Using the Sunyaev-Zel’dovich effect to probe the gas in clusters 0830 – 0910, Thursday 30 January 2003. In this talk emphasize the thermal SZ effect in total intensity. Details of the other effects are another story which would complicate matters too mcu for the present talk. Mark Birkinshaw University of Bristol

Mark Birkinshaw, U. Bristol Outline The origin of the effect SZ effect observations SZ effect science: clusters SZ effect science: cosmology The future: dedicated SZ instruments Summary Mark Birkinshaw, U. Bristol

1. The origin of the effect Clusters of galaxies contain extensive hot atmospheres Te  6 keV np 103 protons m-3 L  1 Mpc CL 0016+16 X-ray image from ROSAT PSPC (Hughes & Birkinshaw 1998; ApJ 501 1). 2 Mpc Mark Birkinshaw, U. Bristol

Inverse-Compton scatterings Cluster atmospheres scatter photons passing through them. Central iC optical depth te  np sT L  10-2 Scatterings changes the average photon frequency by a fraction   kBTe/mec2   Mark Birkinshaw, U. Bristol

Microwave background spectrum Fractional intensity change I/I = -2 (n/n) te  10-4 I  Mark Birkinshaw, U. Bristol

Thermal SZ effect Fractional intensity change in the CMB I/I = -2 (n/n) te  10-4 Effect in brightness temperature terms DTRJ = -2 Tr (n/n) te  -300 K Brightness temperature effect, DTRJ, is independent of redshift Flux density effect, DS, decreases as DA-2, not DL-2, and depends on redshift Note that the flux density effect therefore reaches a minimum at some redshift (1.61 for the currently-popular  = 0.7, m = 0.3 model) and thereafter increases. This means that the highest-flux density, highest-redshift, objects in the Universe at mm wavelengths would be clusters of galaxies if they had the same atmospheres in the past as they have today. Mark Birkinshaw, U. Bristol

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

Predicted SZ effect sky SZ sky predicted using structure formation code (few deg2, y = 0 – 10-4) CMB primordial fluctuations ignored da Silva et al. Mark Birkinshaw, U. Bristol

SZ effect and CMB power spectrum thermal SZ Sandor’s power spectrum plot – shows tSZ, KSZ, RS effects. kinematic SZ RS effect Figure from Molnar & Birkinshaw 2000 Mark Birkinshaw, U. Bristol

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

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 structural and spectral information weather RT has detected: 0016+16, A611, PC1643, A697, A773, A1413, A1423, A1704, A1722, A1914, A2111, A2218, MS1137 (z = 0.78), TexOx L20 (z  1). Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol 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

Mark Birkinshaw, U. Bristol Interferometers restricted angular dynamic range high signal/noise (long integration possible) clusters easily detectable to z  1 Figure from Carlstrom et al. 1999 Carlstrom et al., 1999 (astro-ph/9905255). Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol Interferometers restricted angular dynamic range set by baseline and antenna size good rejection of confusing radio sources available baselines Abell 665 model, VLA observation Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol 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

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

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

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

Mark Birkinshaw, U. Bristol 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

SCUBA 850 µm images: SZ effect measured in one; field too small Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol MITO MITO experiment at Testagrigia 4-channel photometer: separate components 17 arcmin FWHM Coma cluster detection Figure from De Petris et al. 2002 MITO = Millimetre and Infrared Testagrigia Observatory 2.6-m telescope Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol 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

ACBAR cluster observations ACBAR is still operating, will observe only clusters from mid September until the end of November ACBAR makes simultaneous 3-frequency images, subtract primary CMB fluctuations Observing complete luminosity-limited sample (nine REFLEX clusters, 0.05 < z < 0.35) XMM/Chandra and weak lensing data to follow up Romer et al. (see next talk) 2002 cluster observations: three of nine objects detected? Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol 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 RT has detected: 0016+16, A611, PC1643, A697, A773, A1413, A1423, A1704, A1722, A1914, A2111, A2218, MS1137 (z = 0.78), TexOx L20 (z  1). Mark Birkinshaw, U. Bristol

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

Mark Birkinshaw, U. Bristol Integrated SZ effects Total SZ flux density Thermal energy content immediately measured in redshift-independent way Virial theorem then suggests SZ flux density is direct measure of gravitational potential energy Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol Integrated SZ effects Total SZ flux density 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

Mark Birkinshaw, U. Bristol 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 ne2 dependence. SZ should show more about outer gas envelope, but need better sensitivity Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol SZ effects and lensing Weak lensing measures ellipticity field e, and so 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

Mark Birkinshaw, U. Bristol Total and gas masses Inside 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 XMM gave temperature of 9.13  0.23 keV, central electron density of (8.8  0.5)  103 m -3, abundance 0.22  0.04 solar, beta = 0.70  0.01, core radius 36.6  1.1 arcsec (geometric mean) SZ gives central SZ effect of –1.26  0.07 mK Gas is 0.13  0.02 of total mass within 250 kpc of core. CL 0016+16 with XMM Worrall & Birkinshaw 2002 Mark Birkinshaw, U. Bristol

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

Cluster radial peculiar velocity Need good SZ spectrum X-ray temperature Confused by CMB structure Sample  vz2 Three clusters so far, vz  1000 km s LaRoque et al (2002) for Abell 2163 Holzapfel et al (1998) for Abell 1689, Abell 2218 only serious attempts so far. A 2163; figure from LaRoque et al. 2002. Mark Birkinshaw, U. Bristol

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

Cluster Hubble diagram X-ray surface brightness X  ne2 Te½ L SZ effect intensity change I  ne Te L Eliminate unknown ne L  I2 X1 Te3/2  H0  X I2 Te3/2  Mark Birkinshaw, U. Bristol

Cluster distances and masses DA = 1.36  0.15 Gpc H0 = 68  8  18 km s-1 Mpc-1 Worrall & Birkinshaw 2002 XMM gave temperature of 9.13  0.23 keV, central electron density of (8.8  0.5)  103 m -3, abundance 0.22  0.04 solar, beta = 0.70  0.01, core radius 36.6  1.1 arcsec (geometric mean) SZ gives central SZ effect of –1.26  0.07 mK Gas is 0.13  0.02 of total mass within 250 kpc of core. Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol Hubble diagram poor leverage for other parameters need many clusters at z > 0.5 need reduced random errors ad hoc sample systematic errors Picture is from Carlstrom, Holder & Reese (2002; ARAA 40, 643). 38 distances for 26 clusters exist: figure shows the high s/n results. Three results: solid line H0 = 60 km s-1 Mpc-1, m = 0.3,  = 0.7, dashed line m = 0.3,  = 0, dash-dot line m = 1.0,  = 0. From Carlstrom, Holder & Reese 2002 Mark Birkinshaw, U. Bristol

need orientation-independent sample 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

Mark Birkinshaw, U. Bristol 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

Cluster counts and cosmology Cluster counts and redshift distribution provide strong constraints on 8, m, and cluster heating. dN/dz Wm=1.0 WL=0 s8=0.52 Wm=0.3 WL=0.7 s8=0.93 Wm=0.3 WL=0 s8=0.87 z Figure from Fan & Chiueh 2000 Mark Birkinshaw, U. Bristol

ACBAR blind survey Cold spots (diamonds), hot spots (crosses), radio sources (plus signs), and Abell clusters are marked. CMB5 field, filtered, pointing source blanked. Features at s/n > 4. Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol 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 b/m Figure from Carlstrom et al. 1999. Mark Birkinshaw, U. Bristol

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

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

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 SZA AMI = 10 3.7-m antennas plus existing 6 Ryle telescope antennas Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol Survey speeds OCRA will be fastest survey radiometer AMiBA will be fastest survey interferometer Frequencies complementary Mark Birkinshaw, U. Bristol

New SZ interferometers AMIBA 90 GHz SZA 30 GHz AMI 15 GHz Complementary spectral coverage Short baselines crucial for SZ detection Long baselines for radio sources AMiBA SZA AMI AMI = 10 3.7-m antennas plus existing 6 Ryle telescope antennas solid nearby high-M cluster dashed high-z low-M cluster Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol 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

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

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

OCRA Torun Observatory, Jodrell Bank, Bristol, Bologna OCRA-p Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol 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

Mark Birkinshaw, U. Bristol APEX MPI project at Chajnantor 300-element bolometer array at 870 m ideal for SZ (117-element prototype shown) Mark Birkinshaw, U. Bristol

Mark Birkinshaw, U. Bristol 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

Mark Birkinshaw, U. Bristol 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

Mark Birkinshaw, U. Bristol 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

Mark Birkinshaw, U. Bristol 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

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