EARA Herschel Workshop @ IAP

1
A FIR Sub-mm View onGalaxy Formation
EvolutionTaking Stock of Spitzer
SCUBAPlanning for Herschel SCUBA2

• Mattia Vaccari - University of Padova
• A. Franceschini G. Rodighiero
• Thanks also to Carol, Jim, Matt, Michael, Seb
  all the
• ISOCAM, ELAIS, SHADES, SWIRE, SPIRE SCUBA2 folks

2
The Cosmic IR Background RadiationResolved Into
Sources
The integrated extragalactic background light in
the far-infrared and sub-millimeter region of the
spectrum is approximately equal to the integrated
background light in the optical and UV part of
the spectrum. To develop a complete
understanding of galaxy formation, this
background light must be resolved into galaxies
and their properties must be characterized.
3
Modeling the FIR Sub-mm Era

• Long-wavelength observational astronomy is
  arguably experiencing its golden era, with new
  facilities (telescopes instruments) coming
  on-line in an ever accelerating fashion
• The sources making up the very bright and
  violently evolving universe discovered by IRAS
  first and the CIRB light later detected by COBE
  are being (and will increasingly be) detected and
  characterized at a pace largely reminiscent of
  optical astronomy
• The paradigms of semi-analytical models will need
  time to properly adjust to and make sense of new
  results
• Simpler and more flexible phenomenological models
  can be of great help in providing the first clues
  about the nature of such sources as well as in
  timely planning development and use of new
  facilities

4
The AF08 Model

• The AF01 phenomenological model (Franceschini et
  al. 2001) was rather successful in explaining
  exploring ISO results
• Spitzer SCUBA data (re)-analyses, however,
  called for a revamp
• Through a simple backward evolution approach,
  AF08 describes available observables (number
  counts, z-distributions, L-functions, integrated
  CIRB levels) in terms of number and luminosity
  evolution of four populations
• slowly or non-evolving disk galaxies blue dotted
  lines
• type-1 AGNs evolving as shown by UV and X-ray
  selected Quasars Seyferts green
  long-short-dashed lines
• moderate-luminosity starbursts with peak emission
  at z 1 cyan dot-dashed lines
• ultra-luminous starbursts with peak evolution
  between z 2 and z 4 red long-dashed lines

5
Modeling Benchmarks

• Spitzer MIPS SCUBA Number Counts largely
  available
• Spitzer MIPS Deep Narrow GOODS GTO GO
• Spitzer MIPS Wide Shallow SWIRE GTO GO
• SCUBA Deep Lensing Surveys
• SCUBA Wide SHADES
• Spec Phot Redshift Distributions in the works
• Complete Redshift Distributions of 24 ( 70) ?m
  sources
• Redshift Distributions of SCUBA sources related
  headaches
• Ultimate Spitzer Luminosity Functions within
  reach
• Extras AzTEC_at_JCMT BLAST surveys being reduced

6
Spitzer MIPS Counts 24 ?m

• SWIRE (Shupe 08)
• GTO (Chary 04)
• but also (not shown)
• GTO (Papovich 04)
• FLS (Fadda 06)
• Remarkable agreement!
• Most stringent constraint
• provided by Spitzer to date

7
Spitzer MIPS-Ge Counts 70 160 ?m
SWIRE (Vaccari in prep) FLS (Frayer 2006a) GTO
GO (Frayer 2006b)

• SWIRE (Vaccari in prep)
• FLS (Frayer 2006a)

Reasonable agreement but comparison is hampered
by poor areal coverage and sensitivity at 70 and
160 ?m respectively
8
SCUBA 850 ?m Counts
SHADES (Coppin 2006) SCUBA Misc (Scott
2006) SCUBA Lensing (Knudsen 2008)

• SHADES (Coppin 2006)

Without ad hoc adjustments, we generally
overpredict bright counts by some amount and
perhaps more importantly miss the exact locus of
their downturn
9
First Hints At Herschel Counts450 ?m SCUBA Counts

• Smail et al. 2002

10
First Hints At Herschel Counts350 ?m CSO Counts

• Khan et al. 2007
• Fainter data point
• from CSO
• Brighter data points
• from rescaling of
• Smail et al. 2002

11
Redshift Distribution of 24 ?m Sources

• Available in the literature are various bits of
  heterogeneous information reporting an awkward
  and sometimes undocumented mix of spec phot
  redshifts
• ALSO
• Largely unreleased material!
• Understandable but a little annoying from the
  modelers baby food prone point of view

In order to improve this situation, we turned to
Spitzers Legacy SWIRE Public (Dec 05)
Proprietary databases handily available BUT Lack
of spec and/or deep phot coverage, new spec
campaigns crushed by TACs or bad
weather GOODS LOTS of ancillary data and,
crucially, redshifts BUT Only 24 ?m public
catalogue (in GOODS-S) released back in 2004

12
Redshift Distribution of 24 ?m Sources

• Available in the literature are various bits of
  heterogeneous information reporting an awkward
  and sometimes undocumented mix of spec phot
  redshifts
• ALSO
• Largely unreleased material!
• Understandable but a little annoying from the
  modelers baby food prone point of view

In order to improve this situation, we turned to
Spitzers Legacy SWIRE Public (Dec 05)
Proprietary databases handily available BUT Lack
of spec and/or deep phot coverage, new spec
campaigns crushed by TACs or bad
weather GOODS LOTS of ancillary data and,
crucially, redshifts BUT Only 24 ?m public
catalogue (in GOODS-S) released back in 2004

NRAO - Socorro 21 Nov 2007
A FIR Sub-mm View Mattia Vaccari
13
z-dist of GOODS-S 24 ?m Sources

• The Field
• 948 objects within GOODS-S MIPS 24 ?m Public
  Catalogue
• LARGELY reliable complete down to 80 ?Jy
• 666 within GOODS-MUSIC 168 arcmin2 area
• ACS z ISAAC K sel.
• also imaged in UVIRAC
• lots of spectroscopic photometric z-info

14
ID Redshift Information

• Full Likelihood Ratio (LR) Analysis based on
  GOODS-MUSIC catalogue
• MIPS fluxes vs IRAC fluxes of zK-selected
  sources
• Archival search through ESO, COMBO-17 GMASS
  databases
• Our own SED Fitting Flavor
• revamped HyperZ better exploring pars space by
  Giulia Rodighiero
• SWIRE empirical SED templates by Mari Polletta
• Out of 666 sources, 650 (97.5) have robust
  redshifts
• 253 Spec-Z (40) 397 Phot-Z (60)
• 499 via GOODS-MUSIC from LR analysis
• 59 via GOODS-MUSIC through Visual Checks
• 57 via COMBO-17
• 35 via case-by-case SED analysis through HyperZ
• Full Optical/NIR/IRAC ( MIPS-24!) SED info now
  available!

15
ID Redshift Information

• Full Likelihood Ratio (LR) Analysis based on
  GOODS-MUSIC catalogue
• MIPS fluxes vs IRAC fluxes of zK-selected
  sources
• Out of 666 sources, 650 (97.5) have robust
  redshifts
• 253 Spec-Z (40) 397 Phot-Z (60)
• 499 via GOODS-MUSIC from LR analysis
• 59 via GOODS-MUSIC through Visual Checks
• 57 via COMBO-17
• 35 via case-by-case SED analysis through HyperZ
• Full Optical/NIR/IRAC ( MIPS-24!) SED info
  now available!

HyperZ SED Fitting
16
Redshift Distribution of 24 ?m Sources

Overdensities Confirmed Successful Modeling No Ad
Hoc Adjustments
17
Redshift Distribution of SCUBA Sources
Not so bad
NRAO - Socorro 21 Nov 2007
A FIR Sub-mm View Mattia Vaccari
18
Z0-2.5 24 ?m Luminosity Functions

• GOODS-S GOODS-N SWIRE-VVDS Fields (Rodighiero
  in prep)
• 2000 sources with some of the best spec info
  available

The determination of redshift-dependent
Luminosity Functions require large corrections
which depend to a large extent on the adopted SED
templates, and particularly so for IR Bolometric
(8-1000 ?m) Luminosity Functions
19
Comparisons with other Models

• Aim Take stock of models on the market to
  minimize uncertainties
• Occasion Sebs SPIRE High-z Modeling Challenge
  of July 06
• In the pipeline already
• Franceschini et al. 2008
• Elbaz et al. 2008
• Xu et al. 2008
• Lagache et al. 2004
• Guiderdoni et al. 1997 (A)
• Guiderdoni et al. 1997 (E)
• Coming along
• Rowan-Robinson 2008
• Malkan 2008
• Pearson 2008
• Interested? Then drop me a line at
  mattia_at_mattiavaccari.net

20
Comparisons With Other Models

• Differences of up to an order of magnitude near
  the peak

21
Comparisons With Other Models

• Differences of up to an order of magnitude near
  the peak

22
Extragalactic Confusion

• Two methods are routinely used to evaluate EG
  confusion
• The Fluctuations approach measures rms
  background fluctuations due to unresolved sources
  and sets confusion at q (e.g. q3,4,5) times that
  level
• The Counts approach measures source counts and
  sets confusion where a maximum number of sources
  per beam (or rather a minimum number of beams per
  source, e.g. bps10,30,50) is reached
• Values provided by the two mostly depend on
  whats happening
• fainter than the confusion limit (Fluctuations)
• brighter than the confusion limit (Counts)
• Roughly speaking, the level of agreement between
  the two approaches depend on the slope of the
  counts AT the confusion limit

23
Fluctuations Criterion
24
Counts Criterion
Red lines refer to 10 bps levels
25
Extragalactic Confusion
Channel PACS1 PACS2 PACS3 SPIRE1 SPIRE2 SPIRE3
?m 70 110 170 250 350 500
Beam FWHM 4 .74 6.96 10.76 17.1 24.4 34.6
3 ? mJy 0.06803 0.8979 6.958 18.26 23.86 22.16
4 ? mJy 0.1962 2.073 12.20 27.89 34.49 31.03
5 ? mJy 0.3691 3.454 17.52 37.38 44.83 39.57
10 bps mJy 0.1100 1.263 7.090 14.00 15.23 13.23
20 bps mJy 0.3029 2.746 11.82 20.49 21.65 18.31
30 bps mJy 0.4887 4.034 15.23 25.40 26.34 21.49
40 bps mJy 0.6656 5.110 18.12 29.18 29.93 24.09
50 bps mJy 0.8350 6.107 20.45 32.47 33.05 26.18
Due to the different slope in counts, the ? vs
bps is not a one-to-one relation, ? values being
generally consistently worse than bps ones for
SPIRE with respect to PACS
26
A Pre-Launch Consensus Viewon Herschel EG
Confusion Limits
27
A Pre-Launch Consensus Viewon Herschel EG
Confusion Limits
MEAN - RMS of various models4 ? values above
are arguablybest pre-launch indication(modulo
BLAST results)
28
Simulations Spitzer 2 Herschel( keeping an eye
on SCUBA(2) )

• 24 ?m

• 110 ?m

• 350 ?m

• 450 ?m

• 850 ?m

29
What Herschel OT Key ProjectSurveys can we
expect (or aim for)?
Rich opportunities exist to extend GT KP mid-
large-area planned surveys ( 1 deg2 or larger)
to deeper levels wider areas
3 s 10 bps
20 bps
Opp. Area
10 bps
Opp. Area
3 ? 20 bps
An Herschel Survey of Surveys (SOS) www.mattiavacc
ari.net/herschel/sos
NB NOT including last-minute changes to GT but
still largely valid
30
What might be in it for you?

• A handy modeling code simulation tool
  generating
• source counts
• redshift luminosity distributions
• mock catalogues
• at a complete set of relevant FIR sub-mm
  wavelengths
• A companion tool to estimate confusion limits
  based on the model above as well as on competing
  models
• A set of simulated surveys reproducing lots of
  completed planned surveys by ISO, Spitzer,
  Herschel, Planck, Akari, SCUBA2 (and soon SPICA
  FIRI?), e.g. the Herschel GT Wedding Cake and the
  SCUBA2 Legacy Surveys
• Hopefully sometime soon publicly available
  through the www

31
(No Transcript)
32
Thanks!

• Don't clap too loudly - it's a very old world
• Tom Stoppard, Rosencrantz and Guildenstern are
  dead

Mattia Vaccari mattia_at_mattiavaccari.net www.mattia
vaccari.net