The universe today is Galaxies form a “cosmic SDSS
filled with Galaxies web” of clusters and voids
Archeology of the Universe
4G us
ly m
Paolo de Bernardis nce
fro
ta
Dipartimento di Fisica di s
Universita’ La Sapienza, Roma
Colloqui della
Scuola di Dottorato in Scienze Fisiche 2dF
Matematiche e Astronomiche
Bologna 06/06/2008
Moreover, the Universe is expanding.
The Universe has not been always the same. How do we know this ?
The Universe Evolved ! We observe light coming from distant
Galaxies, and notice that all wavelengths
How do we know this ?
are systematically shifted to longer
Looking far away is like looking in the past.: (redder) wavelengths.
light can take billion of years to arrive from the
farthest Galaxies we observe. The farther the galaxy, the larger the shift
(Carl Wirtz, Edwin Hubble 1920-1930).
And the earliest galaxies we see are more
irregular, smaller, redder that the nearby
(current) ones.
So, there has been an evolution in the
morphology of Galaxies cz
Hubble Ultra-deep field modern data
nm
General relativity tells us that in • It is common experience that an expanding gas also cools
an homogeneous expanding down. The Universe does the same.
universe, where all lengths • If the Universe has been expanding, it also has been
expand by the some factor, light cooling from a higher temperature state.
wavelengths also expand by the • The temperature (energy) decrease is responsible for a
same factor (cosmological very strong evolution of the content of the Universe.
redshift, z). • In the Very Early Universe the temperature was so high
The classical analogous is the Doppler that it was impossible to have structures like galaxies and
shift, a wavelenght shift proportional to stars. The early Universe was very homogeneous.
the relative speed of the source wrt the • Structures must have formed at some later stage of the
Hubble law:
observer. evolution, when the temperature became low enough.
redshift distance
If all lengths expand by the same factor, z = (Ho/c) d • When did this process start ?
there is no center of the expansion, and
Hubble
Constant • Can we investigate experimentally the hot, homogeneous
70 km/s/Mpc
recession velocity is proportional to early universe, and then the transition to a colder,
distance structured one ? YES, with the CMB
1
• The sequence of events
• Extrapolating all the way back, we would conclude that depends on the content of
the Universe started in a Big Bang, an initial state with different forms of energy in
infinite density and temperature. the Universe:
• However, we do not really know what happened at the – Radiation
Big Bang. – Matter
– Vacuum
• We know that the energy density was so high to be
beyond the capabilities of current physical description:
at such high energies, general relativity and quantum
physics break down, and we do not have a complete,
unified theory describing ultra-high energy phenomena.
• Physics can describe states with extremely high energy,
occurring a split-second after the Big-Bang, when the
temperature was not infinite, but a more comfortable
1020 K or so.
• There is a lot of photons for each ordinary matter particle • We believe that this situation is
in our Universe (109:1) . On the average, we count 400 the result of matter – antimatter
annihilations occurring a few μs
photons/cm3 and 10 billion light
anisotropy of the CMB. years, has an angular size of about 1 degree.
• The measured spectrum
380000 ly
T1
requires a scale 1°
invariant P(k) (n=1) 1° 10°
380000 ly
• Its incredible T2
smoothness requires an 14 000 000 000 ly
inflationary process • How is it possible that regions separated by more than 1°
happening in the first are seen in the COBE map to have the same temperature,
split second after the within 1 part in 10000 ? They could not interact in all
Big Bang. the history of the Universe, from the Big Bang to
G. Smoot et al. 1992 recombination ! (the “Paradox of Horizons”)
4
Inflation ? Sub-horizon scales
• The only known solution to this paradox is the
inflationary hypothesis • If we want to see the seeds of structure
• Those regions are not in causal connection at formation, we must be able to resolve structures
recombination, but they had been in causal connection in smaller than 1°, so that we can see what
the very early universe, when they occupied a happened in regions of the Universe where
microscopic, isothermal volume of the Universe. forces had time to work, and matter clouds could
interact.
• Later, but still very early in the history of the Universe,
at some phase transition a huge, superluminal inflation • COBE was a small satellite with two small
of space happened, boosting microscopical scales to microwave antennas. The resolution was 10°.
cosmological scales. • What we need to make this measurement is a real
• Can we produce a proof that this process ever happened ? microwave telescope. An angular resolution of
Yes, with the CMB, in the (near ?) future. 1
space 2o
High density Universe
horizon
14 billion light years
Here, now 0.5o
Low density Universe Ω1 Ω=1 Ω 3000 K
0.5o t After recombination T 217 GHz
Independent of redshift !
US
OLIMPO (PI Silvia Masi, Roma) • 4 frequency bands
simultaneously. Uniqueness of
• Focal plane can host >400 bolometers
• from Cardiff (P. Mauskopf) and Genoble (P. Camus)
• Optimally sample the
spectrum of the SZ
OLIMPO
effect.
-4
6.0x10
150 GHz 220 GHz 340 GHz 540 GHz 7keV
• Opposite signals at 10keV
410 GHz and at 150 4.0x10
-4
15keV
GHz provide a clear
20keV
signature of the SZ
ΔI (mJy/sr)
detection. 2.0x10
-4
30’
• 4 bands allow to clean
the signal from dust 0.0 150 240 410 600
and CMB, and even to
measure Te
-4
-2.0x10
• Resolution: 2x(Planck)
• Detectors: 10x(Planck) -4
-4.0x10
• Integration time per 0 200 400 600 800
cluster: 10x(Planck) ν (GHz)
(40 clusters/flight +
blind survey) - 0 + +
11
DARK
MATTER
Cosmic
Microwave γ rays
Background Cosmic
Rays
Power Spectrum of χ−χ annihilation photons
CMB anisotropy
in X and γ-rays spectra
Observation of SZ effect χ−χ annihilation products
in CR spectra ……
in selected clusters
……… ………
Flights: 2007 & 2008
What is Dark Matter ? Dark Matter Annihilation Products
• Hp: Weakly Interacting Supersymmetric Particles
(WIMPs)
• Lightest one predicted by SUSY : Neutralino χ
• Could be measured by LHC
• χs tend to cluster in the center of astrophysical
structures
• Annihilation of Neutralinos would produce fluxes of
– Neutral and charged pions
– Secondary electrons protons
– Neutrinos
– etc.
• They produce various effects
• One of them is the SZ from the charged
component (see Colafrancesco, 2004)
SZ effect from χχ annihilation What is Dark Matter ?
• Hp: Weakly Interacting Supersymmetric Particles (WIMPs)
• Lightest one predicted by SUSY : Neutralino χ
• Could be measured by LHC
• χs tend to cluster in the center of astrophysical structures
• Annihilation of Neutralinos would produce fluxes of
– Neutral and charged pions
– Secondary electrons protons
– Neutrinos
– etc.
• They produce various effects
• One of them is the SZ from the charged component (see
Colafrancesco, 2004)
• Subdominant with respect to SZE from the gas.
• We need clusters where Dark Matter and Baryonic Matter
are separated.
12
1E0657-56 1E0657-56
9’ 7.5 ’
SZ effect from DM [Colafrancesco 2004 , A&A, 422, L23]
SZ effect at clump centres
ΔI(x) = I(x) – I0(x)
0
I0(x) Ith(x)
IDM(x)
ν' 4
thermal e- =
ν 3
ν 4 2
' [Colafrancesco, de Bernardis, Masi, Polenta & Ullio 2006]
relativistic e- = γ −1
ν 3
Isolating SZDM (at 223 GHz)
Mχ = 20 GeV Mχ = 40 GeV Mχ = 80 GeV
The SZE from the hot gas disappears at x0,th (∼ 220-223 GHz)
while the SZDM expected at the locations of the two DM clumps
[Colafrancesco, de Bernardis, Masi, Polenta & Ullio 2006] remains negative and with an amplitude and spectrum
which depend on Mχ.
[Colafrancesco, de Bernardis, Masi, Polenta & Ullio 2006]
13
SAGACE
Spectroscopic Active Galaxies And Clusters Explorer
Did Inflation really happen ?
• The ideal continuation of OLIMPO • We do not know. Inflation has not been
• Selected by ASI for a phase-A study as a small mission
proven yet. It is, however, a mechanism able
to produce primordial fluctuations with the right
• 2.6 m telescope + FTS spectrometer on a Soyuz
characteristics.
• Spectra of thousands of SZ clusters and AGNs
• Four of the basic predictions of inflation have
• Uni. La Sapienza / Uni. Mi. Bicocca / Uni. Genova / Kayser Italiana / ASDC-ASI
been proven:
– existence of super-horizon fluctuations
– gaussianity of the fluctuations
– flatness of the universe
– scale invariance of the density perturbations
• One more remains to be proved: the stochastic
background of gravitational waves produced
during the inflation phase.
• CMB can help in this – see below.
CMB polarization y y
-10ppm +10ppm
• CMB radiation is Thomson scattered at recombination. - +
• If the local distribution of incoming radiation in the
rest frame of the electron has a quadrupole moment, x x
the scattered radiation acquires some degree of linear + +
polarization.
- - - -
Last scatte y
ring surface - +
x
-
= e- at last scattering
Quadrupole from P.G.W.
If inflation really • If inflation really happened:
It stretched geometry of space to
happened… nearly Euclidean
It produced a nearly scale invariant
spectrum of gaussian density
• It stretched geometry of OK fluctuations
space to nearly Euclidean It produced a stochastic background of
gravitational waves: Primordial G.W.
• It produced a nearly scale The background is so faint that even
E-modes
invariant spectrum of density OK LISA will not be able to measure it.
fluctuations • Tensor perturbations also produce
quadrupole anisotropy. They generate
• It produced a stochastic irrotational (E-modes) and rotational
background of gravitational
waves.
? (B-modes) components in the CMB
polarization field.
• Since B-modes are not produced by scalar
fluctuations, they represent a signature of
inflation. B-modes
14
B-modes from P.G.W.
• The amplitude of this effect is very small, but Pure E(left) & B(right)
depends on the Energy scale of inflation. In fact the
amplitude of tensor modes normalized to the scalar
ones is:
1/ 4
⎛ C2 ⎞ Inflation potential
1/ 4
⎛T ⎞
GW
V 1/ 4
⎜ ⎟ ≡ ⎜ Scalar ⎟
⎜C ⎟ ≅
⎝S⎠ ⎝ 2 ⎠ 3.7 ×1016 GeV
• and
l(l + 1) B ⎡ V 1/ 4 ⎤
cl max ≅ 0.1μK ⎢ ⎥
2π ⎢ 2 × 10 GeV ⎥
⎣
16
⎦
• There are theoretical arguments to expect that the
energy scale of inflation is close to the scale of GUT
i.e. around 1016 GeV.
• The current upper limit on anisotropy at large scales
gives T/S
data: July 2005 Piacentini et al. astro-ph/0507507 :
Five papers: Montroy et al. astro-ph/0507514 :
MacTavish et al. astro-ph/0507503 : cosmological PSB Pair
parameters
Silvia Masi Francesco Bill Jones
Piacentini Tom Montroy Carrie MacTavish
BOOMERanG-03
From Page et al. 2006
06/01/2003
Where do we go from here Sensitivity
• Polarization measurements do not constrain
• B03 has shown that Polarization Sensitive
parameters better than anisotropy Bolometers work well for CMB polarization
measurements, yet. measurements.
• Most of the weight in the results above is in • Their sensitivity is close to be photon-noise-
Temperature power spectra. limited. In Planck-HFI the same bolometers
• If we want to constrain better the cosmological will be cooled a factor 3 more and will be
model, and finally detect B-modes, and we need limited only by quantum fluctuations of the
to improve in three ways:
CMB itself. It is useless to improve the
detector noise below the photon noise limit.
1. Sensitivity
2. Control of systematics
3. Knowledge of foregrounds
16
A post-Planck mission Sensitivity
• Planck will or will not detect Inflationary B-Modes (depending on
amplitude, foregrounds, systematics… and if they are really
there). • At variance with interferometers,
• In a diffraction limited 150 GHz survey, CMB BLIP gives 1 μK in Bolometer technology is easily scalable,
1 min of integration. But we need to observe 105 pixels !
and the throughput can be larger than λ2.
• Focal planes hosting thousands of
150 GHz,10% BW, λ
2
10 150 GHz, 10% BW, 1 cm sr
2 bolometers are being developed already.
30 GHz, 10% BW, λ
2
error per pixel (μK)
1
0.1
CMB BLIP
0.01
1 10 100 1000 10000
integration time (s)
• We need to increase the mapping speed using more detectors
than in the Planck focal plane.
Large Bolometer Arrays Large Bolometer Arrays
• > 1000 TES bolometers for the South Pole • > 1000 TES bolometers for SPIDER a proposed
Telescope devoted to SZ (Adrian Lee, Berkeley) spinning polarimeter on a LDB (Andrew Lange,
Caltech) devoted to large scale CMB polarization
Large Bolometer Arrays
• >1000 TES bolometers for the EBEX CMB
polarization balloon telescope (Shaul Hanany,
Minneapolis)
Bolom.
Array
Planck
From the
EBEX
proposal
17
Control of Systematic Effects
• B03 has shown that systematic effects can be
controlled by a combination of
– Multifrequency capabilities
– Scan variation
– Polariziation angle redundancy
– Variations of observing conditions
– Accurate pre-flight and in-flight calibration
• This was OK at the level of sensitivity of B03 (i.e. 3σ
detection of E-modes, 4 μK rms).
• Nobody knows how to control systematics for a B-
modes experiment (100 PSB at >350 GHz.
Frequency range complementary to PILOT
(higher f. J.F. Bernard, Toulouse)
18
A post-Planck mission • European proposal recently
B-Pol submitted to ESA (Cosmic
Vision). PI PdB.
(www.b-pol.org)
• A post-Planck mission, with a large array of • ESA encourages the
sensitive polarized detectors, is needed to development of technology and
detect B-modes and constrain inflationary resubmission for next round
• Detector Arrays development
parameters (energy scale, r, nT, V(φ) …) activities (KIDs in Rome, TES
– NASA – Beyond Einstein : Inflation Probe in Genova etc.)
– ESA - Cosmic Vision : B-Pol • A balloon-borne payload being
developed with ASI (B-B-Pol).
• Meanwhile, laboratory, ground-based, and
balloon-borne experiments are necessary
develop the needed technology
Pre-phase-A study for the B-Pol program Sensitivity and frequency coverage: the focal plane
• Baseline technology: TES bolometers arrays
Corrugated feedhorns Sub-K, 600 mm
• A coordinated effort of the italian CMB community for polarization purity and
to measure the polarization of the CMB beam symmetry
• Study and compare the performance of
– Balloon option
– Small satellite option
– Medium satellite option
IASF-Bo
IRA-Bo
IEIIT
Optical system:
B-B-Pol: The Balloon Option
• Wide field,
WHY ?
• low cross-pol,
• low emissivity • Get important science
(complementary to NASA’s SPIDER, EBEX)
Possible solution:
modified telecontric • Validate needed technology, for next round of
telescope ESA cosmic vision
HOW ?
• ASI polar-night flight -> large sky coverage
• Three instruments to cover from 40 to 220 GHz
• Low angular resolution – large scales
• High-Throughput Channels – High sensitivity
• Single-mode channels – Foregrounds
• Large ground shields
• No optics – no spurious polarization
HWP
19
Worksheet Sensitivity
Worksheet Performance
37 detectors
12 cm
20
Spinning
HWP
2K B-B-Pol: A spinner in the polar night
W Polyethilene 0.3K
ire
Lens
G
ri d • Can provide extremely competitive
Beam 2o FWHM
measurements of CMB polarization at
large angular scales.
25 cm
diam • Is complementary to NASA’s SPIDER and
40 overmoded
Detectors, diam 1.7 cm EBEX
(10 modes @ 150 GHz,
Polyethilene 20 modes @ 220 GHz) • Will qualify, producing great science
Lens
B-B-Pol: High results, italian technology, in view of next
Frequency 0.3K Cosmic Vision call.
Instrument 40 overmoded • Will exploit the unique ASI-ARR capability
(one of the two Detectors, diam 1.7 cm
bands shown) (10 modes @ 150 GHz,
to launch long duration balloons in the
20 modes @ 220 GHz) polar night
• TES arrays are being prepared for
• The readout system for TESs requires SQUIDs
– South Pole Telescope and is very complex.
– APEX (Atacama)
Large Dishes
– ACT (Atacama)
• KIDs (see e.g. Zmuidzinas, Caltech) represent a
– IRAM 30m dish (Pico Veleta)
good alternative because are intrinsically
– PolarBear (White Mountain)
multiplexable
– EBEX
– SPIDER
Balloons • Cold electron bolometers (e.g. Kuzmin, Chalmers)
– OLIMPO
– B-Pol represent a good alternative because the readout
– SAGACE …. Satellites system is much simpler
RC 0.3K - 0.1K RC
GHz RF (…+fN-1+fN+fN+1+…)
CMB CMB CMB
Pixel N-1 Pixel N Pixel N+1
fN-1 fN fN+1
Mazin (Caltech)
21
• First prototype: a 0.3K Al resonator @ 6 GHz
• Currently under test
We have a true image of the early universe,
KIDs (RIC – INFN) And we have new intriguing questions to answer
New techniques to develop .. A lot of work to do.
22