The Sunyaev-Zel’dovich Effect
Jason Glenn
APS
Historical Perspective
Physics of the SZ Effect
-------------------------------------------Previous Observations & Results
Bolocam
Imminent Experiments
Future Work
References
Historical Perspective
•CMB discovered in
1964 by Penzias and
Wilson
•COBE 1989: perfect
blackbody to 1/105,
primary anisotropies
measured
•However, in 1970
Sunyaev & Zel’dovich
predicted the SZ effect:
secondary anisotropies
in the CMB
TCMB = 2.725 K
Physics of the SZ Effect
Mechanism & Thermal Effect
CMB photons
T = (1 + z) 2.725K
observer
z=0
galaxy cluster
with hot ICM
z~0-3
scattered
photons
(hotter)
last scattering
surface
z ~ 1100
Sunyaev &
Zeldovich (1970)
Spectral
shift
CMB photons have a ~1% chance of inverse Compton scattering off of the
ICM electrons; photon number is conserved
Physics of the SZ Effect
Functional Form
y parameter
•Temperature shift proportional to the gas pressure,
neTe, & mass dl
•CMB photon energies boosted by ~kTe/(mec2)
•kTe ~ 10 keV, Te ~ 108 K  relativistic
•x = h/(kTe)
•f(x) is the spectral dependence
•Notice that the temperature shift is redshift
independent  unbiased surveys for clusters
Physics of the SZ Effect
The Kinetic Effect: a Doppler boost from the peculiar velocity of the cluster
Spectral distortion:
Null in thermal 
measure kinetic
Increment
Kinetic effect is
small
Decrement
Physics of the SZ Effect
What the thermal effect looks like
•Simulations, of course!
• = 2 mm
•“Maps” are 1° on a side
•SZ effect is an increment at 2 mm
Physics of the SZ Effect
The Angular Power Spectrum
•Secondary anisotropies can be
measured independent of cluster
detection
•l is the multipole number (as in
quantum mechanics); (°) ~
200°/l
•Vertical units: T2 – power
usually measured as an excess
variance above the noise, Cl is
per l – there are more
independent multipoles at high l
•Dashed and dotted lines are
models
•The signals are small: ~ 15 mK
@ 30 GHz, ~ 5 mK @ 150 GHz
•Tentative detections so far
(more on this Friday)
Green is 30 GHz, or 1 cm
Pink is 150 GHz, or 2 mm
Physics of the SZ Effect
Cosmological Utility
What can be measured when combined with other
observations:
•H0
•Cluster masses
•Cluster abundance as a function of redshift
•, , w
•Spectral index of initial perturbations (nonGaussianity)
•Cluster evolution
Next, we’ll discuss SZ observations and some results
Previous Observations
Images from Interferometers
•Image from Carlstrom group using OVRO/BIMA
interferometer at 30 GHz
•Spectral measurements a compendium – confirms spectrum
through RJ tail
•To date, only pointed observations toward massive clusters
•Measurements of the kinetic effect will be very hard,
depending on precision of multiband calibration
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270 GHz
(2mm/1mm) observations?
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270 GHz
(2mm/1mm) observations?
The amplitude of the SZ thermal effect is larger at 30 GHz
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270 GHz
(2mm/1mm) observations?
The amplitude of the SZ thermal effect is larger at 30 GHz
Contamination by cluster, foreground, and background radio point
sources (quasars) would be a problem at 30 GHz.
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270 GHz
(2mm/1mm) observations?
The amplitude of the SZ thermal effect is larger at 30 GHz
Contamination by cluster, foreground, and background radio point
sources (quasars) would be a problem at 30 GHz.
Contamination by dust from background, lensed galaxies is a
potential problem at 1 mm.
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270 GHz
(2mm/1mm) observations?
The amplitude of the SZ thermal effect is larger at 30 GHz
Contamination by cluster, foreground, and background radio point
sources (quasars) would be a problem at 30 GHz.
Contamination by dust from background, lensed galaxies is a
potential problem at 1 mm.
In practice, the angular resolution achievable with each is about the
same because bolometer arrays are used for short-wavelength
observations and interferometers are used for long-wavelength
observations.
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270
GHz (2mm/1mm) observations?
The amplitude of the SZ thermal effect is larger at 30 GHz
Contamination by cluster, foreground, and background radio point
sources (quasars) would be a problem at 30 GHz.
Contamination by dust from background, lensed galaxies is a
potential problem at 1 mm.
In practice, the angular resolution achievable with each is about the
same because bolometer arrays are used for short-wavelength
observations and interferometers are used for long-wavelength
observations.
1 mm and 2 mm observations are necessary to measure the kinetic
SZ effect.
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270 GHz
(2mm/1mm) observations?
The amplitude of the SZ thermal effect is larger at 30 GHz
Contamination by cluster, foreground, and background radio point
sources (quasars) would be a problem at 30 GHz.
Contamination by dust from background, lensed galaxies is a
potential problem at 1 mm.
In practice, the angular resolution achievable with each is about the
same because bolometer arrays are used for short-wavelength
observations and interferometers are used for long-wavelength
observations.
1 mm and 2 mm observations are necessary to measure the kinetic
SZ effect.
Emission/absorption by the atmosphere is not a huge problem at
long wavelengths for interferometers because the noise between
telescopes is not highly correlated.
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270 GHz
(2mm/1mm) observations?
The amplitude of the SZ thermal effect is larger at 30 GHz
Contamination by cluster, foreground, and background radio point
sources (quasars) would be a problem at 30 GHz.
Contamination by dust from background, lensed galaxies is a
potential problem at 1 mm.
In practice, the angular resolution achievable with each is about the
same because bolometer arrays are used for short-wavelength
observations and interferometers are used for long-wavelength
observations.
1 mm and 2 mm observations are necessary to measure the kinetic
SZ effect.
Emission/absorption by the atmosphere is not a huge problem at
long wavelengths for interferometers because the noise between
telescopes is not highly correlated.
In contrast, atmospheric noise is much worse at short wavelengths
– much worse than anticipated!
Some Questions
•What are the tradeoffs between 30 GHz (1 cm) and 150/270 GHz
(2mm/1mm) observations?
Clearly, we need both.
Atmospheric Noise
Emission, rather than absorption, is the primary problem: fluctuation in the
arrival rate of background photons from water molecules in the sky (and the
telescope, the ground, the instrument…)
300 m
cm band
The sky over
Mauna Kea
2 mm 1 mm
Emission = 1 Transmission
Bolocam
Physics of the SZ Effect
The Angular Power Spectrum
We need
more high-l
data!
Green is 30 GHz, or 1 cm
Pink is 150 GHz, or 2 mm
Bolocam
Detectors
Incoming Photons
Absorber
Q
Weak Thermal
Link
Si3N4 micromesh “spider web” bolometer
JPL Micro Devices Lab
Bath
(T ≤ 270 mK)
Bolocam
Bolometers
In 1878, Samuel Pierpont Langley invented the bolometer.
Oh, Langley devised a bolometer:
It’s really a kind of thermometer
Which measures the heat
From a polar bear’s feet
At a distance of half a kilometer1.
1Anonymous
Bolocam
Bolometers
In 1878, Samuel Pierpont Langley invented the bolometer.
Oh, Langley devised a bolometer:
It’s really a kind of thermometer
Which measures the heat
From a polar bear’s feet
At a distance of half a kilometer1.
1Anonymous
With Bolocam on the CSO, we can detect a polar bear’s foot with a
S/N of one at a distance of 3 km in one second of integration time2.
2(In
good weather!)
Bolocam
Cryostat
Instrument
CSO
5 in.
Focal
Plane
Bolometer Array
•144 bolometers
• = 1.1, 2.1 mm
•300 mK
Thanks to Sunil for
some graphics in this
lecture!
Collaborators
(Cardiff, Caltech, JPL, & CU)
P.A.R. Ade, J.E. Aguirre, J.J.
Bock, S.F. Edgington, A.
Goldin, S.R. Golwala, D.
CU
Haig, A.E. Lange, G.T.
Caltech
Laurent, P.R. Maloney,
P.D.
Mauskopf, P. Rossinot,
J.
JPL
Sayers, P. Stover, H. Cardiff
Nguyen
Bolocam
The reality of sky noise (a must read for theorists)
“Average”
subtraction takes
out 90% of the
noise, but we need
>99% with retention
of large-scale
structure
Residual noise and
itsy-bitsy SZ signal!
“White” noise:
ultimate sky
subtraction
Bolocam is a bolometerarray pioneer and the
other groups are looking
to us; we’re only in the
lead by ~12 months! (this
part is for you, Andrew)
Imminent MM-Wave Experiments
High-l Anisotropies
Nils
References
An excellent review from an observer’s perspective and the source of
some of the graphics in this lecture: “Cosmology with the SunyaevZel’dovich Effect”, Carlstrom, Holder, & Reese, ARAA, 2002, Vol. 40,
pp. 643-680
•H0:
•Cluster mass fraction:
•Cluster peculiar velocities:
Long-Term Future Work
Probing the physics of galaxy cluster evolution
Hallman &
Burns, et al.