The recent results from the Herschel ATLAS survey can be summarized as follows:
1) The Herschel ATLAS survey has detected over 100,000 far-infrared galaxies out to redshift z=3 using Herschel's PACS and SPIRE instruments.
2) Studies of the far-infrared properties and clustering of these galaxies are providing insights into the evolution of dust in galaxies and galaxy formation over cosmic time.
3) The large area of the Herschel ATLAS survey has also led to the detection of strongly lensed high-redshift galaxies and fluctuations in the far-infrared background.
3. The Cosmic IR background
• Contains as much energy as
the optical / UV background
• Half the energy emitted by
stars and AGN since the Big
Bang has been absorbed by
dust and re-emitted at longer
wavelengths
• Herschel presents the
first opportunity to study
large samples of galaxies Dole et al. 2006
selected near the peak
5. HerMES+PEP
GOODS North / HDF North
GOODS South CDFS ECDFS
Lockman wide & deep
Extended Groth Strip
Bootes
XMM/VVDS
SWIRE fields (EN1, EN2, ES1)
Spitzer-FLS
AKARI SEP
Courtesy of S. Oliver
6. The Herschel ATLAS
• The widest area extragalactic survey with
Herschel (~ 570 sq deg)
• Consortium of 150+ astronomers
worldwide led by Nottingham (Dunne)
and Cardiff (Eales)
• Covering 5 bands with PACS and SPIRE
(100 – 500 microns) in fast parallel mode
• 5 sigma sensitivities of 132, 126, 33, 36
and 45 mJy / beam from 100-500µm
• Detect ~105 sources to z~3
• SDP = 3% of data = 7000 galaxies = 16
hrs!
9. The Herschel ATLAS
Pascale et al 2010
• 250/350/500um
• no filtering
• cirrus background
• almost confused
10. The dust SED z=0
• Sensitive to cold and warm dust giving
the total mass of dust (and gas)
• At high redshift, the shape of the curve
means that galaxies don’t get much
fainter at larger distances.
• Study evolution of dusty star forming
galaxies over the past 10 billion years
of cosmic history
• The sub-mm colours of the galaxies will
give us clues to their redshifts
11. Cross-matching: the problem
• 250um:
– beam 18.1”
– positional uncertainty ~2.4”
– minimal z info
– probes dust properties
• SDSS r band:
– PSF ~1-2”
– positional uncertainty ~0.1”
– redshift & colour
information
– probes starlight/AGN
Smith et al. 2011
12. Identifying counterparts
(Smith et al. 2011)
• Likelihood ratio technique (e.g. Sutherland &
Saunders 1992)
“The ratio of the probability that two
f (r)q(m)
sources are associated to the probability LR =
that the same two sources are unrelated” n(m)
1 ⎛ −r 2 ⎞ Radial probability density –
f (r) = exp⎜
⎜ 2σ ⎟⎟ estimate from comparing
2πσ pos ⎝ pos ⎠ €
HATLAS & SDSS positions
n(m) = Probability density of possible counterparts
i.e. SDSS r band number counts
€
q(m) = Probability density of true counterparts – statistical excess
€
13. Identifying counterparts
Introduce the Reliability:
“The ratio of the probability that two
sources are associated to the
Li
Rprobability that the same two sources
i =
∑ (
are unrelated” Q0
L j + 1− ) Radial probability density –
j
estimate from comparing
Can define a catalogue of 5sigma HATLAS & SDSS positions
250um sources with R>0.8 optical
counterparts.
€
Smith et al. (2011)
14. Identifying counterparts
• LR method allows for the fact that not all 250um galaxies are detected in
Sloan r band:
• Q0 = ~63% of 250um sources have an r band counterpart in SDSS
Smith et al. (2011)
15. Identifying counterparts
• LR method allows for the fact that not all 250um galaxies are detected in
Sloan r band:
• Q0 = ~63% of 250um sources have an r band counterpart in SDSS
Smith et al. (2011)
17. Comparisons with other models
• Normalised to LFir
• Binned according to
matched luminosities
• 1sigma uncertainty
regions shown in grey
hatchings
• CE01 models too hot
for 250um selected
galaxies
Chary & Elbaz (2001) vs Smith et al. 2011
19. H-ATLAS: Evolution of dust
Dust mass varies by factor
of 5 - not T
High z SMGs
@ z~2.5(Dunne 2003) T=25K
Dunne et al. 2011
20. H-ATLAS: Environments of dusty galaxies
Herschel sources in and around galaxy clusters
Coppin et al. 2011
Excess of far-infrared
sources towards the
centre of galaxy clusters
in the local Universe
21. H-ATLAS: Environments of dusty galaxies
Herschel sources in and around galaxy clusters
Burton, MJJ, et al. in prep.
Find a tendency for far-IR bright
galaxies to reside in less dense
environments that a matched
sample of non-far-IR galaxies
Far-IR
bright
sources Optical
sources
Suggests that gas is stripped
out of galaxies in dense
environments, thus hindering
star-formation activity
26. H-ATLAS: lenses in the SDP field
Lens subtraction @ F160W
Flux @ 1.6 µm ~ 10 µJy
CREDITS: Rosalind Hopwood
27. H-ATLAS: lenses in the SDP field
To extract the maximum amount of
science from these lenses, accurate
redshifts of both the lens and the
lensed source are required.
SALT is going to be the leading
telescope to obtain accurate
redshifts of the lenses in the
southern hemisphere
(PI Leeuw).
Redshifts for the lensed sources
requires mm-wavelength
observations of redshift CO. ALMA
and ATCA will do this in the southern
hemisphere.
30. Lensing in HerMES
Evidence for lensing induced
cross-correlations between
background (high-z) far-IR
sources and foreground (low-z)
optical galaxies
Wang et al. 2011
32. H-ATLAS: Galaxy Clustering
van Kampen et al.
submitted
Clustering as a
function of z by
combining H-ATLAS
with GAMA
33. HerMES: Fluctuation Analysis
Amblard et al.
2011, Nature
Brightness
fluctuation
analysis of two
HerMES fields
H-ATLAS
fluctuation
analysis to follow
this year, over ~30
degree scale!
34. H-ATLAS: AGN-star formation
One of the key
unknowns in
astrophysics is how
Density of galaxies /magnitude
active galactic nuclei
influence the formation
and evolution of
galaxies.
Luminosity
Benson et al. (2003)
35. H-ATLAS: AGN-star formation
One of the key Supernovae
unknowns in
2 mechanisms
astrophysics is how
Density of galaxies /magnitude
proposed to stop
active galactic nuclei gas cooling to
influence the formation form stars
and evolution of
galaxies.
Feedback is not Active Galaxies
understood in
models of galaxy
formation.
Luminosity
Benson et al. 2003
36. H-ATLAS: BAL QSOs and unification
Cao Orjales, Stevens, MJJ et al.,
in prep
Long standing issue as to
whether BAL QSOs are an
early stage in QSO evolution
when the outflow terminates a
period of star formation, or just
a simple orientation effect
37. H-ATLAS: BAL QSOs and unification
Cao Orjales, Stevens, MJJ et al.,
in prep
Long standing issue as to
whether BAL QSOs are an
early stage in QSO evolution
when the outflow terminates a
period of star formation, or just
a simple orientation effect
40. H-ATLAS: AGN-star formation
With the larger sample we
see a higher star-
formation rate associated
with more powerful radio
galaxies.
In line with current views
that powerful AGN are
fueled by the influx of cold
gas via galaxy mergers,
whereas lower power
radio sources are fueled
by the hotter ICM
Virdee, Hardcastle, MJJ, et al. in prep.
Hardcastle, Ching, MJJ et al. in prep.
41. H-ATLAS: AGN-star formation
One of the key unknowns
is accurate redshifts at
high-z and optical
emission-line
classification of AGN and
star-forming galaxies
SALT observations are
going to address this
issue (PI MJJ)
Virdee, Hardcastle, MJJ, et al. in prep.
Hardcastle, Ching, MJJ et al. in prep.
42. H-ATLAS: Far-IR—radio correlation
• The far-infrared—radio
correlation is key to
using future radio
surveys to measure the
star-formation history
of the Universe
• FIRC looks to be very
similar at low and high
redshift
• Puzzling - as would
expect evolution! Jarvis et al. 2010, MNRAS, 409, 92
43. H-ATLAS: Far-IR—radio correlation
• The far-infrared—radio
correlation is key to
using future radio
surveys to measure the
star-formation history
of the Universe
• FIRC looks to be very
similar at low and high
redshift
• Puzzling - as would
expect evolution! Jarvis et al. 2010, MNRAS, 409, 92
44. The new generation of radio surveys
(a factor of ~10 shallower than LOFAR deep field data and
100 times shallower than MIGHTEE Tier 3)
McAlpine & MJJ in
prep.
10 arcmin
45. The likelihood ratio on the new radio surveys
Resolution does matter
EVLA B-array in continuum radio
surveys for X-matching.
Key to almost all science!
ASKAP-EMU
WODAN MeerKAT will excel at
this compared to ASKAP
and APERTIF!
Currently extending to
fainter fluxes using
COSMOS data.
McAlpine, Smith, MJJ, Bonfield in prep.
46. The likelihood ratio on the new radio surveys
Depth of optical/nearIR
EVLA B-array data also crucial!
Again the MeerKAT-
K=22.6 K=20 MIGHTEE deep fields
ASKAP-EMU will have the best optical/
WODAN near-IR data available!
McAlpine, Smith, MJJ, Bonfield in prep.
47. The likelihood ratio on the new radio surveys
Depth of optical/nearIR
EVLA B-array data also crucial!
Again the MeerKAT-
K=22.6 K=20 MIGHTEE deep fields
ASKAP-EMU will have the best optical/
WODAN near-IR data available!
Redshifts are also
important for science
exploitation.
SALT-MOS observations
McAlpine, Smith, MJJ, Bonfield in prep. will provide these (PI
McAlpine)
48. Radio surveys with SKA precursors
Constraints on the evolution of star-forming galaxies
50. The link to cosmology
Raccanelli et al. (2011) present several predictions of the constraints
that can be obtained on modified gravity and the cosmology using the
new generation of wide-area radio continuum surveys.
51. The link to cosmology
Cosmology with the radio continuum surveys requires information from
most of the science I have presented.
The redshift distribution of radio
sources is fundamental to many
tests, such as ISW, lensing etc
The new surveys will be dominated
by star-forming galaxies and low
luminosity AGN.
We know the least about the redshift
evolution of these objects!
Herschel gives us information on the
evolution of the SFGs
Wilman, MJJ et al. 2010
Raccenelli et al. 2011 Nikhita & Kim both working on this
52. The link to cosmology
Cosmology with the radio continuum surveys requires information from
most of the science I have presented.
The evolution of bias is also key.
This is one of the most uncertain
factors in the prediction
presented in Raccanelli et al.
(2011)
Using GAMA+FIRST and SDSS-
Stripe82+EVLA data we can pin
this down to z~0.7 (Lindsay, MJJ
& Percival in prep) Wilman, Miller, MJJ et al. 2008
Raccanelli et al. 2011
53. The link to cosmology
Cosmology with the radio continuum surveys requires information from
most of the science I have presented.
The evolution of bias is also key.
This is one of the most uncertain
factors in the prediction
presented in Raccanelli et al.
(2011)
Using GAMA+FIRST and SDSS-
Stripe82+EVLA data we can pin
this down to z~0.7 (Lindsay, MJJ
& Percival in prep) Wilman, Miller, MJJ et al. 2008
Raccanelli et al. 2011
But for SFGs and starbursts can use the measurements from Herschel surveys
54. Summary
• Herschel is providing new and important insights into the evolution
of galaxies, from the star-formation history of the Universe, the
evolution of dust, the influence of AGN activity etc.
• Over the next year or so, Herschel will also be working in pinning
down the shape of dark matter haloes through strong lensing,
magnification bias over ~500 sq.deg and clustering of starburst
galaxies at z~2.
• We are using the techniques developed for Herschel and the
science results from Herschel to input into the design and
implementation of the new generation of radio continuum surveys.
• All of this information is key for our understanding of both galaxy
evolution and cosmology