nanjing, march 2003 using the sunyaev-zeldovich effect to probe the gas in clusters mark birkinshaw...
TRANSCRIPT
Nanjing, March 2003
Using the Sunyaev-Zel’dovich effect to probe the gas in clusters
Mark Birkinshaw
University of Bristol
Mark Birkinshaw, U. Bristol 2
Nanjing, March 2003
Outline
1. The origin of the effect
2. SZ effect observations
3. SZ effect science: clusters
4. SZ effect science: cosmology
5. The future: dedicated SZ instruments
6. Summary
Mark Birkinshaw, U. Bristol 3
Nanjing, March 2003
1. The origin of the effect
Clusters of galaxies contain extensive hot atmospheres
Te 6 keV
np 103 protons m-3
L 1 Mpc
2 Mpc
Mark Birkinshaw, U. Bristol 4
Nanjing, March 2003
Inverse-Compton scatterings
• Cluster atmospheres scatter photons passing through them. Central iC optical depth
enpT L
• Scatterings changes the average photon frequency by a fraction
kBTe/mec2
Mark Birkinshaw, U. Bristol 5
Nanjing, March 2003
Microwave background spectrum
I
Fractional intensity change I/I = -2 (/ e 10-4
Mark Birkinshaw, U. Bristol 6
Nanjing, March 2003
Thermal SZ effect• Fractional intensity change in the CMB
I/I = -2 (/ e 10-4
• Effect in brightness temperature terms
TRJ = -2 Tr (/ e -300 K
• Brightness temperature effect, TRJ, is independent of redshift
• Flux density effect, S, decreases as DA-2, not DL
-
2, and depends on redshift
Mark Birkinshaw, U. Bristol 7
Nanjing, March 2003
Spectrum of thermal effect
• spectrum related to gradient of CMB spectrum
• zero at peak of CMB spectrum (about 220 GHz)
• weak dependence on Te
Mark Birkinshaw, U. Bristol 8
Nanjing, March 2003
SZ sky predicted using structure formation code (few deg2, y = 0 – 10-4)
CMB primordial fluctuations ignored
da Silva et al.
Predicted SZ effect sky
Mark Birkinshaw, U. Bristol 9
Nanjing, March 2003
SZ effect and CMB power spectrum
Figure from Molnar & Birkinshaw 2000
thermal SZ
kinematic SZ
RS effect
Mark Birkinshaw, U. Bristol 10
Nanjing, March 2003
Attributes of SZ effect
TRJ is a redshift-independent function of cluster thermal energy, it is a calorimeter
TRJ has a strong association with rich clusters of galaxies, it is a mass finder
TRJ contains a weak redshift-independent kinematic effect, it is a radial speedometer
TRJ has polarization with potentially more uses, but signal is tiny
Mark Birkinshaw, U. Bristol 11
Nanjing, March 2003
2. SZ effect observations• Interferometers: e.g., Ryle, BIMA, OVRO
– structural information– baseline range
• Single-dish radiometers: e.g., OVRO 40-m, OCRA – speed– systematic errors from spillover
• Bolometers: e.g., SuZIE, SCUBA, ACBAR– speed– structural and spectral information– weather
Mark Birkinshaw, U. Bristol 12
Nanjing, March 2003
Ryle telescope
• first interferometric map• Abell 2218• brightness agrees with
single-dish data• limited angular dynamic
range
Figure from Jones et al. 1993
Mark Birkinshaw, U. Bristol 13
Nanjing, March 2003
Interferometers
• restricted angular dynamic range
• high signal/noise (long integration possible)
• clusters easily detectable to z 1
Figure from Carlstrom et al. 1999
Mark Birkinshaw, U. Bristol 14
Nanjing, March 2003
Interferometers
• restricted angular dynamic range set by baseline and antenna size
• good rejection of confusing radio sources
Abell 665 model, VLA observation
available baselines
Mark Birkinshaw, U. Bristol 15
Nanjing, March 2003
Interferometers
• Good sky and ground noise rejection because of phase data
• Long integrations and high signal/noise possible
• 10 years of data, tens of cluster maps• SZ detected for cluster redshifts from 0.02
(VSA) to 1.0 (BIMA)• Could be designed with better baseline range
Mark Birkinshaw, U. Bristol 16
Nanjing, March 2003
Single-dish radiometers
• Potentially fast way to measure SZ effects of particular clusters
• Multi-beams better than single beams at subtracting atmosphere, limit cluster choice
• Less fashionable now than formerly: other techniques have improved faster
• New opportunities: e.g., GBT
Mark Birkinshaw, U. Bristol 17
Nanjing, March 2003
Single-dish radiometers• fast at measuring integrated
SZ effect of given cluster• multi-beam limits choice of
cluster, but subtracts sky well• radio source worries• less used since early 1990s• new opportunities, e.g. GBT
Figure from Birkinshaw 1999
Mark Birkinshaw, U. Bristol 18
Nanjing, March 2003
Distribution of central SZ effects
• Mixed sample of 37 clusters
• OVRO 40-m data, 18.5 GHz
• No radio source corrections
• 40% of clusters have observable T < -100 K
Mark Birkinshaw, U. Bristol 19
Nanjing, March 2003
Bolometers
• Should be fast way to survey for SZ effects
• Wide frequency range possible on single telescope, allowing subtraction of primary CMB structures
• Atmosphere a problem at every ground site
• Several experiments continuing, SuZIE, MITO, ACBAR, BOLOCAM, etc.
Mark Birkinshaw, U. Bristol 20
Nanjing, March 2003
SCUBA
850 µm images: SZ effect measured in one; field too small
Mark Birkinshaw, U. Bristol 21
Nanjing, March 2003
MITO
• MITO experiment at Testagrigia
• 4-channel photometer: separate components
• 17 arcmin FWHM
• Coma cluster detection
Figure from De Petris et al. 2002
Mark Birkinshaw, U. Bristol 22
Nanjing, March 2003
Viper + ACBAR
• Since 2001: 16-pixel bolometer (ACBAR); 150, 220, 280 GHz (+350 GHz in 2001)
• Dry air, 3º chopping tertiary, large ground shield
• 4 – 5 arcmin FWHM• Excellent for SZ work
Mark Birkinshaw, U. Bristol 23
Nanjing, March 2003
ACBAR cluster observations
2002 cluster observations: three of nine objects detected?
Mark Birkinshaw, U. Bristol 24
Nanjing, March 2003
SZ effect status• About 100 cluster detections
– high significance (> 10) measurements– multi-telescope confirmations– interferometer maps, structures usually from X-rays
• Spectral measurements improving but still rudimentary – no kinematic effect detections
• Preliminary blind and semi-blind surveys
Mark Birkinshaw, U. Bristol 25
Nanjing, March 2003
3. SZ effect science: clusters• Integrated SZ effects
– total thermal energy content– total hot electron content
• SZ structures– not as sensitive as X-ray data– need for gas temperature
• Mass structures and relationship to lensing
• Radial peculiar velocity via kinematic effect
Mark Birkinshaw, U. Bristol 26
Nanjing, March 2003
Integrated SZ effects• Total SZ flux density
thermaleeRJ UdzTndS Thermal energy content immediately measured in redshift-independent wayVirial theorem then suggests SZ flux density is direct measure of gravitational potential energy
Mark Birkinshaw, U. Bristol 27
Nanjing, March 2003
Integrated SZ effects• Total SZ flux density
eeeeRJ TNdzTndS If have X-ray temperature, then SZ flux density measures electron count, Ne (and hence baryon count)Combine with X-ray derived mass to get fb
Mark Birkinshaw, U. Bristol 28
Nanjing, March 2003
SZ effect structures
• Currently only crudely measured by SZ methods (restricted angular dynamic range)
• X-ray based structures superior
• Structure more extended in SZ than X-ray: ne rather than ne
2 dependence. SZ should show more about outer gas envelope, but need better sensitivity
Mark Birkinshaw, U. Bristol 29
Nanjing, March 2003
SZ effects and lensingWeak lensing measures ellipticity field e, and so
)(),(1 2
crit θθθθ ii ed
Surface mass density as a function of position can be combined with SZ effect map to give a map of fb SRJ/
Mark Birkinshaw, U. Bristol 30
Nanjing, March 2003
Total and gas massesInside 250 kpc:
XMM +SZ
Mtot = (2.0 0.1)1014 M
Lensing
Mtot = (2.7 0.9)1014 M
XMM+SZ
Mgas = (2.6 0.2) 1013 M
CL 0016+16 with XMMWorrall & Birkinshaw 2002
Mark Birkinshaw, U. Bristol 31
Nanjing, March 2003
Cluster radial peculiar velocity
• Kinematic effect separable from thermal SZ effect because of different spectrum
• Confusion with primary CMB fluctuations limits velocity accuracy to about 150 km s-1
• Velocity substructure in atmospheres will reduce accuracy further
• Statistical measure of velocity distribution of clusters as a function of redshift in samples
Mark Birkinshaw, U. Bristol 32
Nanjing, March 2003
Cluster radial peculiar velocityNeed• good SZ spectrum• X-ray temperature
Confused by CMB structure
Sample vz2
Three clusters so far, vz 1000 km s
A 2163; figure from LaRoque et al. 2002.
Mark Birkinshaw, U. Bristol 33
Nanjing, March 2003
4. SZ effect science: cosmology• Cosmological parameters
– cluster-based Hubble diagram– cluster counts as function of redshift
• Cluster evolution physics– evolution of cluster atmospheres via cluster counts – evolution of radial velocity distribution– evolution of baryon fraction
• Microwave background temperature elsewhere in Universe
Mark Birkinshaw, U. Bristol 34
Nanjing, March 2003
Cluster Hubble diagram
X-ray surface brightness
X ne2 Te
½ L SZ effect intensity change
I ne Te L
Eliminate unknown ne
L I2 X1 Te
3/2
H0 X I2 Te
3/2
Mark Birkinshaw, U. Bristol 35
Nanjing, March 2003
Cluster distances and massesCL 0016+16
DA = 1.36 0.15 Gpc
H0 = 68 8 18 km s-1 Mpc-1
Worrall & Birkinshaw 2002
Mark Birkinshaw, U. Bristol 36
Nanjing, March 2003
Hubble diagram
• poor leverage for other parameters
• need many clusters at z > 0.5
• need reduced random errors
• ad hoc sample • systematic errors
From Carlstrom, Holder & Reese 2002
Mark Birkinshaw, U. Bristol 37
Nanjing, March 2003
Critical assumptions
• spherical cluster (or randomly-oriented sample)
• knowledge of density and temperature structure to get form factors
• clumping negligible
• selection effects understood
need orientation-independent sample
Mark Birkinshaw, U. Bristol 38
Nanjing, March 2003
Blind surveys• SZ-selected samples
– almost mass limited and orientation independent
• Large area surveys– 1-D interferometer surveys slow, 2-D arrays better– radiometer arrays fast, but radio source issues– bolometer arrays fast, good for multi-band work
• Survey in regions of existing surveys– XMM-LSS survey region ideal, many deg2
Mark Birkinshaw, U. Bristol 39
Nanjing, March 2003
Cluster counts and cosmologyCluster counts and redshift distribution provide strong constraints on 8, m, and cluster heating.
z
dN/dzm=1.0
m=0.3
m=0.3
Figure from Fan & Chiueh 2000
Mark Birkinshaw, U. Bristol 40
Nanjing, March 2003
ACBAR blind survey
CMB5 field, filtered, pointing source blanked. Features at s/n > 4.
Mark Birkinshaw, U. Bristol 41
Nanjing, March 2003
Baryon mass fraction
SRJ Ne Te
Total SZ flux total electron count total baryon content.
Compare with total mass (from X-ray or gravitational lensing) baryon fraction
Figure from Carlstrom et al. 1999.
b/m
Mark Birkinshaw, U. Bristol 42
Nanjing, March 2003
Microwave background temperature
• Ratio of SZ effects at two different frequencies is a function of CMB temperature (with slight dependence on Te and cluster velocity)
• So can use SZ effect spectrum to measure CMB temperature at distant locations and over range of redshifts
• Test T (1 + z)
Mark Birkinshaw, U. Bristol 43
Nanjing, March 2003
Microwave background temperature• Test T (1 + z)
• SZ results for two clusters plus results from molecular excitation
Battistelli et al. (2002)
Mark Birkinshaw, U. Bristol 44
Nanjing, March 2003
5. The future: dedicated SZ instruments
Today Future
CBI MITO/MAD AMiBA APEX
OVRO 40-m Ryle OCRA ALMA
VSA ACBAR AMI etc.
MAP BOLOCAM Planck
SuZIE etc. SZA
Mark Birkinshaw, U. Bristol 45
Nanjing, March 2003
Survey speeds
• OCRA will be fastest survey radiometer
• AMiBA will be fastest survey interferometer
• Frequencies complementary
Mark Birkinshaw, U. Bristol 46
Nanjing, March 2003
New SZ interferometers
AMiBA
SZA
AMI
solid nearby high-M clusterdashed high-z low-M cluster
AMIBA 90 GHzSZA 30 GHzAMI 15 GHz
Complementary spectral coverage
Short baselines crucial for SZ detection
Long baselines for radio sources
Mark Birkinshaw, U. Bristol 47
Nanjing, March 2003
AMiBA
• ASIAA/NTU project• Operational in 2004, prototype
2002• 19(?) dishes, 1.2/0.3 m
diameters, 1.2 – 6 m baselines = 95 GHz, = 20 GHz• Dual polarization• 1.3 mJy/beam in 1 hr
Mark Birkinshaw, U. Bristol 48
Nanjing, March 2003
XMM-LSS survey SZ follow-up
• XMM survey of 64 deg2 to 5 10-15 erg cm-2 s-1 (0.5 – 2.0 keV)
• Expect 300 sources deg-2, 12% clusters 2000 clusters
• SZ imaging will give Hubble diagram to z = 1
• Combining X-ray, SZ, shear mapping at z < 0.5 will give baryon fraction and total masses
• possible SZ detection of IGM filaments?
Mark Birkinshaw, U. Bristol 49
Nanjing, March 2003
Cluster finding: X-ray vs SZ
• AMiBA is better than XMM for clusters at z > 0.7
• interferometers provide almost mass limited catalogues
• may find X-ray dark clusters
LX(5)
z
Mark Birkinshaw, U. Bristol 50
Nanjing, March 2003
OCRA
Torun Observatory, Jodrell Bank, Bristol, Bologna
OCRA-p
Mark Birkinshaw, U. Bristol 51
Nanjing, March 2003
OCRA
• 30 GHz• Tsys = 40 K• 1 arcmin FWHM
beam• 5 mJy sensitivity in
10 sec• now on telescope• OCRA-F in progress
Mark Birkinshaw, U. Bristol 52
Nanjing, March 2003
APEXMPI project at
Chajnantor
300-element bolometer array at 870 m ideal for SZ
(117-element prototype shown)
Mark Birkinshaw, U. Bristol 53
Nanjing, March 2003
6. Summary (1)
• SZ effect is a major cluster and cosmological probe
• SZ maps dominated by massive objects at z 0.5, filaments and groups tend to average out
• SZ effect easily detectable to z > 1
• SZ effects appear on lumpy background, adds noise
Mark Birkinshaw, U. Bristol 54
Nanjing, March 2003
Summary (2)
• Individual cluster SZ effects give– total thermal energy contents– total electron contents– structural information (especially on large scales)– cluster masses– microwave background temperature at distant points
Mark Birkinshaw, U. Bristol 55
Nanjing, March 2003
Summary (3)
• Sample studies give– Hubble diagram and cosmological parameters– cluster number counts and cosmological parameters– baryon mass fraction– evolution of cluster atmospheres– evolution of radial velocities– redshift-dependence of microwave background
temperature
Mark Birkinshaw, U. Bristol 56
Nanjing, March 2003
Summary (4)
• Improved SZ data could give– radio source energetics (non-thermal SZ effect)– radial velocities of clusters (kinematic effect)– transverse velocities of clusters (polarization effect)– detections of gas in in-falling filaments
• Many new SZ facilities will come on-line in the next 5 – 10 years
Mark Birkinshaw, U. Bristol 57
Nanjing, March 2003
Attributes of SZ effect
TRJ is a redshift-independent function of cluster thermal energy, it is a calorimeter
TRJ has a strong association with rich clusters of galaxies, it is a mass finder
TRJ contains a weak redshift-independent kinematic effect, it is a radial speedometer
TRJ has polarization with potentially more uses, but signal is tiny