Galaxies and galaxy clusters at mm wavelengths: the view from the South Pole Telescope Gil Holder.

Galaxies and galaxy clusters at mm wavelengths:

the view from the South Pole Telescope

Gil Holder

QuickTime™ and a decompressor

are needed to see this picture.

QuickTime™ and a decompressor

are needed to see this picture.QuickTime™ and a decompressorare needed to see this picture.

On one side: CMB/SZ, “fundamental physics”

on the other side: BLAST, “astrophysics”

Outline• Small scale CMB anisotropy

– Detection of secondary anisotropies– Limits on thermal SZ, kinetic SZ

Lueker et al, Hall et al (both submitted)

• Galaxies– “point sources” in SPT maps– Dusty and/or synchroton-dominated galaxies– a new class of dusty galaxies?

Vieira et al (submitted)

Zoom in on 2 mm map~ 4 deg2 of actual data

QuickTime™ and a decompressor

are needed to see this picture.

Detecting the SZ Power Spectrum

SPT Measured Points(model + Gaussian scatter)

• after removing bright sources, there is still small-scale contamination from residual sources

•Primary CMB looks right

• Poisson Point-sources as expected

10 K-arcmin point sources

guess at SZ power spectrum (8=0.8)

primary C

MB

150GHz

QuickTime™ and a decompressor

are needed to see this picture.

Detecting the SZ Power Spectrum

SPT Measured Points(model + Gaussian scatter)

10 K-arcmin point sources

Hall et al. (2009, arxiv:0912.4315): we report tSZ + kSZ + clustered point-source power of 10K2 at=150GHz and l=3000

primary C

MB

150GHz

What happened to all the thermal SZ power?

Clustering of Point Sources

• Radio and IR/submm sources presumably trace the large scale matter fluctuations

• Back of the envelope:– Power spectrum

contribution: mean T2 x projected clustering amplitude

– Arcminute scales: few Mpc has clustering ~1 in 3D, divide by number of independent cells along line of sight => 1e-3

QuickTime™ and a decompressor

are needed to see this picture.

The importance of multiple frequencies

Frequency scaling of Dusty Galaxy Background

9Scaling of Poisson power with frequency

QuickTime™ and a decompressor

are needed to see this picture.

Hall et al 2010

10

First detection of clustered point source power from CIB sources in the mm bands

Hall et al 2010

Frequency scaling of Dusty Galaxy Backgrounds

Single-SED model assumes all galaxies have same rest-frame properties (T=34 K, =2) spread over a broad range in redshift (peaking at z=2)

Removing dusty galaxies

• Models suggest that nearly all of the residual power (both Poisson and clustered) is from high-z dusty galaxies

• To remove these, SPT constructed a map that is T150-xT220– Subtraction factor is tuned to minimize small scale power [no

noise bias in power spectrum]

– New map has all of tSZ, but has subtracted some fraction of CMB+kSZ

– Subtraction is imperfect: unknown and non-unique spectral behaviour of dusty sources

Temperature scaling

• Radio sources look a lot like CMB

• Dusty galaxies are much brighter at higher frequencies Frequency

(GHz)

Dusty galaxies (z=0,2,5)

Radio galaxies

dTcmb

Power Spectrum: Dusty Galaxy Contributions Largely Subtracted

Direct subtraction of (220 GHz map)/3 from 150 GHz map to

remove dusty galaxies

Power Spectrum: Dusty Galaxy Contributions Subtracted

Best fit 150 GHz power:

tSZ+0.46*kSZ=4.2 1.5 uK2

tSZ

kSZ

QuickTime™ and a decompressor

are needed to see this picture.

What does the low SZ power mean?

• If we assume the fiducial tSZ model is correct (and some fiducial kSZ template), we find 8 = 0.746 ± 0.017

Compare to 8 = 0.794 ± 0.027 for WMAP5 + ACBAR + QUaD

• Allowing the best estimate (from theory considerations) 50% uncertainty on tSZ model amplitude gives 8 = 0.779 ± 0.025

Not Much Room for kSZ• Thermal SZ alone is

already a bit low• Using X-ray-based

profiles and WMAP chains, cl 8

16

• covariances between parameters conspire, in particular , h

• SPT analysis (Lueker et al.) used semi-analytic gas model, with cl 8

11 in CMB

chains

Best fit 150 GHz power:

tSZ+0.46*kSZ=4.2 ±1.5 uK2

Where does SZ Power Come From?

• Broad range in z

• Extends to low mass (relative to SPT SZ-detected clusters)

QuickTime™ and a decompressor

are needed to see this picture.

From Shaw et al 09

Rough SPT mass limit for detection

Non-Gaussianity of statistics

d2cl /(dlnM dz) at ell~2500

SPT Galaxies at 150 & 220 GHz

Distribution of Spectral Indices

sources cleanly separate into two populations

synchrotron

dust

AGN counts

AGN counts as predicted

at high flux

source counts

dominated by

synchrotron-

dominated

sources

at the low flux end of the 1.4 mm band where dusty

sources become

dominant

150 GHz

220 GHz

dust source counts

red = Lagacheblue = Pearson

BCS image of a dusty SPT source with an IRAS counterpart

r band 5σ = 24.65 AB mag

i band 5σ = 24.35 AB mag

S1.4 = 14 mJyS2.0 = 8 mJy

BCS image of a dusty SPT source without any counterpart

r band 5σ = 24.65 AB mag

i band 5σ = 24.35 AB mag

S1.4 = 17 mJyS2.0 = 5 mJy

dust source countsWITH IRAS SOURCES REMOVED

red = Lagacheblue = Pearson

Comparison to Negrello et al. 2007

Current Followup Campaign

• ATCA: 3 mm 4 detections after two weeks of observing

• SMA: 1.4 mm detections for ~ 10 or our most northern sources

• Spitzer: 3.6 + 4.5 um observations down to 1 uJy (data taken, fully reduced)

• NOAO SOAR 4m: R,I,J,K observations done, data being reduced

• Gemini-S: spectroscopy for z>4 candidates in queue

• BCS griz ~ 60 square degrees in the can

Interferometric Follow-up

• ATCA at 90 GHz– Hard frequency

– Good location (Australia)

• SMA at 220 GHz– Easy frequency– Terrible location (Mauna Kea)

QuickTime™ and a decompressor

are needed to see this picture.

QuickTime™ and a decompressor

are needed to see this picture.

SMG 10S1.4mm=21 mJy

SPT,SMA,IRAC,Gemini

SMG 03S1.4mm=37 mJy

BCS

Summary• SPT has measured the small-scale CMB

power spectrum, detecting secondaries– SZ power may be a bit low (or matter power spectrum is a

bit low)Hall et al, Lueker et al (submitted)

• interesting population of galaxies at mm wavelengths– Either nearby galaxies with very cold dust or extremely

bright high-z galaxies– Lensed?– Discovered because of large area (~100 deg^2) searched

compared to existing catalogsVieira et al (submitted)

IR/Submm Source Clustering

• Mean Tcmb~104 uK at 500 um (FIRAS)

• Clustering amplitude 10-3

• => few 105 uK2

• BLAST: 106 uK2

• Mean Tcmb~50 uK at 150 GHz (FIRAS, number counts)

• => few uK2

• We do actually have a clustering model

BLAST: Viero et al 2009

• Extrapolate ARCADE results to 150 GHz: 5uK

• Extrapolate source models: less than 1 uK

– => << 1uK2 clustering power at 150 GHz

• Aside: ARCADE extrapolation to 30 GHz: T~200 uK

• 30 GHz clustering power could be >50 uK2

• However: widely agree that ARCADE results are hard to reconcile with known populations

QuickTime™ and a decompressor

are needed to see this picture.

Radio Source Clustering

Fixsen et al 2009