Transcript Document 7294982

Baryonic Acoustic Oscillations: BAO
Outline:
BAO: standard ruler
Present detections of BAO (SDSS, 2dF)
Simulations: predictions
 SKA: billion galaxies at 21cm, + WL  DE
Françoise Combes
MCCT-SKA, August 2009
History of the universe
2
Hu & White 2004
Acoustic oscillations
Seing sound:
hotter and colder regions
Compton scattering coupling
DM does not oscillate,
but is at the center
of fluctuations
Wayne Hu’s page
3
Perturbation Development (0)
Let us follow a given
perturbation
Standard theory: adiabatic
All components have the
same amplitude
Except 4/3 x for relativitic
Photons and neutrinos
(not taken into account here)
Linear perturbations:
All to be summed up
4
From Eisenstein
Perturbation Development (1)
Neutrinos stream away
DM is continuing
slow collapse
Stays at the center
Gas and photons:
Oscillating plasma
Coupled by scattering
Behaves as single fluid
Overpressure,
Propagates at 57% 5c
From Eisenstein
Perturbation Development (2)
Plotted are relative dr/r
So density ratios of W
not visible
However, rR2 plotted
Mass= area under curve
Matter diffuses, and peaks
Widen in the DM
The streaming has
suppressed the small scales
Turnover in P(k)
6
Perturbation Development (3)
Beginning of recombination
Photons decouple from gas
Small scale blurred
(Silk damping)
The sound speed drops
And the pressure wave stalls
7
Perturbation Development (end)
Photons and neutrinos
have streamed away
Left is a DM concentration
in the center + ring of gas
of 150 Mpc
8
After recombination (1)
DM and baryons are
attracting each other
Gravity homogeneises
the fluctuations
9
After recombination (2)
Finally, DM being
5-6 times heavier,
imposes a central peak,
Although no gas initially
These are the acoustic
peaks to be observed
10
After recombination (3)
Now in density, the
peak is of order 1%
Not rR2 any longer
Expected: a central
concentration at the
same place of initial
fluctuation +
153Mpc away a second
density peak
First Predictions: Peebles & Yu (1970), Sunyaev & Zeldovich (1970)
11
Animated version: mass
12
From Eisenstein, & CMBFAST
Animated version: density
From Eisenstein & CMBFAST
13
Anisotropies
in the CMB
1990’: COBE 7°
2000’: WMAP: 0.3°
T/T ~ 10-5 !
linear oscillations
Rs=153 +2 Mpc
COBE
WMAP
14
Correlation functions
Eisenstein et al 2005
k = 2p/l P(k) = power spectrum
= |dk|2
Position of the peak 150Mpc
With width 15 Mpc (Silk damping)
15
Linear power spectrum
Scale of the bend in P(k)
is the size of the horizon at
matter-radiation equality
60 000 yrs after Big-Bang
fn = Wn/Wdm = 1/3
Smn = 3.7ev
Free-streaming reduces
Small scale structure
16
Waves in Fourier Space
A crest launches a planar sound wave,
which at recombination may or may not be in phase
with the next crest
Get a sequence of constructive and destructive interferences
as a function of wavenumber
Peaks are weak —
suppressed by the baryon fraction
Higher harmonics
suffer from Silk damping
at ~10 h-1 Mpc
17
Eisenstein et al 1998
Expected oscillations
Not in phase
At small scale
(velocities)
And twice the
Wavelength
Hütsi 2005
18
M. White 2007
A single perturbation
Creates a depression
 Sound wave at c /√3
Sound horizon
at recombination
R~150Mpc
Galaxies form in peaks
 baryon correlations
From Daniel Eisenstein
19
Random perturbations
lSignal reduced by
random phases
lNo unique shell
lBut 1% in the P(k)
From Daniel Eisenstein
20
BAO: Standard Ruler
Radial BAO: dr = (c/H)dz
Plane of the sky: dr = DAdq
Observer
cz/H
Much better than CMB
3D instead of 2D!
qD
cz/H = qD
Possibility to
determine H(z)
Alcock & Paczynski (1979)
Test of cosmological cst
Could test the bias b
Or b = Wm0.6/b
21
Advantage of radial BAO
DA and DL are integrated intensities, while H(z)=R’/R is not
(in a flat universe)
(R’/R)2 = 8pG r/3 + /3
3 R¨/R = - 4pG (r + 3 P) + 
WM matter, WQ dark energy Wk curvature,
P =wr
22
Mapping Our History
The subtle slowing
down and speeding
up of the expansion,
of distances with
time: a(t), maps out
cosmic history like
tree rings map out the
Earth’s climate
history.
STScI
Baryonic acoustic peak
Waves detected today
In the distribution of baryons
50 000 galaxies SDSS
Eisenstein et al 2005
24
Detections: SDSS
Eisenstein et al. (2005)
25
Detections: 2dFGRS
Comparaison with
previous estimations
for 2dF
z< 0.3
Percival et al (01, 03)
Tegmark et al 04
Cole et al. (2005)
Final= 221414 galaxies, grey-scale 1s errors
26
BAO Surveys
LAMOST
WFMOS/Subaru
HETDEX/Virus
LSST, etc..
From Schlegel (2008)
27
Detection of radial BAO ?
p
without
s
with lensing
Previously, monopole term,
averaged on a sphere, but
s (transverse), and p (l.o.s.):
pb of peculiar motions + lensing
Ring expected at 100 h-1 Mpc
Enhanced by lensing
p
Gaztanaga et al (2008)
28
Detection of 3D BAO ?
z= 0.15-0.47 (all)
Model Wb=0.045, n=1 ,
z=0.15-0.30
Z=0.15-0.30,
z=0.40-0.47
Model Wb=0.06, n=1.3
29
Gaztanaga et al (2008)
Detection or noise ?
z=0.15-0.30 H(z=0.24) = 79.69+2.32 km/s/Mpc
z=0.40-0.47 H(z=0.43) = 86.45+3.27 km/s/Mpc
Gaztanaga et al (2008)
But "Comments" by Miralda-Escude (2009) astro-ph:, 0901.1219
In fact the BAO radial term is 10 times lower than the present
limit of the observations. Gaztanaga et al claim that this factor 10
increase is due to lensing, but the lensing effect cannot be larger
than 10-4!
Or else unrealistic bias should be considered
30
DR7 release SDSS
Compatible with DR3, less noise (Kazin et al 2009)
31
BAO with DR7 SDSS
WMAP5
Union Supernovae
BAO SDSS
Percival et al (2009)
32
0.2 < z < 0.35
Pivot z=0.275
BAO combined, Wm, H0
w=-0.97+0.10
H0=68.2 +2.2km/s/Mpc
Wm = 0.282+0.015
WMAP5 + BAO
WMAP5+SN
WMAP5+SN+BAO
Percival et al (2009)
33
Future surveys, SDSS, LAMOST..
Slice 500h-1 Mpc across, 10 h-1 Mpc thick 100km from Beijing
34
Luminous Red Galaxies
Common galaxies
LRG
Quasars
LRG are big ellipticals in clusters
Strongly clustered
35
High z or low z? Space or ground?
DE effects are at low z, but degeneracy with curvature
36
Non linearities
Can mix scales
and wash out structures
M. White 2005
Full line:linear prediction
Ratio with a smooth
spectrum
37
Nonlinearities
Real-space
Redshift-space
A maximum wavenumber is set in each
redshift bin to reduce the contamination
of nonlinear evolution.
Seo & Eisenstein (2005)
38
Cosmological Simulations
Millenium
Springel et al 2005
HORIZON
L=1 billion light-yrs
39
T=1, 3 & 13 billion yers
Simulations
Springel et al 2005
Power spectra of
DM and galaxies
in the BAO region
(divided by a CDM
linear power spect.)
Blue: scatter
Black: mean
40
Testing models of gravity
Yamamoto et al 2006
Dvali-Gabadadze-Porrati DGP model
41
KP 4 - Galaxy evolution and cosmology
1- HI line surveys
All-sky survey would contain a billion galaxies out to z~1.5
 Galaxy evolution studies using the most abundant element
2- ‘Dark Energy-measuring-machine’
- acoustic peaks in baryons as function of z
- weak gravitational lensing in large fields
Measure DE parameters w0 and w1 to 1% accuracy
42
Galaxies at high z
Possible in CO, not in HI
M83: optical
HI at z  0.03
CO at z = 4

Ultraluminous galaxy
43 M83
Milky Way-like spiral (109 M of HI):
Maximum redshift for a 360 hour integration with SKA
2000 galaxies/
deg
M 101
100 000
30 000
Star formation with z
M 51
SMC
Crucial epoch
44
Dark energy measuring machine
Goal: improve the accuracy in measurement of w from 10% to 1%
DE equation of state: w(z)=P/r = w0 +w1z
CMB observations are limited for this purpose, need independent
measurements (supernovae, redshift surveys, H0, …)
Planck: measure Wmh to 1%, but degeneracy:
uncertainties in sound horizon size Rs
(sensitive to mix of baryons/CDM/HDM) makes our
cosmic ruler squashable (by 2%) - and w can be measured to 10% only:
45
Planck: degeneracy
Dark energy measuring machine: I-Wiggles
Map the acoustic oscillations, or wiggles, in the galaxy power spectrum P(k)
as function of redshift:
Only the SKA can get the required billion all-sky redshifts out to z=1.5
II- Weak Shear
10 billion galaxies,
10 nanoJy
46
But precision is not all!
Bias fb=Wb/Wm assumed linear (Blake et al 2004)
SKA contour
In red
SDSS in blue
47
Worries about systematics, so that targeted experiments should certainly be cleaner.
Testing DE with wiggles
48
Note, still needs `priors’ on Wm h2 and h (with Planck and/or, e.g., SKA masers
)
/JDEM
JDEM and SKA
compared
Resolution and
Sky area
2000 SN 0.1<z<1.7
49
Conclusion
Baryon acoustic oscillations measure Rs (150 Mpc, CMB-calibrated)
and are used as a standard ruler to measure the distance DA and H(z)
Bias measured by Alcock-Paczynski test
SKA will be unique for large volume and high resolution
1 billion galaxies at 21cm, 10 billion for WL
Together with WL, determine DE parameters with 1% precision
Non-linearities: to be simulated with cosmological simulations
50