EXPERIMENTAL
COSMIC MICROWAVE
BACKGROUND RESEARCH
Silvia Masi and Paolo de Bernardis
Dipartimento di Fisica
Universita’ La Sapienza
Roma – Italy
Genova – LTD-10 – 10/July/2003
Plan of the talk:
! What is the CMB
! Which are the observables
! Experimental problems:
•
•
•
•
Atmospheric and Instrument Emission
Detector sensitivity
Foregrounds
Systematics
! Thermal Detectors (Coherent detectors)
! Current results for CMB anisotropy
! Measurements of CMB polarization
! Future directions:
• High angular resolution - SZ
• High frequency - SZ
• CMB polarization – B-modes
According to modern
What is the CMB cosmology:
An abundant background of
10−6s
photons filling the Universe.
T >1GeV b + b → 2γ
• Generated in the very early
B(ν)
universe, less than 4 µs after the
Big Bang (109γ for each baryon)
• Thermalized in the primeval
1013s
MW
NIR visible
fireball (in the first 380000
T=3000K
years after the big bang) by
λnow rnow
=
1+ z =
repeated scattering against free
λem rem
electrons
B(ν)
• Redshifted to microwave
1017s
frequencies and diluted in the
subsequent 14 Gyrs of
T=3K
MW
NIR visible
t
expansion of the Universe
3
Today 400γ/cm
Why is the CMB important
• Is the most ancient fossil remnant of the
early Universe.
• Its characteristics tell us a lot about the
physical processes happening in the early
Universe.
• Modern cosmology is heavily based on
observations of the CMB.
Primeval
fireball
CMB photons
today
CMB and cosmology
• 1992: COBE-FIRAS
measures the spectrum of
the CMB with incredible
precision (1/10000)
• The thermal spectrum at
2.735K and the high
photons to barions ratio
together with the measured
primordial abundances of
light elements is evidence
for a hot initial phase of
the Universe.
J. Mather et al. 1992
CMB observables
T=2.735 K
• The spectrum
-15
10
hν
x=
kT
• The angular
distribution
xe x
∆T
∆B(ν , T ) = x
B (ν , T )
e −1
T
• The
polarization
state
-2
average brightness
anisotropy (rms)
polarization (rms)
photon noise (rms)
-17
10
-18
10
-19
10
-2
-1
W m sr Hz
xe
∆TP
B (ν , T )
∆BP (ν , T ) = x
e −1
T
• The noise
10
-21
4 k 4T 4 x 4 e x
ch 3 e x − 1 2
(
-1
-20
10
∆W (ν , T ) =
-1
10
x
2
-1
W m sr Hz
-16
CMB (MKS units)
2h ν 3
B (ν , T ) = 2 x
c e −1
-14
10
)
2
-1/2
W (m sr Hz)
10
10
11
10
Frequency (Hz)
Hz
-1/2
12
10
CMB observables
10
• The CMB is ONLY
slightly anisotropic.
• The brightness
(temperature)
fluctuations are due to
small density
fluctuations present in
the primeval fireball,
and to their motions:
10
-14
T=2.735 K
-15
CMB (MKS units)
-2
10
-16
10
-17
10
-18
10
-19
-1
W m sr Hz
average brightness
anisotropy (rms)
polarization (rms)
photon noise (rms)
-2
∆T 1 ∆ργ 1 ∆ϕ v
=
+
+
2
T
4 ργ
3 c
c
Photon
Density
fluctuations
Gravitational
redshift
Scattering
against
moving e-
-1
-1
W m sr Hz
10
-20
10
-21
2
W (m sr Hz)
10
10
11
10
Frequency (Hz)
-1/2
-1
Hz
-1/2
10
1
CMB observables
∆ T (θ , ϕ ) =
∑a
l ,m
lm
c l = a l2m
∆T
2
1
=
4π
∑ ( 2 l + 1) c
l
l
-15
-2
Y l (θ , ϕ )
m
-14
T=2.735 K
10
CMB (MKS units)
• The rms anisotropy
has contributions
from many angular
scales
• The angular power
spectrum cl of the
anisotropy defines
the contribution to
the rms from the
different multipoles:
10
10
-16
10
-17
10
-18
10
-19
-1
W m sr Hz
-1
average brightness
anisotropy (rms)
polarization (rms)
photon noise (rms)
-2
-1
W m sr Hz
10
-20
10
-21
2
W (m sr Hz)
10
10
11
10
Frequency (Hz)
-1/2
-1
Hz
-1/2
10
12
Which is the power spectrum ?
• The angular power spectrum of the CMB depends
on the physical processes happening in the early
universe, during the primeval fireball phase.
• The primeval fireball is an expanding plasma,
slowly decreasing its temperature, where photons
and matter are in thermal equilibrium
• There are 109 photons for each baryon, so photon
pressure is very important.
• For T > 0.8 eV the energy density of photons
dominates, while at later times the energy density
of matter dominates.
• At T = 0.26 eV the plasma neutralizes, H atoms
are formed, and the universe becomes transparent
to photons (recombination). We see the image of
the CMB as it was there, when photons were last
scattered.
Image of Solar Granulation
Plasma in the
solar photosphere
(5500 K)
Here, now
8 light minutes
Image of Solar Granulation
Plasma in the
solar photosphere
(5500 K)
Here, now
8 light minutes
Plasma in the LSS
the cosmic
photosphere
(3000 K)
Here, now
14 billion light years
The BOOMERanG map of the last scattering surface
• How is the structure we expect to
see in the primeval plasma at
recombination ?
• It depends on
–the physics of the primeval fireball
–the physics of the very early Universe
–the geometry of space
Physics
of the
Primeval
fireball and
very early
universe
Geometry
of
space
Here, now
14 billion light years
The BOOMERanG map of the last scattering surface
Physics of the
primeval fireball
Density perturbations (∆ρ
∆ρ/ρ
∆ρ ρ) were oscillating in the primeval plasma (as a result of the
opposite effects of gravity and photon pressure).
T is reduced enough
that gravity wins again
Due to gravity,
∆ρ/ρ
∆ρ ρ increases,
and so does T
overdensity
t
Pressure of photons
increases, resisting to the
compression, and the
perturbation bounces back
Before recombination T > 3000 K
t
After recombination
T < 3000 K
Here photons are not tightly
coupled to matter, and their
pressure is not effective.
Perturbations can grow and
form Galaxies.
After recombination, density perturbation can grow and create the hierarchy of structures
we see in the nearby Universe.
size of perturbation
(wavelength/2)
v
C
R
v
v
C
LSS
R
v
v
C
300000 ly
v
C
0y
Big-bang
time
300000 y
recombination
2nd dip
2nd peak
multipole
v
450
v
1st dip
1st peak
Power Spectrum
220
Size of sound horizon
The angle subtended depends on the geometry of space
In the primeval plasma, photons/baryons density perturbations start to oscillate only when the sound horizon
becomes larger than their linear size . Small wavelength perturbations oscillate faster than large ones.
Geometry of the
Universe (curvature)
LSS
0.5o
Ω<
Low density Universe Ω<1
horizon
2o
High density Universe
Ω>1
Ω>
horizon
1o
Critical density Universe Ω=1
horizon
14 Gly
PS
PS
0
200
High density Universe
Ω>1
2o
l
PS
0
200
Critical density Universe
Ω=1
l
0
200
Low density Universe
Ω<1
1o
0.5o
l
Physics of the
very early Universe
(Inflation)
Power spectrum
of CMB
temperature
fluctuations
l
Processed by
causal effects like
Acoustic oscillations
Radiation pressure
from photons
resists gravitational
compression
horizon
horizon
(∆T/T)
= (∆ρ/ρ)
∆ρ/ρ) /3
(∆
+ (∆φ
∆φ/c
∆φ 2)/3
– (v/c)•n
Unperturbed
plasma
0
10-36s
Big-Bang Inflation
3 min
Nucleosynthesis
l( l+1) cl
k
Gaussian,
adiabatic
(density)
horizon
P(k)=Akn
Scales
larger than
horizon
Quantum
fluctuations
in the early
Universe
Power
spectrum of
perturbations
INFLATION
Scales
smaller than
horizon
The origin of CMB anisotropies
neutral
300000 yrs
Recombination
t
The angular power spectrum depends on
the cosmological parameters
Dependance on Ω (curvature drives the location of first peak).
Not as simple as in these examples (see S.Weinberg, astro-ph/0006276 )
7,00E-010
Ω=1.55
Ω=1
Ω=0.66
6,00E-010
L(L+1)CL
5,00E-010
4,00E-010
3,00E-010
2,00E-010
1,00E-010
0,00E+000
0
50
100
150
200
250
multipole
300
350
400
450
500
Effect of the baryon density
Dependance on Ωb (Relative amplitudes second to first peak):
All the spectra are normalized to the first peak.
-9
1,0x10
2
Ωmh =0.13
2
Ωbh =0.012
-10
2
8,0x10
Ωbh =0.021
2
Ωbh =0.030
-10
l(l+1)Cl
6,0x10
2
Ωbh =0.045
-10
4,0x10
-10
2,0x10
0,0
0
200
400
600
800
l
1000
1200
1400
CMB observables
• The angular power
spectrum cl of the
anisotropy defines
the contribution to
the rms from the
different multipoles:
∆ T (θ , ϕ ) =
∑
l ,m
a l m Y lm (θ , ϕ )
c l = a l2m
∆T
2
1
=
4π
∑ ( 2 l + 1) c
l
l
• A real experiment will not
be sensitive to all the
multipoles of the CMB.
• The window function wl
defines the sensitivity of the
instrument to different
multipoles.
• The detected signal will be:
1
∆T 2
=
( 2 l + 1) w l c l
∑
meas
4π l
• For example, if the angular
resolution is a gaussian
beam with s.d. σ, the
corresponding window
function is
LP
− l ( l +1)σ 2
wl = e
6000
Expected power spectrum:
∑
l ,m
cl = a
a l m Y lm (θ , ϕ )
4000
3000
2
lm
∆T 2 =
2
∆ T (θ , ϕ ) =
l(l+1)cl/2π (µK )
5000
2000
1
4π
∑ ( 2 l + 1) c
l
1000
l
0
0
200
400
600
800
1000 1200 1400
multipole l
1.0
20' FWHM
10' FWHM
5' FWHM
0.8
o
7 FWHM
0.6
wl
An instrument
with finite angular
resolution is not sensitive
to the smallest scales
(highest multipoles). For a
gaussian beam with s.d. σ:
0.4
0.2
w
LP
l
=e
− l ( l +1)σ
2
0.0
0
200
400
600 800 1000 1200 1400
multipole
6000
Expected power spectrum:
2
l(l+1)cl/2π (µK )
5000
∑a
l ,m
lm
Y l (θ , ϕ )
c l = a l2m
2000
1000
∑ ( 2 l + 1) c
l
0
l
w
=e
LP
(
2
l
+
1
)
w
∑
l cl
l
− l ( l +1)σ
2
120
100
∆Trms (µK)
LP
l
200
400
600
800
1000 1200 1400
multipole l
rms signal in an instrument
with gaussian beam σ :
1
∆T 2
=
meas
4π
0
∆T
T
= 4.2 × 10 −5
rms
80
60
40
20
0
1000
100
BOOMERanG
1
=
4π
3000
WMAP
∆T
2
4000
COBE
∆ T (θ , ϕ ) =
m
10
FWHM - gaussian beam (arcmin)
CMB observables
∆ T (θ , ϕ ) =
∑a
l ,m
lm
c l = a l2m
∆T
2
1
=
4π
∑ ( 2 l + 1) c
l
l
-15
-2
Y l (θ , ϕ )
m
-14
T=2.735 K
10
CMB (MKS units)
• The rms anisotropy
has contributions
from many angular
scales
• The angular power
spectrum cl of the
anisotropy defines
the contribution to
the rms from the
different multipoles:
10
10
-16
10
-17
10
-18
10
-19
-1
W m sr Hz
-1
average brightness
anisotropy (rms)
polarization (rms)
photon noise (rms)
-2
-1
W m sr Hz
10
-20
10
-21
2
W (m sr Hz)
10
10
11
10
Frequency (Hz)
-1/2
-1
Hz
-1/2
10
12
The polarization
of the CMB
• CMB photons are last scattered by electrons at
recombination
• It’s a Thomson scattering
• The cross-section depends on the scattering angle
• The scattered
radiation can get some
degree of linear
polarization in the
scattering, even if the
incoming radiation is
not polarized.
• In terms of the Stokes
parameters:
r
E (t ) = E x cos(ωt − θ x ) xˆ + E y cos(ωt − θ y ) yˆ
I =

Q =

U =
V =

E x2 + E y2
E x2 − E y2
2 E x E y cos(θ x − θ y ) = E x2' − E y2'
2 E x E y sin(θ x − θ y )
• Q’ is non-zero, i.e. the scattered radiation is polarized, if the
incoming radiation I(θ,φ) has a Quadrupole distribution
• If the local distribution of incoming radiation in the rest
frame of the electron has a quadrupole moment, the
scattered radiation acquires some degree of linear
polarization.
Last scatte
rin
g surface
• Before recombination there is no quadrupole component in the
radiation: it arrives at recombination unpolarized.
• During recombination, Gradients in the velocity field can
produce a quadrupole in the framework of the scattering electron.
converging flow
same flow in e- rest frame
diverging flow
same flow in e- rest frame
• Before recombination there is no quadrupole component in the
radiation: it arrives at recombination unpolarized.
• During recombination, Gradients in the velocity field can
produce a quadrupole in the framework of the scattering electron.
converging flow
radiation in e- rest frame
redshift
Quadrupole!
blueshift
diverging flow
radiation in e- rest frame
blueshift
Quadrupole!
redshift
• Before recombination there is no quadrupole component in the
radiation: it arrives at recombination unpolarized.
• During recombination, Gradients in the velocity field can
produce a quadrupole in the framework of the scattering electron.
converging flow
radiation in e- rest frame
Radial
polarization
diverging flow
Quadrupole
radiation in e- rest frame
Tangential
polarization
Quadrupole
•
This component of the CMB polarization field is called E
component, or gradient component. This is the only kind of
polarization produced at recombination by scalar perturbations.
• It is related to velocity fields. For acoustic oscillations, it will be
maximum for perturbations with maximum velocity and zero
density contrast.
• So we expect peaks in this polarization power spectrum where we
have minima in the temperature power spectrum.
• The amplitude of the polarization signal depends on the length of
the recombination process (it is not produced before, nor later).
• Tensor perturbations (gravity waves) also produce quadrupole
anisotropy. The generation of a faint stochastic background of
gravity waves is a generic feature of all inflationary processes.
• The resulting polarization pattern is shear-like.
• The amplitude of the effect is very small.
• This component of the CMB polarization field is called B
component, or curl component.
• Velocity fields cannot produce B modes.
• Weak lensing can, but is subdominant at scales larger than 1 deg.
• Mathematical alghoritms exist to separate B modes and E modes.
• The description in terms of Q and U is not invariant under
rotation of the coordinate system:
• The description in terms of E and B is rotationally invariant.
• Velocity fields (scalar perturbations) produce E modes only.
• Inflation (tensor perturabtions) produces both E-modes and
B-modes
• Four independent power spectra can be measured (the other
combinations are 0 by symmetry):
c ,c ,c ,c
TT
l
TE
l
EE
l
BB
l
Power spectra of anisotropy and polarization.
The polarization signal is faint; the B modes are very faint.
[l(l+1)Cl/2π]
1/2
(µK)
10
1
τ=0
E-like
scalar T
scalar E
tensor T
tensor E
tensor B
E
B-like
0.1
B
0.01
10
100
10o
multipole
1000
1o
Model parameters: ΩΛ=0.7, Ωb=.05, ΩΜ=0.25, ns=0.8, nT=0.2, T/S=1.4
Expected Patterns of Polarization in the Sky
From the BICEP website (Caltech)
• Finally, CMB is partially rescattered when the
formazion of the first stars re-ionizes the universe
(z>6).
• Since these additional scatterings happen much closer
to us, they enhance polarization spectra only at large
angular scales (l of the order of 20, corresponding to the
size of the horizon at the epoch of reionization)
Power spectra of anisotropy and polarization.
The polarization signal is very small, expecially at large angular scales.
[l(l+1)Cl/2π]
1/2
(µK)
10
1
τ=0
E-like
scalar T
scalar E
tensor T
tensor E
tensor B
E
B-like
0.1
B
0.01
10
100
10o
multipole
1000
1o
Model parameters: ΩΛ=0.7, Ωb=.05, ΩΜ=0.25, ns=0.8, nT=0.2, T/S=1.4
[l(l+1)Cl/2π]
1/2
(µK)
Reionization enhances polarization at large angular scales:
CMB photons are re-scattered much closer to us
10
τ = 0.2
scalar T
scalar E
tensor T
tensor E
tensor B
1
E
0.1
B
0.01
10
100
10o
multipole
1000
1o
Model parameters: ΩΛ=0.7, Ωb=.05, ΩΜ=0.25, ns=0.8, nT=0.2, T/S=1.4
Why do we care about the
polarization of the CMB ?
• We can give an independent confirmation of
the model
• We can detect isocurvature fluctuations mixed
to the dominant adiabatic ones.
• We can test the velocity field present at
recombination
• We can detect the reionization happening
when the first structures form
• We can detect the signature of inflation in the
B-modes pattern of polarization
CMB observables
∆T
xe x
∆BP (ν , T ) = x
B (ν , T ) P
e −1
T
∆TP
≈ 4 ×10 −6
T rmsE
∆TP
T
≈ 2 × 10 −7
rmsB
• Extremely weak !
CMB (MKS units)
• The polarization
state
10
-14
10
-15
10
-16
10
-17
10
-18
10
-19
10
-20
10
-21
10
-22
10
-23
T=2.735 K
-2
-1
W m sr Hz
-1
average brightness
anisotropy (rms)
polarization (E rms)
photon noise (rms)
polarization (B rms)
-2
-1
W m sr Hz
2
W (m sr Hz)
-2
-1
-1/2
Hz
W m sr Hz
10
10
11
10
Frequency (Hz)
-1
-1/2
-1
10
12
Experimental
Approach
• These signals
are faint with
respect to:
– detector noise
– background
emission of the
instrument and
of the earth
atmosphere
– background
emission of the
astrophysical
environment.
CMB (MKS units)
CMB observables
10
-14
10
-15
10
-16
10
-17
10
-18
10
-19
10
-20
10
-21
10
-22
10
-23
T=2.735 K
-2
-1
W m sr Hz
-1
average brightness
anisotropy (rms)
polarization (E rms)
photon noise (rms)
polarization (B rms)
-2
-1
W m sr Hz
2
W (m sr Hz)
-2
-1
-1/2
Hz
W m sr Hz
10
10
11
10
Frequency (Hz)
-1
-1/2
-1
10
12
CMB observables
-14
T=2.735 K
10
-15
-2
CMB (MKS units)
• The spectrum
peaks at 150 GHz
• The anisotropy,
polarization and
noise peak at 210220 GHz
• These frequencies
are high for
coherent
detectors, and low
for thermal
detectors.
10
10
-16
10
-17
10
-18
10
-19
-1
W m sr Hz
-1
average brightness
anisotropy (rms)
polarization (rms)
photon noise (rms)
-2
-1
W m sr Hz
10
-20
10
-21
2
W (m sr Hz)
10
10
11
10
Frequency (Hz)
-1/2
-1
Hz
-1/2
10
12
Detectors
Detectors
• Coerent detectors measure amplitude and
phase of the em wave
• Thermal detectors measure the energy of
the em wave
• On both sides, CMB research drove the
development of new devices:
– Cryogenic, ultra-low noise HEMT amplifiers
(coherent)
– Cryogenic “Spider Web” and “Polarization
Sensitive” Bolometers (thermal)
– Low sidelobe corrugated antennas …..
• Also, the two worlds are progressively
mixed: for example waveguides and
striplines are now used with cryogenic
bolometers
Cryogenic Bolometers
• The CMB spectrum is continuum and bolometers are wide band
detectors. That’s why they are so sensitive.
Load resistor
Thermometer
(Ge thermistor (∆R)
at low T)
Incoming
Photons (∆B)
∆V
Integrating
cavity
Feed
Horn
(angle selective)
filter
(frequency
Radiation
selective)
Absorber (∆T)
• Fundamental noise sources are Johnson noise in the thermistor
(<∆V2> = 4kTR), temperature fluctuations in the thermistor
((<∆W2> = 4kGT2), background radiation noise (Tbkg5)
need to
reduce the temperature of the detector and the radiative
background.
Cryogenic Bolometers
• In steady conditions the
temperature rise of the sensor is
due to the background radiative
power absorbed Q and to the
electrical bias power P:
G (T − T0 ) = Q + P
• The effect of the background power
is thus equivalent to an increase of
T0’
the reference temperature:
Q 

P = G T − (T0 + ) = G (T − T0 ' )
G 

Q
T0 ' = T0 +
G
0.28K
0.27K
Q(pW)
0.26K
0
1
2
Cryogenic Bolometers
1
dT
• In presence of an additional signal
=
dQ Geff 1+τ 2ω 2
∆Q ejωt (from the sky)
d∆T
C
+ Geff ∆T = ∆Q
dt
C
τ=
G
• There is a tradeoff between high
sensitivity and fast response. The
heat capacity C should be
minimized to optimize both.
• Using a current biased thermistor to
readout the temperature change:
Responsivity
Small sensor
at low
temperature
1 dR (T )
α=
⇒ dV = idR = iα RdT
R (T ) dT
dV
dT
iα R
ℜ=
= iα R
=
dQ
dQ G eff 1 + τ 2ω 2
Cryogenic Bolometers
1 dR (T )
α=
R (T ) dT
iα R
ℜ=
G eff 1 + τ 2ω 2
• A large α is
important for
high
responsivity.
−1
• Ge thermistors: α ≈ 10 K
• Superconducting
transition
−1
thermistors: α ≈ 1000 K
S.F. Lee et al. Appl.Opt. 37 3391 (1998)
Cryogenic Bolometers
Again, need
of low
Temperature
And low
Background
• Johnson noise in the thermistor
d ∆ V J2
= 4 kTR
df
• Temperature noise
d ∆ W T2
4 kT 2 G eff
= 2
2
df
G eff + (2π fC )
• Photon noise
5
d ∆WPh2
4k 5TBG
x4 (ex −1+ ε )
dx
= 2 3 ∫ε
2
x
df
ch
(e −1)
• Total NEP (fundamental):
2
2
2
d
V
d
W
d
W
∆
∆
∆
1
J
T
Ph
2
NEP = 2
+
+
df
df
ℜ df
Q
Development of thermal detectors for far IR and mm-waves
17
10
Langley's bolometer
time required to make
a measurement (seconds)
Golay Cell
12
Golay Cell
10
Boyle and Rodgers bolometer
1year
7
F.J.Low's cryogenic bolometer
10
Composite bolometer
1day
Composite bolometer at 0.3K
1 hour
2
10
1 second
Spider web bolometer at 0.3K
Spider web bolometer at 0.1K
Photon noise limit for the CMB
1900
1920
1940
1960
1980
year
2000
2020
2040
2060
•The absorber is micro
machined as a web of
metallized Si3N4 wires, 2
µm thick, with 0.1 mm
pitch.
•This is a good absorber for
mm-wave photons and
features a very low cross
section for cosmic rays.
Also, the heat capacity is
reduced by a large factor
with respect to the solid
absorber.
Spider-Web Bolometers
Built by JPL
Signal wire
Absorber
•NEP ~ 2 10-17 W/Hz0.5 is
achieved @0.3K
•150µKCMB in 1 s
•Mauskopf et al. Appl.Opt.
36, 765-771, (1997)
Thermistor
2 mm
Crill et al., 2003 – BOOMERanG 1998 bolometers
Cryogenic Bolometers
• Ge thermistor bolometers have
been used in many CMB
experiments:
– COBE-FIRAS, ARGO, MAX,
BOOMERanG, MAXIMA, ARCHEOPS
• Ge thermistor bolometers are
extremely sensitive, but slow:
the typical time constant C/G is
of the order of 10 ms @ 300mK
• Transition Edge Superconductor
(TES) thermistors can do much
better using electro-thermal
feedback (100 µs) – Recent
development (hear Adrian Lee,
next talk)
Bolometer Arrays
• Once bolometers reach BLIP
conditions (CMB BLIP), the
mapping speed can only be
increased by creating large
bolometer arrays.
• BOLOCAM and MAMBO are
examples of large arrays
with hybrid components (Si
Bolocam Wafer (CSO)
wafer + Ge sensors)
• Techniques to build fully
litographed arrays for the
CMB are being developed.
• TES offer the natural
sensors. Hear A. Lee, D.
MAMBO (MPIfR for IRAM)
Benford, A. Golding ..
Mapping speed
• Mapping speed will be enormously increased by
the use of arrays of bolometers.
• These are being developed in several labs
• See e.g.
–
–
–
–
–
–
–
–
Holland et al. MNRAS 303, 659, (1999)
Kreysa et al. SPIE 3357, 319, (1998)
Glenn et al. SPIE 3357, 326, (1998)
Turner et al., Appl. Opt., 40, 4291, (2001)
Griffith et al., ESA-SP460, 37, (2001)
Lamarre et al., Astroph. Lett. & Comm, 37, 161, (2000)
Dowell et al., proc. AAS 198, 05.09 (2001)
And the papers presented here.
Coherent Detectors
• Very low noise HEMT amplifiers,
cooled at 20K have been
developed (NRAO).
• They have been used in many CMB
experiments: TOCO, DASI, CBI,
WMAP and are the baseline for
Planck-LFI.
• Hear the talk from L.Terenzi
Atmosphere
• The spectrum
peaks at 150 GHz
• The earth
atmosphere is
emissive (and not
very transparent)
in the same range.
• Sensitive
observations must
be carried out
above the earth
atmosphere:
space carriers
are required.
CMB (MKS units)
CMB observables
10
-14
10
-15
10
-16
10
-17
10
-18
10
-19
10
-20
10
-21
10
-22
10
-23
T=2.735 K
-2
-1
W m sr Hz
-1
average brightness
anisotropy (rms)
polarization (E rms)
photon noise (rms)
polarization (B rms)
-2
-1
W m sr Hz
2
W (m sr Hz)
-2
-1
-1/2
Hz
W m sr Hz
10
10
11
10
Frequency (Hz)
-1
-1/2
-1
10
12
h=41 km, z=45 deg
CMB
CMB anisotropy (rms)
250K BB
250K BB , ε=0.1
-12
10
-13
10
2
Brightness (W / m sr Hz)
• The average emission of
the atmosphere, of the
instrument, and of the
CMB, can be rejected
using modulation
techniques.
• While average brightness
measurements of the CMB
definitely require spaceborne measurements,
anisotropy measurements
can be carried out at
lower altitudes, in selected
atmospheric windows.
• Also, atmospheric
emission is basically
unpolarized.
-14
250K BB , ε=0.01
10
-15
10
-16
10
-17
10
-18
10
-19
10
-20
10
-21
10
10
10
11
10
Frequency (Hz)
12
10
How to do this
• Atmospheric transmission at different altitudes
Atmospheric emission and noise
• Even an in-band
transmission of
95% results in a
15K background,
loading the
detectors, and
with significant
fluctuations,
increasing
detector noise
• Going to 40Km of
altitude,
atmospheric
emissivity drops
significantly.
NEP (10-17 W/Hz0.5)
1000
• Assuming ideal
4.2 km
bolometers,
14 km
matched to the
41 km
atmospheric
100
background, the
noise equivalent
power degrades
10
with
atmospheric
background for
1
two reasons:
optical loading
500
1000 1500 2000 2500
wavelength (µm)
and photon
noise. BALLOON BORNE OBSERVATIONS CAN
BE 4-10 TIMES MORE SENSITIVE THAN
AIRBORNE ONES (assuming same integration
time, λ/∆λ=10, only photon noise, no turbulence)
Modulation
Techniques
Scanning telescopes
• The beam scans the sky at constant speed v (o/s)
• Different multipoles in the CMB temperature field produce
different sub-audio frequencies in the detector (see e.g. astroph/9710349)
Γ(f)
l(l +1) cl
f = v l/π
π
l
• Examples:
– v=1 o/s , l=200
->
f (Hz)
f=1.1Hz
– v=1 o/s , l=1000 -> f =5.5Hz
• This technique allows to produce wide sky maps, so that a wide
multipoles coverage of the power spectrum can be obtained in a
single experiment.
10000
1/f noise
white noise
2
system noise spectral density (µK /Hz)
bolometer + electronics transfer function
1000
2
CMB 1-D power spectrum x beam: Γm (µK )
pendulations
100
first acoustic peak
10
second acoustic peak
1
beam 20' FWHM
τbol=50 ms
0,1
o
azimuth scan @ 3 /s
o
z=50
0,01
0
1
2
3
f (Hz)
4
5
6000
Expected power spectrum:
∑
l ,m
3000
2
lm
∆T 2 =
2000
1
4π
∑ ( 2 l + 1) c
l
1000
l
1/f noise is removed by
cutting low frequencies
(f<fc). This is equivalent to
removing the lowest
multipoles.
w lHP
4000
− 1  2π f c 
= 1 − sin 

π
 lv 
(E. Hivon)
2
0
0
200
400
600
800
1000 1200 1400
multipole l
1.0
fc = 60 mHz
0.8
o
v=0.7 /s
wl high pass
cl = a
a l m Y lm (θ , ϕ )
2
∆ T (θ , ϕ ) =
l(l+1)cl/2π (µK )
5000
0.6
0.4
0.2
0.0
0
200
400
600
800
multipole
1000 1200 1400
The sky scan
• The image of the sky is obtained by
slowly scanning in azimuth (+30o) at
constant elevation
• The optimal scan speed is between 1
and 2 deg/s in azimuth
crosslink in BOOMERanG LDB scans (1 scan/hour sho
0-11h
-35
• The scan center
constantly tracks the
azimuth of the lowest
foreground region
• Every day we obtain a
fully crosslinked map.
declination (degrees)
12-23h
-40
-45
-50
-55
elev. = 45
3
4
5
Right Ascension (hours)
6
o
TOP-HAT
Antarctica Jan.2001
NASA-GSFC
http://topweb.gsfc.nasa.gov
Map of 6% of the sky
Spinning during a polar night flight, ARCHEOPS has covered
25% of the sky, and mapped the CMB over 13% of the sky
A. Benoit, et al. A&A 2003
Foregrounds
From the WMAP web site.
In reality at 150 GHz the dust anisotropy
is << of the CMB anisotropy in most of the sky
Interstellar Foreground
Map of mm-wave emission of dust in our galaxy
as derived from IRAS and DIRBE measurements
(Schlegel et al 1999)
Northern
Hemisphere
270o
270o
180o
0o
90o
180o
Log scale
Minimum
Brightness (0.33 MJy/sr)
Southern
Hemisphere
90o
Maximum
Brightness (30 MJy/sr)
Interstellar Foreground
Map of mm-wave emission of dust in our galaxy
as derived from IRAS and DIRBE measurements
(Schlegel et al 1999)
Northern
Hemisphere
270o
270o
180o
0o
90o
180o
Log scale
Minimum
Brightness (0.33 MJy/sr)
Southern
Hemisphere
90o
Maximum
Brightness (30 MJy/sr)
Current
results
CMB and cosmology
• 2000:
BOOMERanG
and MAXIMA
map the
temperature
fluctuations of
the CMB at
sub-horizon
scales (<1O).
• The signal is
well above the
noise and has
the correct
frequency
spectrum.
BOOMERanG 150 GHz
CMB and cosmology
• 2000:
BOOMERanG
and MAXIMA
map the
temperature
fluctuations of
the CMB at
sub-horizon
scales (<1O).
• The signal is
well above the
noise and has
the correct
frequency
spectrum.
BOOMERanG 150 GHz
CMB and cosmology
• 2000:
BOOMERanG
and MAXIMA
map the
temperature
fluctuations of
the CMB at
sub-horizon
scales (<1O).
• The signal is
well above the
noise and has
the correct
frequency
spectrum.
CMB and cosmology
COBE
6000
Archeops
BOOMERanG
CBI
DASI
5000
l(l+1)cl/2π (µK)
• The power
spectrum of the
CMB anisotropy
features a
distinctive peak
at multipole 210
and overtones at
multipoles 540
and 830.
• There are
acoustic
oscillations in the
primeval plasma
4000
3000
2000
1000
0
1
10
200
400
600
800
1000
1200
1400
multipole
June 2002
0.90
7000
dimensioni delle strutture (gradi)
0.30
0.225
0.45
0.18
6000
intensita'
5000
4000
3000
2000
1000
0
0
200
400
600
multipolo
800
1000
CMB and cosmology
• The peak at multipole 210 means that
typical size of the anisotropies is 1o,
which means that the geometry of the
universe is flat (Ω=1, as predicted by
inflation)
14 billion light years
ct
1o
300000
Light years
Ω=1
ct
2o
Ω>1
Ω>
0.5
o
ct
Ω<1
Ω<
CMB and cosmology
• The amplitudes of the first, second and third
peaks allows to estimate Ωb=0.02, in
agreement with BBN.
• The slope of the power spectrum of the CMB
anisotropy agrees with the expectations of the
basic inflationary model (n=1).
P. de Bernardis et al., Nature, 404, 955-959, 2000
S. Hanany et al., Ap.J., 545, L5-L9, 2000
A. Lange et al., PRD 63, 042001, 2001
R. Stompor et al., Ap.J., 561, L7, 2001
A. Lee et al., Ap.J., 561, L1, 2001,
B. Netterfield et al. Ap.J. 571, 604, 2002
P. de Bernardis et al. Ap.J. 564, 559, 2002
N. Halverson et al., Ap.J., Astro-ph/0104488-89-90, 2002
A. Benoit et al., A&A , 399, L19 and L25, 2003
J. Ruhl et al. Ap.J submitted astro-ph/0212229, 2003
• It is shown that Galactic CMB
contamination at 150 GHz is
less than 1% of the CMB
fluctuations PS at multipole
200. This confirms that
precision cosmological
observations will be possible
from satellite experiments in a
wide area of the sky.
S. Masi et al. Ap.J. 553, L93, 2001
• The image of the CMB is
shown to be accurately
gaussian as predicted by
inflation. G. Polenta et al. Ap.J. 572,
L27, 2002; G. De Troia et al. MNRAS
in press
and cosmology
2002:CBI
CBI: Cosmic Background Imager
• Same technology as DASI (larger
telescopes)
• Operation from Atacama desert
(5000 m o.s.l.)
• l-space resolution still coarse
• New data :
astro-ph/0205384,5,6,78
• Very good consistency with other
experiments at l < 1000.
• New data up to l= 3500.
• Detected fluctuations with the
same mass of clusters of galaxies.
• The damping tail is evident !
• Excess at l=2500.
• Only one frequency (30 GHz).
• 2002: The Degree Angular
Scale Interferometer (PI J.
Carlstrom)) at the south pole
has recently detected for the
first time the linear
polarization of the CMB
• astro-ph/0209478
• astro-ph/0209476
• at a level consistent with
the concordance model.
DASI
• 2003: First results from
WMAP, the CMB
anisotropy mission of
NASA, working from
L2.
• A beautiful, firm
confirmation of all we
know about the
CMB…
• …. and more:
• Optical depth of
reionization (from
polarization)
• Anisotropy low at
large scales
NASA-2001
• 2003: First results from
WMAP, the CMB
anisotropy mission of
NASA, working from L2.
• The TT power spectrum,
limited by cosmic
variance up to l=350
• The power spectrum of
TE (correlation between
anisotropy and
polarization) in
agreement with the
acoustic oscillations
scenario, and featuring an
excess at low l.
WMAP: 94 GHz
BOOM/98: 150 GHz
WMAP & BOOM/98: Power Spectra
Cosmological Parameters
Compare with same weak prior on 0.5<h<0.9
WMAP
BOOMERanG
(100% of the sky)
Bennett et al. 2003
(4% of the sky)
astro-ph/0212229
• Ω =1.02+0.02
• Ω = 1.03+0.05
•
•
•
•
•
•
•
•
•
ns = 0.99+0.04 *
Ωbh2 =0.022+0.001
Ωmh2 =0.14+0.02
T = 13.7+0.2 Gyr
τrec= 0.166+0.076
ns = 1.02+0.07
Ωbh2 =0.023+0.003
Ωmh2 =0.14+0.04
T=14.5+1.5 Gyr
•
The Future :
"Polarization of the CMB
"High Resolution
"High Frequency
CMB
polarization
measurements
Linear Polarimeter
source
polarizer
θ
Intensity detector
• A polarimeter is a device able to detect polarized light
and measure its polarization characteristics.
• The simplest polarimeter we can imagine is a linear
polarimeter, which can be built with a rotating
polarizer in front of an intensity detector.
• An intensity detector is represented by a Stokes vector
D=(1,0,0,0). The power detected by the detector from
an optical beam with Stokes vector S is simply
w=DS=So (here S=(I,Q,U,V))
• If we put a polarizer in front of the detector, the
polarizer is called analyzer, and the power detected
will be w(θ) =DMP(θ)S
Polarizer or Diattenuator
• It attenuates the orthogonal
components of an optical beam
unequally:
• Using the definitions of S and S’
 I '   E x' E x'* + E y' E y'* 
 '   ' '*

'
'*
 Q   Ex Ex − E y E y 
U '  =  ' '*
'
'* 
+
E
E
E
E
y x 
   x y
 V '   i ( E ' E '* − E ' E '* ) 
y x 
   x y
 E x' = p x E x
 '
 E y = p y E y
*
*
 I   Ex Ex + E y E y 

  
*
*
 Q   Ex Ex − E y E y 
U  =  E E * + E E * 
y x 
   x y*
 V   i( E E − E E * ) 
   x y
y x 
• And inserting the expressions for E’
we get
 I' 
 p x2 + p y2 p x2 − p y2
 '
 2
 Q  1  p x − p y2
 '= 
U  2  0
 0
V ' 

 
p x2 + p y2
0
0
0
2 px p y
0
0
 I 
 
 Q 
 
U 
2 p x p y  V 
0
0
0
Rotated Polarizer
 px2 + py2
 2 2
1  px − py
MC (0) = M P (0) = 
2 0
 0

• so
• and
1

1 0
M P (θ ) = 
2 0

0

0
0 
px2 − p2y
Σ


2
2
0
0  1 ∆
px + py
= 

def
0
2 px py
0  2 0

0

0
0
2 px py 

0 0
c2 − s2
s2 c2
0 0
0 Σ

0 ∆
0 0

1  0
∆
Σ
0
0
0
0
X
0
∆
Σ
0
0
0
0
X
0
0  1 0 0

0  0 c2 s2
0  0 − s2 c2

X  0 0 0
0

0
0

1 
0
c2 ∆
s2 ∆
 Σ


2
2
1  c2 ∆ c2 Σ + s2 X s2c2 (Σ − X ) 0 
M P (θ ) = 
2 s2 ∆ s2c2 (Σ − X ) s22Σ + c22 X 0 


 0
0
0
X 

0

0
0

X 
 Σ = p x2 + p y2

2
2
p
p
∆
=
−
x
y

 X = 2 px py
 s = sin 2 θ
 2
 c 2 = cos 2 θ
Linear Polarimeter
source
Polarizer (analyzer)
θ
Intensity detector
c2 ∆
s2 ∆
 Σ

2
2
1  c2 ∆ c2 Σ + s 2 X s 2 c2 ( Σ − X )
w = DM P (θ ) S = (1,0,0,0) 
2 s2 ∆ s2c2 (Σ − X ) s22 Σ + c22 X

 0
0
0

w =
1
2
(Σ I
+ Q ∆ cos 2 θ + U ∆ sin 2 θ
0  I 
 
0  Q 
⇒



0 U
 
X  V 
)
This polarimeter is not sensitive to circular polarization (no V).
It is sensitive to linear polarization (Q and U) and to
unpolarized light (I).
If the polarizer is ideal: ∆ = 1 ; Σ = 1 ; X = 0
w=
1
2
(I
+ Q cos 2θ + U sin 2θ
)
Linear Polarimeter
• If we are interested to the linear polarized component
only, we can rotate continuously the polarizer: θ=ωt
and look only for the AC signal at frequency 2ω.
• This allows to reject the unpolarized component, even if
it is dominant, and to remove all the noise components
at frequencies different than 2ω (synchronous
demodulation).
source
ω
Rotating analyzer
Intensity detector
(Σ I + Q ∆ cos 2ω t + U ∆ sin 2ω t )
V (t ) = Rw (t ) + N (t ) = 12 R [Σ I + ∆ (Q cos 2ω t + U sin 2ω t )] + N (t )
w=
1
2
detector
responsivity
constant
signal (DC)
modulated
signal (AC)
noise
(AC)
Linear Polarimeter
source
Log P(ω)
<…>T
ω
Detector
Rotating
analyzer
Rw+N
Ref(2ω)
C
A
x
A(Rw+N)AC
A[Rw(2ω)+N(∆ω)]
Demodulated signal
noise
signal
σ2 = ∫ P(
ω)dω
∆ω
∆ω-=1/T
1/RC
2ω
Log ω
R
How do we separate Q and U
V ( t ) = Rw ( t ) + N ( t ) =
1
2
R [Σ I + ∆ (Q cos 2ω t + U sin 2ω t )] + N ( t )
• Neglecting the stochastic effect of noise (we
integrate enough that N becomes negligible) and
of the constant term (which we remove with the
AC decoupling)
V (t ) = Rw ( t ) = 12 R [∆ (Q cos 2ω t + U sin 2ω t )]
• We measure V and we want to estimate Q and U.
We can use two reference signals, out of phase
by T/8 and synchronously demodulate with them:
How do we separate S1 and S2
T
T


1
1 R∆
X = T ∫ V (t ) sin 2ωtdt = 2 T Q∫ cos 2ωt sin 2ωtdt + U ∫ sin 2ωt sin 2ωtdt
0
0

 0
T
T
T


1
1 R∆
Y = T ∫ V (t ) cos 2ωtdt = 2 T Q∫ cos 2ωt cos 2ωtdt + U ∫ sin 2ωt cos 2ωtdt
0
0
 0

T
[ R ∆ ]U
= [18 R ∆ ] Q
X =
Y
1
8
• So the double linear polarimeter is
insensitive to I and it is easy to
calibrate.
• Is this a troubleless instrument ? No !
• It is inefficient (factor 1/8 from
modulation and demodulation)
• It can be microphonic.
Modulation techniques
for polarization experiments
All these techniques suffer for the need of long integration time:
One needs to point to the same sky pixel during many cycles of
the analyzer. It is thus difficult to produce extended maps of the
CMB polarization.
A two-bolometers polarimeter ?
We can map the two orthogonal components of the linear
polarization with two separate bolometers, and combine the two
signals to retreive the Stokes parameters Q and U.
I1
I2
Q = I1(t1) – I2(t2)
scan
The modulation is obtained by scanning the sky at constant rate,
so that the polarization signals are detected at a frequency far
from the 1/f knee of the noise and far from the effect of
instrumental drifts (similar to the anisotropy measurements with
B98)
f = v l/π :
with l=300-1000 and v=1o/s
f = 1.7-5.5 Hz
Terminology
• Co-polar response of the polarimeter C:
response of the polarimeter to incoming
radiation 100% polarized along the
principal axis of the polarimeter. It is the
product of the detector responsivity times
the integral of the Co-polar beam response.
Units: V/W or V/K
• Cross-polar response of the polarimeter X:
response of the polarimeter to incoming
radiation 100% polarized and orthogonal to
the principal axis of the polarimeter. It is
the product of the detector responsivity
times the integral of the cross-polar beam
response. Units: V/W or V/K
• Polarization efficiency = 1 – X/C
Polarimeter
Incoming
diffuse
radiation
Principal
axis
I
S=CI={(Σ+∆
Σ+∆)/2}I
Σ+∆
Principal
axis
S=XI ={(Σ
Σ−∆)/2}I
I
Frequency content of detected signal
• The detected signal is the sum of several contributions.
• First order approximation, for bolometer #1:
S1 = C1 x ( TCMB + O + D + ∆B/2 + ∆TU /2 + ∆P1) + N1+ X1x (.. +∆P2)
Co-polar
response
CMB &
Offset
0 Hz
Instrument
Drifts
0-0.01 Hz
These dominant
components can be
removed using a
high-pass filter
∆BKG
In the
f range
of
interest
unpola
rized
CMB
Anisotropy,
Unpolarized
Component
Along dir. 1
0.5-5.5 Hz
CMB
Polarized
Component
Along dir.1
1.7-5.5 Hz
Crosspolar
response
Noise
0-20Hz
These components are detected
PSB: Polarization Sensitive Bolometers
(JPL+Caltech)
• 150 GHz
• Two wire-grid-like
absorbers with
matched NTD
thermistors
• Rotated 90°
• Very close each
other (60 µm)
inside the same
groove of a
corrugated circular
feedhorn
Metalized
Si3N4 wires
B. Jones
A. Lange
T sensor
Metalized
Si3N4 wires
B. Jones
A. Lange
T sensor
Minimize Xpol
by concentrating
the E field in the
center
HFSS simulations
ε ~ 5.5%
ε ~ 1.3%
B.Jones et al. Astro-ph/0209132
Polarization-sensitive bolometers
JPL-Caltech
3 µm thick
wire grids,
Separated by
60 µm, in the
same groove
of a circular
corrugated
waveguide
Planck-HFI
testbed
B.Jones et al. Astro-ph/0209132
Polarization Sensitive Bolometers
PSB Pair
Corrugated cilindrical feedhorns
The Back to Back input feeds
2-Color Photometer
elevation
30’
30’
30’
azimuth
30’
Receiver Specifications
Polarization Efficiency
S(θ) = γ(1 – β sin2(θ-θ0))
β is the polarization efficiency
PSB’s have an efficiency of 90-95%
Photometer channels have an efficiency > 97%
PSB pair
Two 245 GHz detectors 45° apart
Measuring Polarization with BOOMERANG
Stokes Parameters
I = Ex2 + Ey2
Q = Ex2 - Ey2
U = 2ExEy
Each Detector is sensitive to one linear polarization.
A pair of orthogonal detectors measures:
Sx = γx{ (1 – βx/2) I + (βx/2) Q}
Sy = γy{ (1 – βy/2) I - (βy/2) Q}
If γx = γy and βx = βy, we have
I=
Sx + Sy
2 γ (1 – β/2)
Q = Sx - Sy
γβ
Instrument Calibration
• Absolute Calibration with respect to CMB anisotropies
- Primary method: cross-calibration with WMAP and B98
- Secondary methods: CMB dipole and RCW 38
• Beams
- Primary: Quasar in Deep Scan Region
- Secondary: Pre-flight using tethered source 1 km from Telescope
• Polarization
- No measured polarized astrophysical sources at our frequencies
- Primary polarization calibration done with a polarized far-field simulator
Cross-Polarization effects
∆T
X=10%
Theory
Q
Scan and
Map
Making
X=1%
Observation
Cross-Polarization effects
X=10%
+ 10% accurate
correction
∆T
Scan and
Map
Making
Q
Theory
X=1%
Observation
Calibration
• We need 1% calibration.
• This can be done, in the lab by means of a special full beam
calibrator {measures Σ, ∆ (or C and X) and the principal
axis direction for each bolometer}
• In flight, thanks to WMAP (which is calibrated to better
than 1% !).
• If we correlate the two bolometers separately with the
unpolarized WMAP maps, we can estimate the responsivity
to better than 1%.
E.Hivon on B98
Channel
Pixel based
C(l) based
B150A
0.95 +/- 0.03
0.96 +/- 0.02
B150A1
0.91 +/- 0.03
0.92 +/- 0.03
B150A2
0.98 +/- 0.03
0.98 +/- 0.02
B150B2
0.95 +/- 0.03
0.95 +/- 0.02
Sum
0.95 +/-0.03
0.95 +/- 0.01
Pre-flight Beam Mapping
Dirigible
Thermal Source
Kevlar String
1 km
1 km
Collimated Polarized Source
Used to measure the
polarization efficiency and the
polarization angles of the full
integrated telescope.
M
Pre-flight calibrations
• The calibration of a polarimeter at ground is
more difficult than the usual photometer
calibrations.
• In particular is very important to study the
co and cross polarization response (beam
and integral) of the polarimeter
• We have developed a polarized, sinemodulated source filling the beam of the
instrument to carry out a through
polarization characterization of all the
detectors.
• There are two wire grid polarizers (P1 and
P2), and a 77K blackbody source with a
diaphragm in the focus of the 1.3 m off-axis
paraboloid, producing a 10’ beam.
• Rotating P2 at constant speed we modulate
the signal (sine wave).
• Rotating P1 (in steps) we change the
illuminator from co-polar to cross-polar (and
all intermediate directions).
p
o
T
r
e
t
e
m
i
r
a
l
o
S
77K
1.3m
P1
P2
The Calibrator
Payload
SCAN
Modulation
•Additional Pointing Sensors with 16 bit abs. encoders
Tracking Star Camera
Pointed Sun Sensor
BOOM03 Flight
Launched:
January 6, 2003
From:
McMurdo Station,
Antarctica
11.7 days of good data
Measurements OK
for 11.6 days
BOOMERanG
landed near
Dome Fuji (h=3700m)
after 14 days
of flight.
The data have been
recovered. The payload
is still there.
Scan Strategy
Region
Size (sq deg)
Goal
Time per 7’ pixel
Deep CMB
115
<EE>
60 sec
Shallow CMB
1130
<TE> and <TT>
3.3 sec
Galactic Plane
390
Polarized Foregrounds
4.7 sec
Imaging the CMB in the Sub-mm!
140 GHz (Coadd of 8 detectors)
340 GHz (single detector: X)
Polarization Maps:
Quick-look data / single pair (of 4) of 143 GHz PSBs /raw data (no compensation for gain drifts!)/ coarse attitude solution
Noise ~ 3µK/20’
pixel in 100 square
degree “deep” region
X+Y
Resolution ~ 10’
Polarized dust
emission evident
near galactic plane
X-Y
Optimal maps obtained with IGLS, the Rome pipeline (Natoli et al. 2001)
Optimal maps obtained with IGLS, the Rome pipeline (Natoli et al. 2001)
Optimal maps obtained with IGLS, the Rome pipeline (Natoli et al. 2001)
Optimal maps obtained with IGLS, the Rome pipeline (Natoli et al. 2001)
E. Hivon
E. Hivon
E. Hivon
Boomerang 2002: ~ 200 square degrees
<TT>
<TE>
<EE>
ell bins ( ∆ ell = 75) < 10% correlated
Bill Jones
Boomerang 2002: ~900 square degrees
<TT>
<TE>
<EE>
ell bins ( ∆ ell = 75) < 10% correlated
Bill Jones
MAP 2 year data
<TT>
<TE>
<EE>
Bill Jones
Filters
Front Horn
Detector
Back Horn
4K Back-to-Back Horn
100mK Horn
1.6K Filter
Holder
QMW
Caltech
Planck HFI
Center Frequency (GHz)
857
545
353
217
143
100
Center
Wavelength
(mm)
Operating T (K)
0.35
0.55
0.85
1.38
2.1
3.0
0.1
0.1
0.1
0.1
0.1
0.1
Fractional Bandwidth
0.33
0.33
0.33
0.33
0.33
0.33
Bandwidth (GHz)
286
182
118
72
48
33
Number of unpolarised
bolometers
4
4
4
4
4
4
Number of polarised bolometers
0
0
8
8
8
0
FWHM (arcmin)
5
5
5
5.5
7.1
9.2
dT/T CMB
microK/K
Sensitivity
6600
17
15
3.8
2.4
2.2
dT/T CMB
microK/KvHz
Sensitivity
2900
0
61.7
54.4
13.8
12.2
14.4
109
27.6
24.4
dT/T Sensitivity
polarised
(U&Q)
CDE Compressor
Harnesses
PHDFA - PPO
PHDFB - Force
PHDFC - Drive
Cooler Current
Regulator
Focal Plane Unit
JT Orifice
Cooler Drive
Electronics
PHDC
4K
CDE Ancillary
Harness
PHDFD
Heat Exchangers
18K
To
CDE
Filter
Flow meter
P
Cryoharness
50K
Buffer
Filter
JT Compressors
PHDA
Getter
P
Ancillary and gas
cleaning equipment
PHDB
Connecting
Pipework
PHDE
Low Temperature
plumbing PHDD
(Includes cryoharness)
150K
50K
50K
The JFET BOX: 72 diff. Channels, < 200mW @ 50K, 3nV/sqrt(Hz)
150K
50K
50K
High Resolution
and
High Frequency
APEX, SP, ALMA …
OLIMPO
An arcmin-resolution
survey of the sky
at mm and sub-mm
wavelengths
PI: Silvia Masi
2.6 m
(Dipartimento di Fisica,
La Sapienza, Roma)
Collaboration with IFACCNR, INGV, Univ. of
Cardiff, CEA Saclay,
Univ. of Santa Barbara
(http://oberon.roma1.infn.it/olimpo)
CMB anisotropy
SZ clusters
30’
Galaxies
Olimpo: list of Science Goals
• Sunyaev-Zeldovich effect
– Measurement of Ho from rich clusters
– Cluster counts and detection of early clusters ->
parameters (Λ
Λ)
• Distant Galaxies – Far IR background
– Anisotropy of the FIRB
– Cosmic star formation history
• CMB anisotropy at high multipoles
– The damping tail in the power spectrum
– Complement interferometers at high frequency
• Cold dust in the ISM
– Pre-stellar objects
– Temperature of the Cirrus / Diffuse component
Olimpo: The Primary mirror
• Lo specchio
primario (2.6m
diametro) e’ stato
verificato nel
laboratorio
proponente.
• Si tratta dello
specchio piu’
grande mai
lanciato su un
pallone
stratosferico.
• Viene fatto
oscillare
lentamente per
realizzare la
modulazione.
Olimpo: The Payload
La navicella
e’ stata
disegnata e
verificata.
E’ in fase di
realizzazione
presso una
piccola
impresa
nazionale
There is still a lot to learn from CMB
photons, and new technologies are
ready …