Transcript HST Key Project to Measure the Hubble Constant from

HST Key Project
on the Extragalactic Distance
Scale
Presentation by Alaine Ginocchio
May 5, 2006
HST Key Project


Measuring an
accurate value of H0
was one of the
motivating reasons
for building
NASA/ESA Hubble
Space Telescope
(HST).
Measurement of H0
with the goal of 10%
accuracy was
designated as one of
the three “Key
Projects” of the HST
–Mid 1980’s
Measuring H0
Overall Goal: measure H0 based on a
Cepheid calibration of a number of
independent, secondary distance
determination methods
 Locate Cepheids, measure distances to
galaxies, calibrate secondary methods:

–
–
–
–
–
Tulley-Fisher relation
Type Ia supernovae
the fundamental plane for elliptical galaxies
surface brightness fluctuations
Type II supernovae
Outline
 The
status of H0 prior to HST
 Using Cepheid variable stars to
measure distances
 Secondary Methods
 The Key Project
 Results
The Hubble Constant

1929 Edwin Hubble publishes results: correlation between the
distances to galaxies and their recession velocities v=H0d
(evidence that the universe is expanding)

It has been over 7 decades trying to pin down accurate H0,
primary difficulty establishing accurate distances over
cosmologically significant scales
Prior to HST


Arguing over H0 by factor of 2 (50-100km s-1 Mpc-1)
Atmospheric seeing set limit for resolving Cepheids to
only a few Mpc’s, confined to search Local Group of
galaxies and very nearest surrounding groups:
– 5 galaxies with well-measured Cepheids provided absolute
calibration for Tully-Fisher relation
– A single Cepheid distance provided calibration for surface
brightness fluctuation method
– No Cepheid calibrators for Type Ia Supernovae

HST extends seeing limit 10 fold (vol thousand fold)
Cepheid Variable Stars

1907, Henrietta
Leavitt (1868-1921),
Harvard Observatory
Astronomer, discovers
CV’s luminosity and
period follow very
tight correlation

Longer period =
brighter CV
Why Cepheids Pulsate


Vary in luminosity because they actually pulsate in and out
in size (get brighter as they grow larger, dim as they shrink)
Can’t achieve balance of power welling up from core and
power radiating from surface




Upper layers of star too opaque, pressure blds beneath photosphere, star expands in
size
Outer layer puffs, becomes more transparent, energy escapes, underlying pressure
drops, star contracts
Larger and more luminous CVs take longer to pulsate in and
out , period ranges from 2-150 days
Light curve with saw tooth signature P~50 days
Type I: heavy element content similar to sun
Cepheid Variables







Det luminosity to within ~10%
Most important primary distance indicator for nearby
galaxies
2 populations of CVs (Hubble’s mistake )
Type one: classical type,
~4 times brighter than type 2,
have a high metallicity, found in disk.
Type two: older stars, low metallicity,
found in halo
Reach is limited to ~ 30Mpc
But you can use it to calibrate secondary methods to
extend reach
Secondary Methods
Tully-Fisher Relation








Most commonly applied distance
indicator (2001)
Relationship between total luminosity
spiral galaxy and rotation speed of
disk
Faster rotation=more luminous
L~V4 scaling law appears to be a
general property of galaxies
Scatter +-0.3mag or +- 15% distance
for a single galaxy
Total (corrected to face on) luminosity
is strongly correlated with the max
rotation velocity of galaxy (corrected
to edge-on inclination)
Flat rotation curves of spirals which
produce well defined circular
velocities
Find speed from Doppler shift, tells
you L, makes galaxy a standard
candle
Secondary Method: Ia Supernovae






White dwarf supernovae,
exploding white dwarf stars that
have reached the 1.4 solar mass
white dwarf limit
Should have same luminosity
because come from stars of same
mass that explode in same way
Determine true luminosity by
using CVs to measure distance to
galaxies
Confirm L of all WD SNs are
about the same
BRIGHT: 10 bill solar luminosities
at peak, detect ~ 10 bill lt-yrs
away
But, occur only once every few
hundred years and need to catch
within 1-2 days of explosion
White dwarf supernovae in galaxies
approx 10 billion lt yrs away.
Upper image shows what galaxies
look like without supernovae .
HST Images
Supernova 1994D in Galaxy NGC 4526, May 25, 1999, HST Key Project
Secondary Methods

Fundamental plane for elliptical galaxies
– EG: correlation between the stellar velocity dispersion and the intrinsic
luminosity
– EGs found to occupy a “fundamental plane” in which a defined
effective radius is tightly correlated with the surface brightness with
that radius and central velocity dispersion
– Scatter ~10-20% distance for individual cluster

surface brightness fluctuations
– The resolution of stars in galaxies is distance dependent
– Elliptical or spirals with prominent bulges
– Normalizing to the mean total flux and correcting for an
observed color dependence, relative distances to galaxies can
be measured
– Intrinsic scatter factor of 3 improvement to T-F or F-Plane

Type II supernovae
– Supernova from massive stars, fainter and show wider
variation in luminosity than IaSN
– Baade-Wesselink technique: follow spectral fits of color T, flux
and radial vel of envelope over time to det dist.
– Applied independent of local calibration of extragalactic
distance scale but verified with galaxies of known distances &
diversity of methods constrains overall systematic errors
Key Project





1984
1991
1998
2001
Conceptually Begins
First Observations
Last observations finished
Final results published
Main goal: determine H0 to an accuracy of
10% by:
(1) refining the cosmic distance scale by observing
18 spiral galaxies within 20 Mpc, searching for and
measuring the periods and magnitudes of Cepheid
variables
(2) calibrating the secondary distance indicators
(3) Measure H0 with secondary distance indicators
and combine
Total Cepheid Calibration Sample
18 galaxies: Key Project
 8 galaxies: reanalyzed HST archival data
observed by other groups
 5 very nearby galaxies
 31 Galaxies: subsets used to calibrate
secondary models



Discovered 800 CVs in 18 galaxies out to
65 million lt yrs.
CV positions magnitudes and periods available on
http://www.ipac.caltech.edu/H0kp/H0KeyProj.html
Velocity v. distance for galaxies with CV distances
Strengths and Weaknesses of CV s






Among brightest stellar
indicators, relatively young
stars, found in abundance in
spiral galaxies
Many independent objects can
be observed in a single galaxy
Large amplitude & distinct
sawtooth light curve facilitate
discovery
Long lifetimes: can be
observed at other times and
wavelengths
CV P-L relationship has small
scatter (I-band dispersion
~+-0.1 mag)
Have been studied and
theoretically modeled
extensively and their physics
is understood





Young stars so found in
regions with dust scattering,
absorption and reddening.
Corrections must be made
for extinction with
assumptions
Metallicity: dependence of PL
relationship on chemical
composition very difficult to
quantify (PL relationship has
slight dependence on
metallicity)
Accurate calibration of PL
relation at any given
metallicity not yet
established
Crowding effects
Limited reach ~30 Mpc
Methodology






Combine values of H0 from different secondary
methods
All based on a common Cepheid 0-point but with
different subsets of Cepheid calibrators
Take out peculiar velocity
Used 3 methods: classical frequentist analysis,
Bayesian analysis and weighting scheme based
on numerical simulations
All in good agreement
Conclusion: combined value is robust, the choice
of statistical method not determine result
Uncertainties in H0 for Secondary Methods
Method
H0
Error
(random, systematic) Refs
36 Type Ia SN, 4000\cz\30,000 km s~1 . . . . . . 71
2
21 TF clusters, 1000\cz\9000 km s~1 . . . . . .. . 71
(%)
6
1, 2, 3, 4
3
7
5, 6, 7
11 FP clusters, 1000\cz\11,000 km s~1 . . .... . 82
6
9
8, 9
SBF for 6 clusters, 3800\cz\5800 km s~1 . . ... . 70
5
6
10, 11
4 Type II SN, 1900\cz\14,200 km s~1 . . . . . .. . 72
9
7
12
“\” = “<“
NOTE: Combined values of (random) km s~1 Mpc~1 (Bayesian), H0: H0\722 H0\723
(random) km s~1 Mpc~1 (frequentist) ; (random) km s~1 Mpc~1 (Monte Carlo) H0\723
REFERENCES.È(1) Hamuy et al. 1996; (2) Riess et al. 1998; (3) Jha et al. 1999; (4) Gibson et al. 2000a;
(5) Giovanelli et al. 1997; (6) Aaronson et al. 1982, 1986; (7) Sakai et al. 2000; (8) Jorgensen et al. 1996;
(9) Kelson et al. 2000; (10) Lauer et al. 1998; (11) Ferrarese et al. 2000a; (12) Schmidt et al. 1994.
Hubble diagram of distance vs. velocity for secondary distance indicators calibrated by
CVs. Velocities are corrected for the nearby flow model of Mould et al. (2000). A slope of
H0=72 is shown, flanked by +-10% lines. Beyond 5000 km s-1 (vertical line), both
numerical simulations and observations suggest that the effects of peculiar motions are
small.
Combined
RESULT FOR H0
H0=72 +-8 km s-1 Mpc-1
THE END