Posts: 119
Joined: March 02 2005
Affiliation: University of Helsinki

[astro-ph/0603647] The Hubble Constant: A Summary of the HST

Post by Syksy Rasanen » February 27 2007

This HST Hubble parameter summary paper gives a lower value than the widely cited Freedman et al paper from 2001, astro-ph/0012376, which found [tex]h=0.72\pm0.08[/tex], consistent with the CMB parameter fits of the [tex]\Lambda[/tex]CDM model.

These authors find [tex]h=0.62\pm0.05[/tex]. They claim that the difference with Freedman et al is due to the "untenable" treatment of Cepheid distances by the latter. (The authors note that with the improved treatment of Cepheids, Freedman et al would get [tex]h=0.60[/tex].)

Sandage and Tammann also have a more recent paper supporting the low value of [tex]h[/tex], astro-ph/0608677.

It seems that people have not paid much attention to these results. (For example, the Freedman et al value is still commonly applied as an external prior in CMB analysis.) is there some physics reason for this? The paper appears authoritative and careful. (This of course does not guarantee correctness - and I know nothing about Cepheids.) Is there something wrong with the analysis?

Garth Antony Barber
Posts: 59
Joined: July 19 2005
Affiliation: Published independent

[astro-ph/0603647] The Hubble Constant: A Summary of the HST

Post by Garth Antony Barber » February 28 2007

h = 0.72 ?
h = 0.62 ?

Why don't we just split the difference and call h = 0.67, which is within the error bars of both values.

Wait a minute, haven't we seen that value before??

Garth

Simon DeDeo
Posts: 44
Joined: October 26 2004
Affiliation: Santa Fe Institute
Contact:

[astro-ph/0603647] The Hubble Constant: A Summary of the HST

Post by Simon DeDeo » February 28 2007

I found this strange (from the conclusions):
The constancy of H0 from global cosmic scales down to 4.4Mpc or even 2.5Mpc (see also Ekholm et al. 2001) in spite of the inhomogeneous mass distribution requires a special agent. Vacuum energy as the solution has been proposed by several authors (e.g. Baryshev et al. 2001; Chernin 2001; Chernin et al. 2003a,b; Thim et al. 2003; Teerikorpi et al. 2005). No viable alternative to vacuum energy is known at present. The quietness of the Hubble flow lends support for the existence of vacuum energy.
What is going on here? I was left unsure if the authors were claiming that, the predictions of linear theory for the velocity field on largeish (e.g., 30 Mpc) scales were being called into question.

I took a quick look at Baryshev's paper from 2001, and do not yet understand the claims. It seems actually to be a rather subtle thing (as far as I can tell) that is being claimed: that velocity dispersions in the "translinear" regime (< 10 Mpc) are lower than "expected" (relative to what, I am not yet sure.)

But there are two questions here:

1. is the dispersion of the measured Hubble constant at a particular distance lower than expected?

2. are there systematic deviations due to coherent velocity flows that show up as a distance-dependent H_0? ("Hubble bubble", etc.)

Baryshev's paper seems to address #1 (which talks about velocity dispersions on the outskirts of the local group -- do we really understand the velocity profile of clusters well enough to say this is a problem yet?), whereas the conclusion of the HST paper seems to address #2 -- and #2 should really be something that we can address with linear theory. Is the apparent constancy of H_0 given the HST error bars consistent with what you would expect from the standard LCDM?

Anyway, sorry if this is a somewhat confused comment. Just poking around in the paper and this caught my eye.

Syksy Rasanen
Posts: 119
Joined: March 02 2005
Affiliation: University of Helsinki

[astro-ph/0603647] The Hubble Constant: A Summary of the HST

Post by Syksy Rasanen » March 01 2007

One can separate two things: the coincidence of the local mean and global Hubble parameters, and the small dispersion around the local mean. The latter issue is known as the "coldness" of the local Hubble flow.

How much of a problem the dispersion is, I don't know (see astro-ph/0107104, astro-ph/0201524, astro-ph/0212538, astro-ph/0308415, astro-ph/0412583, astro-ph/0603226 for analyses). However, the agreement between the global and local values is certainly surprising. Given that the scale of homogeneity is at least 70-100 [tex]h^{-1}[/tex] Mpc, having the global Hubble parameter at already 4 Mpc seems strange.

Sandage et al (and Baryshev, Chernin and Teerikorpi, astro-ph/0011528, astro-ph/0603226) advocate that the coldness is local evidence for vacuum energy. However, it's not clear to me whether the coldness could not be due to some peculiarity of the local region. (See e.g. the conference proceedings by van de Weygaert and Hoffman, http://adsabs.harvard.edu/cgi-bin/nph-b ... c12bf05039 .)

At any rate, the theoretical explanation of this problem is separate from the issue of the observational global value of the Hubble parameter.

Simon DeDeo
Posts: 44
Joined: October 26 2004
Affiliation: Santa Fe Institute
Contact:

[astro-ph/0603647] The Hubble Constant: A Summary of the HST

Post by Simon DeDeo » March 01 2007

I guess what I don't understand about the problem here is "cold relative to what" -- what about our current understanding about how velocities work on 4-30 Mpc scales fails? Yes, the distribution of matter is getting very inhomogenous on these scales, but is it sufficiently inhomogenous to require some new homogenizing mechanism for velocities?

Just browsing the papers on this subject -- including Sandage's remarks -- I haven't seen where someone goes through the "standard" picture to show that the constancy of the Hubble parameter given the error bars is unusual.

For example, we've been doing velocity-mass relationship studies for many years (POTENT) -- did those studies discover anything unusual about the mass-velocity relationship?

I totally get that this is separate from the determination of the global H_0, though -- just taking the discussion in a different direction here.

Syksy Rasanen
Posts: 119
Joined: March 02 2005
Affiliation: University of Helsinki

Re: [astro-ph/0603647] The Hubble Constant: A Summary of the

Post by Syksy Rasanen » March 01 2007

Simon DeDeo wrote:Just browsing the papers on this subject -- including Sandage's remarks -- I haven't seen where someone goes through the "standard" picture to show that the constancy of the Hubble parameter given the error bars is unusual.
You're right - people have looked at the velocity dispersion for different models (both for the background cosmology and the local environments), but I haven't seen similar work on the Hubble parameter.

Naively the agreement of the Hubble rates seems surprising, because there are regions of a few Mpc in size which are collapsing, and 20 Mpc Mpc underdense regions which expand faster than the average (and occupy a large fraction of the universe), so one would expect the variance on Mpc scales to be large. (Also, the galaxy distribution is very inhomogeneous on scales below 100 Mpc or so, astro-ph/0411197, astro-ph/0406202, astro-ph/0501583, so one would expect the same from the expansion rate.)

So happening to be located in a region around which things are very smooth seems surprising. But I don't know whether anyone has actually calculated the odds.

Post Reply