Diapositiva 1

1
Understand Galaxy Evolution with IR
Surveys Comparison between ISOCAM 15-mm and
Spitzer 24-mm Source Counts as a Tool
Carlotta Gruppioni INAF OAB
La Thuile 08/03/05
2
Summary

• General Perspective
• Past ISOCAM Surveys
• Present from ISOCAM to Spitzer What can we
  learn from the comparison between 15- and 24-mm?
• Future from ISOCAM and Spitzer to Herschel and
  ALMA

3
General Perspective

• Locally stars form in giant
• molecular clouds, where optical
• and UV light is strongly
• absorbed by DUST
• Thanks to IRAS we know that galaxies forming
  stars at gt 20 M?/yr radiate the bulk of their
  luminosity above 5 mm
• LIGS 11 ? log(LIR/L?) ? 12
• ULIGS 12 ? log(LIR/L?) ? 13

4

• In the Past, when galaxies were more
• gaseous and formed the bulk of their
• present-day stars, it would be logical
• to expect to detect a large population
• of LIGs/ULIGs.
• ISO observations showed that Galaxy
• Formation could not be understood
• without accounting for dust extinction
• as a major ingredient

5
ISO SURVEYS

• Mid-IR ISOCAM (at 15 mm a LIG is visible up to
  z1.3)
• - Several Surveys were performed in
• Garanteed Time (IGTES 0.1 lt S lt 0.5 mJy Elbaz
  et al. 1999) and Open Time (ELAIS 0.5 lt S lt 150
  mJy Oliver
• et al. 2000 Rowan-Robinson et al. 2004)

6
? Source counts exhibit a strong excess of
sources below S15 2 mJy. Galaxies above this
flux density do fall within the no-evolution
region (ELAIS counts Gruppioni et al. 02)
The predicted Extragalactic Background Light at
15 mm is EBLmodels(15 mm) 3.3 nW m-2 sr-1 ?
ISOCAM resolves 73 of EBL
7
Nature of ISOCAM galaxies

• Most are star-forming galaxies, often in small
  groups and showing irregular/merging
  morphologies.
• from Shallow Surveys (i.e. ELAIS
• La Franca, Gruppioni et al. 04)
• ltL15gt 1010 L? , ltzgt 0.2
• from Deep Surveys (i.e. IGTES Elbaz et
• al. 99,01) ltL15gt 1011 L? , ltzgt ? 0.8
• ? LIG is an important phase in galaxy life a
  galaxy might experience several bursts of intense
  SF

lt z gt 0.8 HDFN
lt z gt 0.2 ELAIS-S1
8
Cosmic Evolution
Several authors have produced backwards evolution
models to reproduce source counts and redshift
distributions of ISOCAM galaxies (and IR
galaxies) i.e. Devriendt Guiderdoni 00 Dole
et al. 00 Chary Elbaz 01 Pearson 01, 05
Franceschini et al. 01, 03 Malkan Stecker
01 Xu et al. 01, 03 King Rowan-Robinson
03 Lagache, Dole Puget 03 Pozzi,
Gruppioni, Oliver et al. 04
9
Cosmic Evolution
All use a combination of luminosity and density
evolution as a function of z of the IR luminosity
function at 15 or 60 mm The major output of
these models was to show that LIGs/ULIGs were
much more common in the past than they are
today (i.e. Chary Elbaz 01 comoving IR
luminosity due to LIGs 70 times larger at z1
than today)
Pozzi et al. 04 Lagache et al.04 Pearson et al.
01 Franceschini 01 Franceschini upd.
10
The Pozzi et al. (04) 15 ?m model (a)

• Data
• ELAIS larger OT ISO survey, ?12 deg2
• (PI M. Rowan-Robinson)
• ELAIS-S1 field (?4 sq.deg) completely
  analysed
• Bologna Roma
• 15 ?m 406 sources (Lari method, Lari
  et al. 01)
• R-band 81 (330/406) of the 15 µm
• sources optically
  identified R?23
• Spectra 72 (293/406) of the 15 ?m
• source spectroscopically
  classified
• R23 at ESO2dF (La Franca
  et al 04)
• 
  

11
Relation L15/L_opt of versus L15
More luminous IR galaxy having larger L15/Lopt
Starbursts
Arp220
M82
M51
Spirals
(Pozzi et al. 04)
12
The Pozzi et al. (2004) 15 ?m model (b)
2) Luminosity Function Method
Quantitative estimators Maximum
likelihood (Marshall et al.83)
1/Vmax formalism (Schmidt 68) 4
Populations Spiral (M51), Starburst (M82), AGN1
(Elvis et al. 94) AGN2 (Circinus)
Evolution Starburst density luminosity
Normal Spiral no
evolution Agn1 AGN2
luminosity NEW !!!! L_15/L_opt to divide
starburst/spiral
13
ISOCAM extragalactic counts at 15?m
Pop. kl kd zb ----------------------
--- Spiral 0 0 0 Starb
3.5 3.8 1 Agn1 2.6 0 2 Agn2
2-2.6 0 2 (NGC1068 or Circinus) Starbursts
evolve in luminosity as (1z)3.5 and density
(1z)3.8 up to z1
14
Luminosity Function
z-distribution
15
From ISOCAM to Spitzer ...
Spitzer Telescope is now providing new insight
into the IR population contributing to the CIB
In particular with the MIPS 24-mm band, which is
starting to detect the high-z (z1.5-3.0)
analogs of the 15-mm galaxies
16
FLS Extragalactic Source Counts at 24
mmcomparison with some existing models
Galaxy evolution models
Franceschini et al. (2001) Rodighiero et al.
(2004) non-evolving normal pop, fast-evolving
type-II AGNs starbursts, evolving type-I AGNs
Lagache, Dole Puget (2003) non-evolving normal
spirals and starbursts with L density evolving
with redshift
IRAS data points (transformed to 24 µm) (Hacking
Soifer 1991 Sanders et al. 2003)
No-evolution model normalized to IRAS counts
Marleau, Fadda, Storrie-Lombardi, et al. 2004
17
Spitzer 24 ?m FLS compared to 15 mm source
counts (Marleau et al. 04)
Shaded region 24 mm counts

• Confirm existence
• of rapidly evolving
• population discovered
• by ISOCAM
• Question how to
• compare with previous ISOCAM 15 ?m counts?
• ONLY one ratio assumed to convert
• 15 ?m to 24 ?m (S24/S15?1.2)

? ELAIS range
empty symbol 15 ?m counts
Gruppioni et al.02
18
Ratio for different IR prototype populations
Considering M82 (dominant population) ratio
strongly dependent on z (because of PAHs)
S24/S15 ?2-2.5 z ?0 S24/S15 ?1.2 z
?1 S24/S15 gt 5 2ltzlt 3 The
ratio assumed by FLS team is fine for objects at
z?1 but NOT for ELAIS sources or higher z ones
19
Model predictions S24/S15 as a function of z, S24
S gt 2-3 mJy dominated by objects with
S24/S15?2-2.5 S ? 0.3 mJy dominated by
objects with S24/S15 ?1.5 S lt 0.2-0.3 mJy
dominated by objects with S24/S15 gt 2-3
-gt NEW POPULATION !
20
Contributions from different z
While galaxies with zlt1.3 dominate up to S?0.3
mJy, at lower fluxes the new high-z population
starts contributing for a significant fraction
Anyway, ?70 background qt z lt 1.5 Spitzer
deep counts resolve ?75 background
21
A bit of criticism ...
After some months Swire team (Shupe et al.
2005, in preparation)
First published 24 ?m data FLS team (Marleau et
al. 04)
22
Our Model fit to other Spitzer Bands
23
From ISOCAM to the longer ls of Spitzer and to
HERSCHEL

• For the longer ls ? Fundamental Ingredient
  Galaxy SEDs (FIR BUMP) !!!
• For starburst galaxies we need colder
  SEDs than the prototypical M82 to fit the
  observed 70 and 160 mm source counts
• Use the phenomenological evolution model
• to make predictions also in the Herschel bands
  (PACS75, 110, 170 mm
• SPIRE 250, 350, 500 mm)

24
Strength of our Model

• The evolution parameters are determined with a
  Maximum Likelihood fit of the 15-mm LF, source
  counts and redshift distributions
• The starburst/normal galaxy separation is based
  on a physical property of galaxies (L15/Lopt
  ratio) instead of being an arbitrary variable

25
Weakness of our Model

• Is based on 15-mm data only
• ? Extend the ML fit to all the MIR/FIR
  observables (i.e. source counts and redshift
  distributions) ? find the best-fitting multi-l
  solution
• Use a single SED for each galaxy population
• Better using a SED library, with SEDs changing
  (i.e. becoming colder) as function of LIR or z
  (mainly for starbursts)

26

• Separation between evolving and non-evolving
  population is sharp
• Better considering a smoother variation (i.e.
  different evolutions for different intervals of
  L15/Lopt)
• ?
• MUCH WORK TBD
• SPITZER WILL HELP !!!