Download Cosmology with the CBI

CMB Polarization with the
CBI
Jonathan Sievers (CITA)
The CBI Collaboration
Caltech Team: Tony Readhead (Principal Investigator), John Cartwright, Clive Dickinson, Alison
Farmer, Russ Keeney, Brian Mason, Steve Miller, Steve Padin (Project Scientist), Tim Pearson,
Walter Schaal, Martin Shepherd, Jonathan Sievers, Pat Udomprasert, John Yamasaki.
Operations in Chile: Pablo Altamirano, Ricardo Bustos, Cristobal Achermann, Tomislav Vucina,
Juan Pablo Jacob, José Cortes, Wilson Araya.
Collaborators: Dick Bond (CITA), Leonardo Bronfman (University of Chile), John Carlstrom
(University of Chicago), Simon Casassus (University of Chile), Carlo Contaldi (CITA), Nils
Halverson (University of California, Berkeley), Bill Holzapfel (University of California, Berkeley),
Marshall Joy (NASA's Marshall Space Flight Center), John Kovac (University of Chicago), Erik
Leitch (University of Chicago), Jorge May (University of Chile), Steven Myers (National Radio
Astronomy Observatory), Angel Otarola (European Southern Observatory), Ue-Li Pen (CITA),
Dmitry Pogosyan (University of Alberta), Simon Prunet (Institut d'Astrophysique de Paris), Clem
Pryke (University of Chicago).
The CBI Project is a collaboration between the California Institute of Technology, the Canadian
Institute for Theoretical Astrophysics, the National Radio Astronomy Observatory, the
University of Chicago, and the Universidad de Chile. The project has been supported by funds
from the National Science Foundation, the California Institute of Technology, Maxine and Ronald
Linde, Cecil and Sally Drinkward, Barbara and Stanley Rawn Jr., the Kavli Institute,and the
Canadian Institute for Advanced Research.
Whence Polarization?
Polarization comes from Thomson
scattering during recombination
 Quadrupole seen by electrons
produces polarization.
 Dominant source from velocity
modes – so out of phase with TT
 Pol’n lines up with modes, so no BB
type
 Strongly predicted by TT – good
check on model.

No Dipole


No net extra
radiation from
left/right or
top/bottom, so
no polarization
Doppler shift
makes dipole
pattern, so
moving
electron in an
isotropic field
has no
polarization.
Quadrupole

More radiation
from left/right
than
top/bottom.
Electron moves
up/down, and
so scattered
radiation is
polarized.
E Mode Only



If velocity
converges,
then electron
moves normal
to wave +E
If diverges,
electron
moves along
wave –E
Never moves
tilted, so no B
radiation
The Instrument







13 90-cm Cassegrain
antennas
• 78 baselines
6-meter platform
• Baselines 1m – 5.51m
10 1 GHz channels 26-36 GHz
• HEMT amplifiers (NRAO)
• Cryogenic 6K, Tsys 20 K
Single polarization (R or L)
• Polarizers from U. Chicago
Analog correlators
• 780 complex correlators
Field-of-view 44 arcmin
• Image noise 4 mJy/bm
900s
Resolution 4.5 – 10 arcmin
The CBI Adventure…
CBI located at 5080 meters in
Atacama desert, Chile.
Area is used by NASA as a proxy
for Mars for testing/developing
equipment.
Land mines along border w/ Bolivia
Steve Padin wearing the cannular
oxygen system
The CBI Adventure…

Two winters a year!
The roads fill with
snow.
The CBI Adventure…

Volcan Lascar (~30 km away)
erupts in 2001
CMB Interferometers
Interferometers are a very clean way of measuring CMB.
Robust with respect to instrumental systematics.

Interferometers directly measure Fourier Plane
• Resolution set by field of view

Interferometers do not pick up large scale
power
• Gain fluctuations don’t see monopole/dipole

Polarization:
• Directly Measure E and B (modulo FOV)
• Correlate circular electric fields to directly
measure linear pol’n
• Do not subtract two nearly equal signals
uv coverage with mosaic beam
Polarization – Stokes parameters



CBI receivers can observe either RCP or LCP
• cross-correlate RR, RL, LR, or LL from antenna pair
Mapping of correlations (RR,LL,RL,LR) to Stokes
parameters (I,Q,U,V) :
Intensity I plus linear polarization Q,U important
• CMB not circularly polarized, ignore V (RR = LL = I)
• parallel hands RR, LL measure intensity I
• cross-hands RL, LR measure polarization Q, U
 R-L phase gives Q, U electric vector position angle
E and B modes
A useful decomposition of the polarization
signal is into “gradient” and “curl modes” – E
and B:



~
~
~
~
i 2 v
Q ( v)  i U ( v)  E ( v)  i B ( v) e
 v  tan
RL
k
V
1
v u 
E & B response smeared by
phase variation over aperture A
~
~
i 2 ( v  k )
RL
(u k )   d v Pk ( v) [ E ( v) i B ( v)] e
 ek
2
interferometer “directly” measures E & B!
CBI 2000+2001 (TT), WMAP, ACBAR, BIMA
Readhead et al. ApJ, 609, 498 (2004)
astro-ph/0402359
CMB
Primary
SZE
Secondary
EE – A Separate View
Excellent check on
consistency of standard cosmological
model.
One example: pathological primordial
spectra with very
different params can
mimic TT. However,
EE changes dramatically.
Sep ‘04: CBI, DASI, Capmap
Spectrum Extraction




Two step pipeline.
First step – compress ~few 10^6 visibilities into
10^4 polarization estimators. Keep track of
signal covariances, sources, ground signal etc.
(Myers et al.)
Second – do exact maximum likelihood on 10^4
estimators, get spectrum, window functions,
likelihood surfaces etc.
Analysis done on CITA McKenzie cluster. Stage 1
~couple hours. Stage 2, ~10 minutes for
spectrum, longer for other stuff.
CBI Polarization Power Spectra
Previously published pol’n detection – Science 306,836
NEW DATA ~ 50% increase presented here



7-band fits (Dl =
150 for
600<l<1200)
7-band spectra
consistent with
WMAPext model
(TT from WMAP,
ACBAR, 2000 +
2001 CBI)
Consistent with
old pol’n data,
errors smaller
New Spectra!


Shaped Cl Fit
Use WMAP’03+CBI TT+
Acbar best-fit Cl in signal
covariance matrix
• bandpower relative to
fiducial PS
• compute for single band
encompassing all l
Results for new CBI pol’n
data
• EE qB = 0.97 ± 0.14
(68%)
• EE likelihood vs. zero :
equivalent significance
10.1 σ
• TE qB = 0.85 ± 0.25
• BB qB = 1.2 ± 1.8 μK2

No evidence for
foregrounds
Likelihood of EE Amplitude
vs. TT Prediction
Parameters w/CBI (old)




Paramaters calculated using Antony Lewis’s
MCMC code, COSMOMC
Old CBI mosaics (Readhead et al. 2004)
overlap with polarization mosaics. Not
allowed to combine sample-limited part of
spectra.
Thermal limited (ℓ>1000) old spectrum
included. New spectrum only for ℓ<1000.
First time EE included for measuring
parameters (though impact of EE quite small)
Blue=WMAP
Red=WMAP+Sep ’04 (Working on new data)
Green=WMAP+Sep ‘04+CBI7 high-ℓ
Params, contd…
What Can We Get From EE?
Sensitivity not so high that can get
precision cosmology in many
dimensional space.
 Ask question on restricted parameter
range gives tighter results (3-15%,
depending on questions asked)

θ/θ0




Angular size of sound
horizon at LSS should
be same for TT and EE.
CBI only has multiple
solutions (shift
spectrum by one
peak).
DASI removes
degeneracy, but less
sensitive.
CBIold+DASI give scale
vs. TT of 1.02 +/0.03.
CBI θ/θ0
Peaks can
slide to left
for a degeneracy with
CBI data
only.
DASI θ/θ0
DASI goes to
lower ell removing degeneracy.
DASI θ/θ0
Combination
of CBI+DASI
zeroes in on
proper peak
with few %
accuracy.
pattern shift parameter 0.998 +- 0.005 WMAP1+CBIpol TT/TE/EE
Evolution: Jan00 11% Jan02 1.2% Jan03 0.9% Mar03 0.4%
EE ONLY: 4% phase check of EE cf. TT pk/dip locales, amp too
1.00 +- 0.04 CBIpol+DASIpol EE only
Isocurvature
Isocurvature puts
peaks in different
places from adiabatic. We use
seed isocurvature
model (ask Antony
for details) and find
both EE and TE
prefer adiabatic w/
iso consistent with
zero.
Isocurvature Cont.
Normalize seed iso spectrum to total
power expected from TT adiabatic
prediction
 Fit shapes for both EE, TE
 EE adi =1.00±0.24, iso=0.03±0.20
 TE adi = 0.86±0.26, iso=0.04±0.25

2D Iso/Adi Likelihood - EE

EE only, isocurvature
and adiabatic are rather
strongly
(anti)correlated.
2D Iso/Adi Likelihood - TE

TE only, isocurvature
and adiabatic are somewhat weakly
correlated.
Noisier than
EE.
2D Iso/Adi Likelihood – TT+EE

Combined EE
and TE does
well as TE
valley is tilted.
Effectively
lops of ends
of EE valley.
Summary





Precision cosmology here in total power, coming
along in polarization (CBI EE now > 10σ)
Interferometers are a very clean, stable way of
directly measuring power spectrum
EE polarization consistent with TT prediction both
in shape and amplitude
Can measure distance to last scattering surface
to 4% in pol’n only.
Find most polarization power observed on sky
comes from adiabatic, not (seed) isocurvature.
Non-Gaussianity cont.



Check nonGaussianity in
each bin
Might show ldependent effect
(such as
foreground)
Individual bins
consitent with
Gaussian.
Non-Gaussianity






Decompose data into
uncorrelated S/N
eigenmodes for each bin.
Pick out modes expected to
have signal
Check distribution for nonGaussianity
Keep total of 5500 modes
TT, 3800 EE – everything
consistent with Gaussian
First check of EE
Gaussianity!
More tests coming…
New: CBI EE Polz’n Phase

Parameterization 1: envelope plus
shiftable sinusoid
• fit to WMAP+ACBAR+CBI TT fiducial
spectrum using rational functions
 = 0° : EE prediction
 = 180°: aligned with TT
  1CEE 
a  f    g  sin k   
CBI EE Polarization Phase

Peaks in EE should be offset one-half
cycle vs. TT
• allow amplitude a and phase  to vary
best fit: a=0.94
= 24°±33° (Dc2=1)
Dc2(1, 0°)=0.56
CBI Fine EE w/ Best Fit Phase
CBI Projections

Will BB (lensing) be foreground
limited?
The CBI Collaboration
Caltech Team: Tony Readhead (Principal Investigator), John Cartwright, Clive Dickinson, Alison
Farmer, Russ Keeney, Brian Mason, Steve Miller, Steve Padin (Project Scientist), Tim Pearson,
Walter Schaal, Martin Shepherd, Jonathan Sievers, Pat Udomprasert, John Yamasaki.
Operations in Chile: Pablo Altamirano, Ricardo Bustos, Cristobal Achermann, Tomislav Vucina,
Juan Pablo Jacob, José Cortes, Wilson Araya.
Collaborators: Dick Bond (CITA), Leonardo Bronfman (University of Chile), John Carlstrom
(University of Chicago), Simon Casassus (University of Chile), Carlo Contaldi (CITA), Nils
Halverson (University of California, Berkeley), Bill Holzapfel (University of California, Berkeley),
Marshall Joy (NASA's Marshall Space Flight Center), John Kovac (University of Chicago), Erik
Leitch (University of Chicago), Jorge May (University of Chile), Steven Myers (National Radio
Astronomy Observatory), Angel Otarola (European Southern Observatory), Ue-Li Pen (CITA),
Dmitry Pogosyan (University of Alberta), Simon Prunet (Institut d'Astrophysique de Paris), Clem
Pryke (University of Chicago).
The CBI Project is a collaboration between the California Institute of Technology, the Canadian
Institute for Theoretical Astrophysics, the National Radio Astronomy Observatory, the
University of Chicago, and the Universidad de Chile. The project has been supported by funds
from the National Science Foundation, the California Institute of Technology, Maxine and Ronald
Linde, Cecil and Sally Drinkward, Barbara and Stanley Rawn Jr., the Kavli Institute,and the
Canadian Institute for Advanced Research.
New: DASI EE θ/θ0

Leitch et al. 2004
5 bin EE bandpowers + covariance matrix plus
approximate window functions
New: CBI + DASI EE θ/θ0

Combined constraints on θ model:
CBI a=0.67 overtone island:
suppressed by DASI data
CBI+DASI phase lock:
θ/θ0= 1.02±0.03
a=0.78±0.15 (low DASI)
CBI Projections
2

EE phase: July 2004 vs. 2006
July 2004 (Readhead et al.)
2006 Forecast
Consistency w/ WMAP
Spectra consistent with the
cosmological model from WMAPext
dataset
 χ2 = 7.98 TT, 3.77 EE, 4.33 BB (vs.
0), and 5.80 TE for 7 dof.

CMB Interferometers



CMB issues:
• Extremely low surface brightness fluctuations < 50
mK
 Large monopole signal 3K, dipole 3 mK
 Polarization less than 10%  signal < 5 mK
• No compact features, approximately Gaussian
random field
• Foregrounds both galactic & extragalactic
Traditional direct imaging
• Differential horns or focal plane arrays
Interferometry
• Inherent differencing (fringe pattern), filtered images
• Works in spatial Fourier domain
• Element-based errors vs. baseline-based signals
• Limited by need to correlate pairs of elements
• Sensitivity requires compact arrays