David Hughes

1
Submillimetre galaxies in the high-z Universe
David Hughes Project Scientist LMT Instituto
Nacional de Astrofisica, Optica y Electronica
(INAOE) Tonantzintla, Mexico
2
Extragalactic Backgrounds
Cosmic Optical Background COB Cosmic Infrared
Background CIB
CIB 24 nW m-2 sr-1
COB 23 nW m-2 sr-1
Dole et al. astroph/0603208
3
Local Ultra-luminous Infrared Galaxies (LFIR gt
1012 L?) An analog to SMGs ?
4
Plan of lectures

• The Cold Universe.
• Basic observational properties of the ISM
  and galaxies at FIR-mm wavelengths.
• Submm observations understanding the
  evolutionary history of galaxies and clusters.
• Dust grains and thermal radiation. IR-mm
  wavelength spectral energy distrubutions
• of starburst galaxies. K-corrections.
• Submm telescopes instruments. Confusion limits,
  biases survey design.
• Calculation of dust masses SFRs.
• Results (statistical properties) from
  blank-field submm galaxy surveys.
• Evolutionary history of high-redshift
  submm galaxy population.
• III. Multiwavelength follow-up (including
  molecular gas content) of high-redshift
• submm population.
• Biased galaxy surveys (towards high-z
  AGN/protoclusters/clusters)
• Sunyaev-Zeldovich effect identifying
  clusters at submm wavelengths.
• Future submm experiments and instruments.
• Future problems/projects

5
Feasibility of high-redshift sub-millimeter
galaxy surveys
6
Arp 220 (z0.018) LFIR 2x1012 L?
Feasibility of mm surveys
7
optically-thin and thick FIR-mm thermal emission
from dust
A(?) 1.086 t transmission e-t attenuation 1
e-t
if A(V) 30 (trans 9e-14) towards Galactic
Centre, then microns transmission 1
1e-4 8 0.7 100
0.97 1000 0.998
extinction
SF region A(V) Orion 8 Rho Oph
9 Serpens 10 M82 5-25
a 0.0003 0.025µm
8
Arp 220 (z0.018) LFIR 2x1012 L?
Detecting dust ( galaxy formation) at
high-redshift
redshift
850µm
.
Flux density
z0.1
.
.
.
z1
.
Current SCUBA detection limit
z3
z6
z11
850µm
radio
IR
observed wavelength
9

• Building galaxies in the early universe
• feasibility of mm surveys

Arp 220 (z0.018) LFIR 2x1012 L?
age of universe 13.7 Gyr
redshift
T12.4Gyr
.
T5.9Gyr
Flux density
T2.2Gyr
z0.1
.
.
.
.
z1
T0.9Gyr
T400Myr
z3
z6
1 mm _at_ LMT
z11
0.25 mm _at_ BLAST
radio
IR
sub-millimetre
observed wavelength
10
Balloon-borne Large-Aperture Submillimetre
Telescope (BLAST)
11
Colour-colour diagram, without errors
Recipe to derive photometric redshifts (Hughes et
al. 2002) Add photometric and calibration errors
to the maps, and extract the galaxies that can be
detected in the considered multi-wavelength
surveys under analysis (taking into account
confusion and survey depth). Calculate the
probability of identifying the colours of mock
galaxies with the colours of a real sub-mm galaxy
through a multi-variate gaussian distribution.
Colour-colour diagram, with errors
Observed colours and 1? error box
12
Bias towards the detection of colder sources?
potentially, an important bias to understand
since a cold, low-redshift galaxy has the same
colours as a hot, high-redshift galaxy
TFIR70K
TFIR10K
Flux density
z0
z0
TFIR70K
z6
radio
IR
250µm
850µm
observed wavelength
13
Bias towards the detection of colder sources?
potentially, an important bias to understand
since a cold, low-redshift galaxy has the same
colours as a hot, high-redshift galaxy
Both SEDs have same Bolometric (FIR) luminosity
TFIR10K
Flux density
TFIR70K
radio
IR
250µm
850µm
observed wavelength
14
Same flux, but a factor x2 increase in
temperature provides x10 increase in bolometric
(FIR) luminosity
Flux density
TFIR30K
TFIR15K
radio
IR
250µm
850µm
observed wavelength
15
Bias towards colder sources?
x 2 increase in temp x 10 increase in
luminosity
Blain et al. 2002
16
Evolution of morphology, number-density,
physical scale luminosity of structure
optical HST
submm JCMT
provide N(gtflux, massdistance)/unit area

• surveying the universe (i.e. counting the number
  of galaxies as a function of
• distance and brightness, and spatial
  distribution) is the most fundamental
• observation to understand the evolution of
  galaxies and clusters
• the observed (submillimetre-wavelength
  luminosity provides a measure of the
• star formation rate (since dust absorbs
  UV-optical radiation from young stars
• and themally re-radiates at longer FIR-mm
  wavelengths)

17

• SCUBA a feedhorn-coupled bolometer array
• 
  (NTD Ge thermistors)

850µm 37 pixels
450µm 91 pixels
18
15-m James Clerk Maxwell Telescope (JCMT)
Mauna Kea, Hawaii 4100 m
19
Mauna Kea, Hawaii (13,000 ft.)
1 mm
3 mm
350 um
450 um
850um
20
250µm
350µm
500µm
(Balloons)
BLAST/Herschel filters
microns
Mauna Kea transmission
21

• also Galactic confusion from cool dust (T 15
  25K)

100 ?m cirrus emission
lt 1 MJy/sr (500mJy/beam, 150 FWHM)
lt 2 MJy/sr
Schlegel, Finkbeiner, Davis 1998, ApJ, 500, 525
22
Example BLAST/Herschel confusion limits
3? cirrus confusion B100?m
250?m 500?m 1 MJy/sr 6
mJy 33 mJy 0.2 MJy/sr 0.5 mJy 3 mJy
Cirrus
Gautier etal. 1992, AJ, 103, 1313
Extragalactic
(extrapolation from 850?m HDF survey) 3?
20-30 mJy at 250 - 500 ?m 1 source/30 beams
(Condon 1974) 30 mJy at
250?m
BLAST/Herschel must survey low NH fields if
? galaxies gt 10 ?cirrus we need
galaxy confusion to dominate !
23
Holes in the Galactic cirrus (HI dust)
emission allow a view to the high-z Universe with
minimal Galactic foreground contamination e.g.
Lockman Hole, ELAIS N1 N2 with I(100um) lt
0.3 MJy/sr
24

• ACT survey region
• centred at -53 degs
• maximize the area with the
• lowest cirrus surface-brightness
• includes SEP, bright clusters

25
major SCUBA 850µm blank-field surveys
MAMBO 1.2mm
1998
2002
2004
2002 - 2005
SCUBA Half Degree Extragalactic Survey
Hughes et al. 1998, Eales et al. 1999, Lilly et
al. 1999, Scott et al. 2002, Greve et al 2004
26
UK 8 mJy SCUBA survey of the Lockman Hole ELAIS
N2 fields
(Scott et al. 2002, Fox
etal. 2002)
Lockman Hole
Hubble Deep Field (850um SCUBA survey Hughes
et al. 1998)
UK 8mJy survey 260 sq. arcmin 850um
survey 2.5mJy r.m.s. at 850um 19 sources (S/N
gt 4), 38 sources (S/N gt 3.5)
27
SHADES (2002-2005)
http//www.roe.ac.uk/ifa/shades/
Data-reduction analysis catalogues
source-counts Mortier et al. 2005, MNRAS, 363,
509 Coppin et al. 2006, MNRAS, 372, 1621

• 720 sq. arcmins in Lockman Hole and Subarus-XMM
  Deep Field mapped with SCUBA (at 850 450um)
• 120 sources detected at 850 microns (gt 3s, 1s
  2mJy)
• 20 of submm background resolved
• Extensive multi-wavelength follow-up in progress
• (search astro-ph for SHADES)
• Complete 1800 sq. arcmins mapped by AzTEC camera
  at 1.1mm in Dec 2005

28
Lockman Hole
29
Subaru-XMM Deep Field (SXDF)
30

• Sub-mm surveys (source-counts) imply a strongly
  evolving, luminous
• (LFIR gt 1012L?), optically-obscured, galaxy
  population in the high-z Universe

X 1000
No evolution
(Coppin et al. 2006)
31
IRAS FIR (60um) luminosity function
z1
z2
L(z) L (1z)3
at z1, L(z) 8 L
10
100
at z2, L(z) 27 L
HLIRG
ULIRG
Saunders etal. 1990 MNRAS, 242, 318 Soifer
etal. 1986, ApJ Lett. 303, L41
32
BLAST false-colour image - 50 hour survey
simulation 250um350um500um
Use simulations to test source-extraction
algorithms, accuracy of extracted counts,
severity of confusion, significance of
clustering, cosmic variance determine optimal
depth and area of surveys.
33
Simulated submillimetre extragalactic surveys to
guide future experiments

• luminosity evolution
• (consistent with submm counts)
• SEDs (luminosity, z)
• galaxy clustering
• multi-wavelength
• (60 - 3000um)
• variable resolution
• (SMA, GBT, LMT, BLAST,
• HERSCHEL, SIRTF)
• 1 sq. degree surveys
• (examples show 0.1 sq. degrees)

Hughes Gaztañaga 2000 (astro-ph/0004002)
34
Variance in the submm-counts due to galaxy
clustering cosmic variance
Large difference ( ?mean 3, ?max 10) in
source-counts between randomly selected regions
on the scale of HDF ( 6 sq. arcmins)
model
real SCUBA HDF survey
35
The impact of resolution source-confusion
Not just source-counts that are affected -
extraction of colours requires careful
attention in non-confused surveys
36
The effect of galaxy confusion on submm/mm
source-counts
2 mJy
20 mJy
1 mJy
0.1 mJy
Source-confusion flattens the extracted
galaxy-counts below the confusion limit and
severely affects the ability to disentangle
galaxy colours (hence constrain redshifts)
model
LMT
JCMT
BLAST
CSO
37
The effect of galaxy confusion on submm/mm
source-counts
2 mJy
20 mJy
1 mJy
0.1 mJy
model
LMT
JCMT
BLAST
CSO
SCUBA HDF - 130 beams 5 sources gt 2 mJy
Confusion level rule of thumb 1
source/30 beams (Condon 1974) SCUBA HDF area
(130 beams) /30 4.3
sources
5 observed (gt 3.5 sigma)
38
Hughes et al. 2005
39
Scientific rationale for next-generation
(sub)millimetre surveys

• improved accuracy in source-counts
• greater dynamic range in measured source-counts
• increased wavelength coverage

40
higher resolution
larger telescopes
lower confusion limit
increased sensitivity
170µm SPITZER
850µm JCMT
SPITZER
3mm LMT
increased field-of-view
larger-format cameras
faster mapping speed
improved source statistics
41
higher resolution
larger telescopes
lower confusion limit
increased sensitivity
1 sq. deg
170µm
850µm JCMT
Survey Depth
GTM confusion-limited wide-area surveys
(0.85mm 3mm)
SPITZER
3mm GTM/LMT
30 sq. deg
area
larger-format cameras
increased field-of-view
faster mapping speed
improved source statistics
42
Redshift distribution of SMGs
Radio-detected radio-dim SMGs
Radio-detected SMGs
(Aretxaga et al. 2007)
(Chapman et al. 2005)
43
simulation by Min Yun
Direct measurement of molecular gas (CO) and
spectroscopic-redshifts with the LMT
redshift-receiver
LMT

• perform efficient measurements of CO
  spectroscopic redshifts without the prior
  necessity to have accurate X-ray, optical, IR or
• radio positions.

(simulation by Yun et al. 2006)
44
Summary from submillimeter galaxy (SMG) number
counts

• The population undergoes a strong evolution,
  (1z)3,
• luminosity evolution from 0 lt z lt 2 (3)
• Break in the power-law is detected at S850µm 8
  mJy
• or at LFIR gt 6 x 1012L? (SFR 500 M?/yr )
• The co-moving volume of galaxies with SFR1000
  M?/yr
• is 10-5 Mpc-3 (zgt1.5, only weakly dependent on
  the precise z-distribution)
• 50 of (massive galaxy) SMG population at 2 lt z
  lt 3
• where are the remaining 50 ?
• 20-30 of the 850µm FIR background resolved into
  sources
• with S850µm gt 3 mJy