the Towards Arrival of SZ Galaxy Cluster Surveys: The Arcminute Microkelvin Imager Small Array (AMI-SA)
Tak Kaneko
Astrophysics Group (MRAO), Cavendish Laboratory, University of Cambridge, UK SPIE Astronomical Telescopes & Instrumentation, Orlando. 31 May 2006 [6267-152]
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
1/21
Structure Formation
Structure formation & cluster evolution scenarios have been tested with data from the local universe. Need high-redshift data to test the models.
http://uchicago.edu/lss/filaments.html
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
2/21
Clusters of Galaxies (Perseus)
Optical view (Jim Misti)
X-ray (Fabian et al 2003)
The collapse of the cluster heats the gas up to 107 108 K. ROSAT is the only recent X-ray telescope capable of conducting a large-scale survey.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
3/21
Sunyaev-Zeldovich Effect
Carlstrom et al 2002
Some of the CMB photons gain energy from the energetic electrons in the cluster gas by inverse-Compton scattering. In the Rayleigh-Jeans regime, the SZ effect appears as a temperature decrement in the CMB.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
4/21
Sunyaev-Zeldovich Effect
SZ contours are 0.75 K and X-ray scales are the same.
SZ and X-ray maps (insert). Mohr 2002 Z = 0.17 Z = 0.54 Z = 0.83
SZ Temperature decrement is redshift-independent. SZ effect does not suffer from cosmological dimming in the same way that optical & X-ray surveys do.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
5/21
Next-Generation Dedicated SZ Instruments
Interferometers: SZA, AMI, AMiBA Low-cost and is a mature technology Good handle on systematics Single dish bolometers: APEX-SZ, ACT, Planck, SPT. Wideband and good sensitivity Can operate at high-frequency, where synchrotron emission from radio sources is less.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
6/21
Ryle Telescope SZ Program
Abell 2218: Jones et al 1993
1984: First widely recognised SZ detection. 1993: First image of the SZ effect by the Ryle Telescope. Also, first interferometric detection. But each detection typically takes a few weeks of observation.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
7/21
*
Why the Ryle isnt Optimum
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
8/21
Why the Ryle isnt Optimum
4 sensitivity in a 200-hour observation of a 0.1 deg2 field.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
9/21
Why the Ryle isnt Optimum
Interferometers are Insensitive to the worst effects of the primordial CMB!
4 sensitivity in a 2-hour observation of a 0.1 deg2 field.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
10/21
Why the Ryle isnt Optimum
5 mJy source
4 sensitivity in a 2-hour observation of a 0.1 deg2 field.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
11/21
Why the Ryle isnt Optimum
5 mJy source
Residual Confusion
4 sensitivity in a 2-hour observation of a 0.1 deg2 field.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
12/21
Source Subtraction in Action
A1914 on the long baselines of the Ryle and on the short baselines (Jones et al. 2005).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
13/21
The AMI Team
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
14/21
AMI Small Array
10 x 3.7 m dishes (18.6 arcmin FOV) Baselines 420 m (3.417 arcmin). Sea-level site near Cambridge, UK RF 1218 GHz Bandwidth = 6 GHz IF 612 GHz Tsys = 25 K
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
15/21
First Results: Abell 1914
AMI Collaboration, MNRAS 2006
High-significance (17) detection of the SZ effect. Also detected the spectrum of the SZ effect.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
16/21
Early Targets Program
A263 at z = 0.3, 8 hrs (Natasha Hurley-Walker)
0717+374 at z = 0.545, 6 hrs (Jon Zwart)
Currently observing known clusters to constrain cluster scaling relations (eg. Mass vs Temperature). This is needed to calibrate our cluster survey.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
17/21
Ryle Move (aka AMI-LA)
The 3 outlying antennas were moved to form a 2D array (Autumn 2004). Currently being upgraded with the same frontend amplifiers and backend electronics as AMI-SA.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
18/21
Status & Timeline
Currently (May 2006):
AMI-SA operational Engaged in the Early Targets program of observing known clusters (8 detected).
Midlate summer 2006:
Shallow Cluster Survey with AMI-SA using the source list from the VSA fields.
Beginning 2007:
AMI-LA operational. Start Deep Cluster Survey.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
19/21
Conclusions
Cluster surveys with the SZ effect is a powerful means of finding high-redshift clusters. Their high resolution allows interferometers to spatially filter out radio sources very effectively. Interferometers will lead the first wave of SZ surveys (mature technology, lowcost, low systematics).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
20/21
Thank You
SPIE Proceedings 6267-152 www.mrao.cam.ac.uk/telescopes/ami
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
21/21
Bonus Slides
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
22/21
SZ Effect: Equation
In the RayleighJeans Regime:
S T k Z = 2 T B nT l d e e 2 TM m c CB e
The SZ temperature is a line-of-sight pressure integral of electron density (ne) and gas temperature (Te). The SZ temperature decrement is redshiftindependent: When the CMB photons interacted with the cluster, TCMB(z) was hotter and this compensates for cosmological dimming.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
23/21
Thermal & Kinetic SZ Effect
Thermal and kinetic SZ effect for a cluster with a peculiar velocity of 500 km s1 (Carlstrom et al. 2002).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
24/21
Costs
AMI is a highly cost-effective project. The hardware costs for the AMI-SA and upgrade of AMI-LA is 850K ($1.6M) The total cost, including salaries, etc is 2M ($3.7M). This was made possible because we were able to re-use the Ryle Telescope. Building Ryle the Telescope from scratch would have cost 10M ($19M).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
25/21
Cluster Counts
Cluster source count and cumulative count (courtesy of Marko Velic).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
26/21
AMI Survey Strategy
Shallow Survey:
Survey 100 deg2 with short (1 hr) observation per AMI-SA pointing. Should start to detect clusters ~1015 Msun.
Deep Survey:
Survey 10 deg2 with long (50 hr) observations per AMI-SA pointing. Should detect clusters ~1014 Msun.
Other Survey Options:
XMM/Subaru LSS Survey Fields, Lockman Hole, Medium depth survey
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
27/21
How Many Clusters will AMI Detect?
Shallow Per pointing 6 hrs Clusters per deg2 0.6 Area covered in a 74 deg2 1 year survey Clusters found in 47 a 1 year survey Medium 24 hrs 2.4 18 deg2 45 Deep 48 hrs 3.9 9 deg2 35
From model 2 in Tom Culverhouses Thesis 2006
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
28/21
Selection Function
6hr and 24hr integration selection function for AMI (model 2 from Tom Culverhouses Thesis 2006)
Selection Function of AMI (top of the scale is ~1 x 1015 Msun).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
29/21
AMI-LA Specs
8 x 12.8 antennas (5.3 arcmin FOV). 18120 m baselines (30 arcsec resol) RF: 1218 GHz Bandwidth: 6 GHz Tsys = 25K Flux sensitivity = 1.6 mJy s-1/2
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
30/21
Science with AMI 1
Cluster redshift distribution Cluster structure and evolution Cluster abundance relation (deep SZ surveys will probe the epoch of cluster formation) Volume-redshift relation Measure dn(M, z) / dM to constrain cosmology. Cluster scaling relations and any evolution with redshift. Any evolution in baryon fraction (clusters should be a good sample of matter in the Universe).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
31/21
Science with AMI 2
Excess CMB power at higher multipoles 20003000. Cosmic Strings (Kaiser-Stebbins effect)?, OstrikerVishniac effect? Source count at 1218 GHz, evolution of radio sources in cluster environment. Cosmological parameters Mostly better constrained by CMB + SN but provides independent tests. Estimate the normalisation of matter power spectrum on scales of 8h1 Mpc (sigma-8). An independent estimate of the Hubbles constant (needs X-ray data).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
32/21
Science with AMI 3
Optical and X-ray follow-up needed. After commissioning and surveys, 25% of the time will be available to external observers.
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
33/21
Excess CMB at l > 2000
Left: ACBAR and CBI excess at high multipoles (Kuo et al. 2004). Right: Expected contribution of the SZ effect to the CMB power spectrum (Cooray et al 2004).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
34/21
Sigma-8
The normalisation of the power spectrum on scales of 8h 1 Mpc. This will affect the level of contribution of the SZ effect to the CMB. Cooray et al. 2004
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
35/21
Cosmological Parameters
Left: Cluster surveys can be an independent check for DE density and density of matter (Vikhlinin et al. 2004). Right: Can help constrain the mass density and equation of state (Haiman et al. 2004).
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
36/21
Sources in Clusters
There are about 14 times more sources in the cluster core (r < 0.25 Mpc) than in the field (Kaneko 2006)
Tak Kaneko, Astrophysics, Cavendish, University of Cambridge
37/21
Confusion Noise
With evolution
Confusion noise drops off with redshift, flattening out above z &g...