Constraining%20dark%20energy%20with%20Con-X%20galaxy%20clusters

1
Constraining dark energy with Con-X galaxy
clusters

• David Rapetti, KIPAC (Stanford/SLAC)

In Collaboration with
Steve Allen (KIPAC) Robert Schmidt
(Heidelberg) Harald Ebeling (Hawaii) Andy Fabian
(Cambridge) Jochen Weller (UCL) Alexey Vikhlinin
(CfA) and the Con-X Facility Science Team
2
Outline

• The X-ray gas mass fraction, fgas(z) experiment
  constrains dark matter and dark energy
• Current constraints (Allen et al 2004, 2006
  prep., Rapetti et al 2005)
• Notes on systematics
• Simulated Constellation-X fgas(z) data set
• Projected constraints on dark energy from Con-X
  fgas(z)
• Other Con-X dark energy studies X-raySZ, growth
  of structure
• Conclusions

3
The fgas(z) experiment
4
Measuring the X-gas mass fraction of X-ray galaxy
clusters
Mgas X-ray gas mass Mtot the total cluster mass
We define
From White et al 93, Fukugita et al 98, Voevodkin
Vikhlinin 04
Then
The matter content of rich clusters of galaxies
is expected to provide an almost fair sample of
the matter content of the Universe (White Frenk
91, White et al. 93, Eke et al.98). Then
b is the bias factor that accounts for the
relatively small amount of gas expelled when the
cluster forms.
HSTBBNS priors when clusters alone
5
Spanning a redshift range to constrain dark energy

• Absolute distance measurements at different
  redshift between 0.06ltzlt1.07 for the current
  sample of 41 Clusters (to be published).

Mgas ? dA(z)2.5 Mtot ? dA(z)

• It is crucial to use dynamically relaxed clusters
  for the fgas measurement
• Regular X-ray morphology
• Low ellipticities
• Minimal centroid variation
• Sharp central brightness peaks centred on their
  dominant elliptical galaxies.
• Weighted mean scatter about the best-fitting 12
  (8 in distance).
• For the cosmological parameter estimation
  together with SNe Ia and CMB data we use the
  Montecarlo Markov chain technique implemented in
  COSMOMC (Lewis Bridle 02) and CAMB (Lewis,
  Challinor Lasenby 00) to calculate the cmb
  power spectrum.

6
Current constraints
7
Constraints on ?CDM model
8
Constant equation of state parameter w

• 68.3 and 95.4 confidence limits
• assuming flat prior.
• Note the small overlaping
• region for the combination of the
• three data sets. It comes from the
• complementary nature of SNCMB
• (WMAP1stCBIACBAR) and
• ClustersCMB.
• Marginalized constraints (68.3)
• w0 -1.05 0.11
• ??????Wm 0.29 0.03
• Note that using only Clusters
• ??????Wm 0.24 0.04 (Allen et al 04)

Rapetti, Allen Weller (2005)
9
Clusters data breaks the degenaracy Wbh2,t,ns
SNCMB (WMAP1stCBIACBAR)
ClCMB (WMAP1stCBIACBAR)
WMAP3 nicely confirms the ClCMB trend ns
0.95 0.015 - 0.019 ?bh2 0.0223 0.0007 -
0.0009 ? 0.09 -0.03
Plus ?80.740.05-0.06 WMAP3 ?80.72-0.04
(Voevodkin Vikhlinin 04) ?80.70-0.04 (Allen
et al 03) ?80.75-0.05 in preparation
10
New fgas(z) constraints (41 clusters) combined
with two SNe Ia data sets
SNLS (Astier 05)Clusters
SNLS (Riess 04)Clusters
11
Notes on systematics
12
No evidences for systematics in current data
For r0.25rvir(Chandra obs) b0.83-0.03 (10
systematics)
Assuming ?CDM we fit for evolution in this
parameter
Relaxed (sim) Unrelaxed (sim) Observations
Plus, we find no evidence for trend of fgas with
temperature kT.
13
Constellation-X analysis
14
Constellation-X

• Con-X is one of two flagship missions in NASAs
  Beyond Einstein Program ranked second only to
  JWST. Due for launch 2018.
• The most important single feature for dark
  energy studies is its large collecting area (two
  orders-of-magnitude with respect to previous)
  with 5 angular resolution.
• Con-X will carry out two powerful and independent
  sets of tests of dark energy
• Measure the absolute distances to clusters
  primarly using measurements of the fgas(z) in the
  largest, dynamically relaxed clusters with
  additional constraining power provided by
  follow-up SZ observations.
• Together with theoretical models for the mass
  function and X-ray and SZ cluster surveys Con-X
  will help to constrain the growth of structure.

15
Proposed sample to constrain dark energy with the
Con-X fgas(z) experiment

• Construct an expanded catalogue of the most X-ray
  luminous, X-ray brightest clusters reaching to
  faint flux levels with an X-ray all-sky survey.
  Large-area SZ surveys like the South Pole
  Telescope and Planck can also provide Con-X
  targets.
• From the 2000 most luminous clusters in the
  catalogue, select the most dynamically relaxed
  clusters with short Con-X snapshot (1ks)
  observations (2Ms total). The current MACS sample
  indicates 1/4 of them are enough relaxed. We
  conservatively assume 1/8 (500 clusters) for our
  study (we also checked 1/16 obtaining similar
  constraints assuming the same total Con-X time).
• Modest 10-12Ms of Con-X observing time (10 over
  5 first years). This implies a typical exposure
  time of 20ks per cluster. This leads to 5
  statistical error bars.
• We adopt the baseline effective area and spectral
  resolution describe at http//constellation.gsfc.n
  asa.gov/ and 5 angular resolution which is
  enough for the removal of point source flux
  contamination and to identify relaxed clusters at
  high z.

16
Predicted Con-X fgas(z) data
Allen et al. (2003)

• Evrard et al (2002)

Jenkins et al (2002)
17
Predicted Con-X fgas(z) data set
Required Flux Limit Survey
ROSAT All-Sky
The predicted fgas(z) values for the proposed
Con-X survey of 500 clusters with individual fgas
uncertainties of 5. A systematic scatter of 4
due to cluster-cluster variations in the bias
parameter b is included. The median redshift for
the sample is z1.
The predicted number density of clusters with
bolometric X-ray luminosities LXgt2x1045h-270
erg/s for a cluster survey flux limit of
5x10-14erg/cm2/s in the 0.1-2.4 keV band
(redgas(z) values for the proposed Con-X survey
of 500 with a median redshift z 1. Blue curve
shows for the ROSAT All-Sky survey flux limit
5x10-12erg/cm2/s.
18
Constraints from Con-X alone

• Con-X simulated data with 5 measurement errors,
  4 systematic scatter in b and using 2 priors on
  ?bh2, h and b.
• Evolution model w(a)w0wa(1-a)w02w(1-a).
• Comparable accuracy and beautifully complementary
  to LSST, SNAP, BAO, Planck, and cluster growth.

19
Current Chandra vs projected Con-X constraints
Clusters Chandra

• 68.3 and 95.4 confidence limits.
• Note the combination of different data sets will
  be crucial for controling the systematics of each
  individual experiment.

SNIa Gold
Clusters Con-X
CMB
20
Constraints on dark energy evolution using
Con-XWMAP8 simulated data

• Here instead of using the 2 priors on ?bh2, h
  we combine Con-X clusters with a simulated cmb TT
  power spectra after 8 years of WMAP as described
  in Upadhye, Ishak Steinhardt 2005.
• We use the dark energy model w(a)w0wa(1-a)w02
  w(1-a) to compare with other future experiments
  of SNe Ia and BAO shown in Linder 04,05.
• Our constraints are comparable and nicely
  complementary to those other experiments.

21
Other Con-X dark energy studies
22
Constraints from the Con-X X-raySZ experiment
Using the same Con-X data (most luminous and
hottest clusters), i.e., no extra time. We can
also measure distance independently of z by
combining X-raySZ observations of galaxy
clusters.
X-raySZ 2
X-raySZ 0.1
yobs observed by a SZ survey (we assume
2errors) yref predicted by the X-ray data
assuming a reference cosmology (5 errors).
fgas Con-X

• Though the X-raySZ experiment is less powerful
• (y?d0.5) than the fgas?d1.5 but provides
• i) Precise redshift independent distance
  measurment
• ii) Test for the assumption of fgas constant with
  z
• iii) X-raySZ independent of b and hydrostatic
  equilibrium.

Growth of Structure G(z) will be a powerful
complementary dark energy observable 1000
clusters constraints lt0.5 on G(z) leading to
-0.06-0.08 in w0 (Majumdar Mohr 04)
23
Conclusions

• In terms of direct distance measurements the
  X-ray method provides similar accuray to SNIa and
  BAO studies and has several valuable features
• The physics of galaxy clusters is relatively
  simple and can be accurately modeled by
  simulations.
• Clusters can be revisited with X-ray
  observatories to improve signal-to-noise.
• The fgas technique includes an additional
  constraint on ?m from the normalization of the
  curve.
• The combination of fgas and CMB data breaks a
  number of important degeneracies.
• The systematic scatter in the fgas(z) is small
  (undetected and lt12 in current data).
• Direct checks on the key assumptions in the fgas
  method are possible using the spectral/resolution
  capabilities of Con-X and by combining with other
  data (X-raySZ).
• Additional powerful and independent constraints
  on dark energy will be provided by Con-Xs
  contribution to growth of structure studies.