Hubble constants reported in the literature since 1920
Hubble constants reported in the literature since 1970
- Microlensing searches; Calibrating the S/LMC P-L relation:
Galactic Cepheid measurements are of limited utility because
there are relatively few near the Sun and the ones that are near
suffer from major extinction problems that are difficult to
quantify. The LMC and SMC provide samples of Cepheids all at
about the same distance and in large quantities with little
extinction. Microlensing searches such as EROS, MACHO and OGLE have had
telescopes trained on the S and LMC since the early 90s. This
has allowed the light curves for literally millions of stars and
thousands of Cepheids to be measured. Specific cepheid papers for
EROS, MACHO and OGLE can be, respectively, found here: Sasselov
et al. 1997, Alcock
et al. 1995 and Udalski
et al. 1999.
Unfortunately, both the EROS and MACHO campaigns use non-standard
bands and are not of much use for refining Cepheid P-L
relations. The OGLE campaign does use standard bands and has
come up with very precise values for the slope and zero-point of
the P-L relation.
- The
Hubble Space Telescope Key Project (Freedman
et al. 2001 and references therein):
This large team (abbreviated as HSTKP hereafter) has over 29
papers since it's inception and is tasked by STScI with measuring
H0 via Cepheids. They use the S/LMC P-L relation and a
distance modulus of 18.50 mag (50 kpc) as a fiducial starting
point (see Freedman,
Wilson and Madore 1991 and Madore
and Freedman 1991). Their goals are:
- Using HST's high resolving power to measure Cepheids to 20Mpc
- Apply this Cepheid calibration to several secondary distance
indicators that go farther than 20Mpc into the Hubble flow.
- Compare the Cepheid and other distances to compare
uncertainties
- Conduct tests as to the universality of the P-L relation, with
specific attention to metallicity effects.
Their results? They get H0 = 72
±3 (random) ±7 (systematic) km/s/Mpc by
using 5 secondary distance techniques that are calibrated with
Cepheids and 3 weighting schemes between the 5 H0
obtained.
Galactic Cepheid measurements are of limited utility because there are relatively few near the Sun and the ones that are near suffer from major extinction problems that are difficult to quantify. The LMC and SMC provide samples of Cepheids all at about the same distance and in large quantities with little extinction. Microlensing searches such as EROS, MACHO and OGLE have had telescopes trained on the S and LMC since the early 90s. This has allowed the light curves for literally millions of stars and thousands of Cepheids to be measured. Specific cepheid papers for EROS, MACHO and OGLE can be, respectively, found here: Sasselov et al. 1997, Alcock et al. 1995 and Udalski et al. 1999.
Unfortunately, both the EROS and MACHO campaigns use non-standard bands and are not of much use for refining Cepheid P-L relations. The OGLE campaign does use standard bands and has come up with very precise values for the slope and zero-point of the P-L relation.
This large team (abbreviated as HSTKP hereafter) has over 29 papers since it's inception and is tasked by STScI with measuring H0 via Cepheids. They use the S/LMC P-L relation and a distance modulus of 18.50 mag (50 kpc) as a fiducial starting point (see Freedman, Wilson and Madore 1991 and Madore and Freedman 1991). Their goals are:
- Using HST's high resolving power to measure Cepheids to 20Mpc
- Apply this Cepheid calibration to several secondary distance indicators that go farther than 20Mpc into the Hubble flow.
- Compare the Cepheid and other distances to compare uncertainties
- Conduct tests as to the universality of the P-L relation, with specific attention to metallicity effects.
Their results? They get H0 = 72 ±3 (random) ±7 (systematic) km/s/Mpc by using 5 secondary distance techniques that are calibrated with Cepheids and 3 weighting schemes between the 5 H0 obtained.
Hubble diagram for HSTKP Cepheid distances
Hubble diagram for HSTKP secondary indicators
- The Sandage Consortium (Saha,
Sandage et al. and references therein):
The Freedman et al. H0 Cepheid group was awarded the
time for their project... except for the calibration of SNe Ia
against Cepheids. The time for this part of the Key Project went
to Allan Sandage and his team. It's interesting to note that the
Sandage group get H0 = 58.7
±6.3 (random + systematic) km/s/Mpc while, using
virtually the same data, the Key Project gets H0 = 71 ±2 (random) ±6
(systematic) km/s/Mpc.
Most of this discrepancy lies in the fact that the two groups use
different P-L relationships. The Key Project uses the large
sample from the OGLE-II microlensing group (Udalski
et al. 1999) while the Sandage group uses an older set of P-L
relations that were based on Madore
and Freedman 1991 that "are based, however, on limited
samples of stars with photometry obtained many years ago mostly
with photoelectric and photographic techniques what in crowded
fields may lead to systematic uncertainties." (from the OGLE-II
paper).
Even if Sandage et al. suck it up and adopt the Freedman et
al. Cepheid distances (which reduces their moduli by 0.25 mag on
average) they still get H0 = 67.7
km/s/Mpc
Why the heck is this? Well, the Sandage folks say that there are
three "rather subtle" reasons for this (from Saha,
Sandage et al.):
- Freedman et al. use only six SNe Ia (Sandage et al. use
nine) as calibrators and one (SN 1972E in NGC 5253) has a
questionable distance and contributes a higher weight than
it should. (2.5% reduction in H0 to 69.2
km/s/Mpc)
- There is a mismatch in the colors of the Freedman et
al. calibrating SNe Ia and the SNe Ia that they use at
farther distances. (2.6% reduction in H0 to 67.4
km/s/Mpc)
- "The calibrating SNe Ia lie in spirals and hence have slower
decline rates than the distant SNe Ia that come in all types
of galaxies." (~2% reduction in H0 to 66.1
km/s/Mpc)
- Other interesting papers I've read:
-
In "Median
Statistics, H0, and the Accelerating Universe"
Gott, J. Richard III et al. (2001) use Huchra's
gigantic H0 compilation and use median
statistics to come up with a value of H0 = 67 km/s/Mpc
- In "Arguments
for a Hubble Constant near H0 = 55" H. Arp
argues that the difference between Tully-Fisher and redshift
distances at larger distances point to spirals having an
"intrinsic redshift" similar to as if these galaxies were
being gravitationally attracted to unseen galaxies behind
them. I must admit I don't fully appreciate his line of
reasoning or even really understand it.
Next
Home
The Freedman et al. H0 Cepheid group was awarded the time for their project... except for the calibration of SNe Ia against Cepheids. The time for this part of the Key Project went to Allan Sandage and his team. It's interesting to note that the Sandage group get H0 = 58.7 ±6.3 (random + systematic) km/s/Mpc while, using virtually the same data, the Key Project gets H0 = 71 ±2 (random) ±6 (systematic) km/s/Mpc.
Most of this discrepancy lies in the fact that the two groups use different P-L relationships. The Key Project uses the large sample from the OGLE-II microlensing group (Udalski et al. 1999) while the Sandage group uses an older set of P-L relations that were based on Madore and Freedman 1991 that "are based, however, on limited samples of stars with photometry obtained many years ago mostly with photoelectric and photographic techniques what in crowded fields may lead to systematic uncertainties." (from the OGLE-II paper).
Even if Sandage et al. suck it up and adopt the Freedman et al. Cepheid distances (which reduces their moduli by 0.25 mag on average) they still get H0 = 67.7 km/s/Mpc
Why the heck is this? Well, the Sandage folks say that there are three "rather subtle" reasons for this (from Saha, Sandage et al.):
- Freedman et al. use only six SNe Ia (Sandage et al. use nine) as calibrators and one (SN 1972E in NGC 5253) has a questionable distance and contributes a higher weight than it should. (2.5% reduction in H0 to 69.2 km/s/Mpc)
- There is a mismatch in the colors of the Freedman et al. calibrating SNe Ia and the SNe Ia that they use at farther distances. (2.6% reduction in H0 to 67.4 km/s/Mpc)
- "The calibrating SNe Ia lie in spirals and hence have slower decline rates than the distant SNe Ia that come in all types of galaxies." (~2% reduction in H0 to 66.1 km/s/Mpc)
- In "Median Statistics, H0, and the Accelerating Universe" Gott, J. Richard III et al. (2001) use Huchra's gigantic H0 compilation and use median statistics to come up with a value of H0 = 67 km/s/Mpc
- In "Arguments for a Hubble Constant near H0 = 55" H. Arp argues that the difference between Tully-Fisher and redshift distances at larger distances point to spirals having an "intrinsic redshift" similar to as if these galaxies were being gravitationally attracted to unseen galaxies behind them. I must admit I don't fully appreciate his line of reasoning or even really understand it.