Transcript The history of a discovery - Institut d`Astrophysique Spatiale

45 years of developments
around the CIB
1967-2012
Jean-Loup Puget
Institut d’Astrophysique Spatiale
CNRS, Université Paris Sud
- from CIB to CIB anisotropies
- early interest and predictions for the Cosmic Backgrounds
- the search for the CIB
-evolution of
the ideas
the technology
the data analysis
leading to the observations
- what to we hope to learn today Herschel, Planck, SPT, ACT,…
01/04/2016
Cosmic backgrounds: CMB, CIB, Gamma,…
• UV redshifted in near IR cosmic background
– following the observation of the CMB (1965), work developed
on the cosmic backgrounds at other frequencies
– one key “astrophysical cosmology” question was
“where is integrated radiation from the nucleosynthesis of the
heavy elements we see today”
– Partridge and Peebles 1967
“Are young galaxies visible ?”
(ApJ 148,377) , young galaxies
700 times more luminous than
the milky way, z =10-30
Sciama and Rees 1967
(ApJ 147, 868) heavy elements
01/04/2016
detectable or not in the E.G.M
Growing interest for a FIR redshifted in submm
cosmic background (1/2)
– F. Low The Infrared-Galaxy Phenomenon 1970 (ApJ 159, L173)
– 50-300 µm observations of the Galactic center and NGC1068 and M82
(2 1012 and 6 1011 Lo) by Franck Low and H. Aumann 1970, (ApJ 162, L79)
– in 1972 the Fundamental panel of ESRO (not yet ESA !) asked me a
report on what can space observations bring to cosmology and
fundamental physics and I stressed the potential measurements of the
cosmic backgrounds
– the driving questions were two fold:
• the opacity of the universe to ultra high energy particles and photons
• learning about the early stages of galaxy formation and the history of
nucleosynthesis
01/04/2016
predictions from a report to ESRO (1972)
windows in foregrounds
Zodi
Zodi
CMB
EBL / CIB
Puget J.L. 1972,
report to the fundamental physics panel
of ESRO
CIRES project
01/04/2016
Growing interest for a FIR-submm CB (2/2)
– Setti and Woltjer: Infrared Background and Universal Cosmic Rays 1971
(Nature Phys Sc 231, 57)
– J.L. Puget, F.W. Stecker, J. Bredekamp :
Photonuclear interactions of ultrahigh
energy cosmic rays and their astrophysical
consequences1976, (ApJ 205, 638),
F.W. Stecker, J.L. Puget, G. Fazio
Cosmic far infrared background at high galactic
lattitudes 1977, (ApJ let 214, L 1)
the relative contribution to FIR
luminosity from
galactic nuclei vs star formation
remained unclear for 20 years
01/04/2016
1994
1996
01/04/2016
predictions of the anisotropies of the CIB
Not only the SED of the CIB but its angular power spectrum
was predicted to be a powerful tool to explore the star
formation history in the universe
R. Bond, B. Carr and C. Hogan 1986 (ApJ 306, 428), 1991 (MNRAS) , D. Scott and M.
White (A &A 346,1), Z. Haiman and L. Knox 2000 (ApJ)
This is still to come !
In the 80s it was not clear what was the best strategy to search
the CIB for two reasons:
- what fraction of the UV stellar energy was absorbed by dust
and re-emitted in the FIR ? (near IR or FIR CIB ?)
- COBE-DIRBE was aiming at both
- in the e.m. spectrum where are the best extragalactic
“windows” between the foregrounds ?
- solar system : zodi cloud scattering solar radiation, zodi emission
- interstellar dust emission in the FIR
01/04/2016
high latitude sky emission
21 Sept 2010 - Orsay
Hervé Dole, IAS - HDR - la nuit n'est pas noire Leinert et al., 1997, A&ASS
8
integrated emission of galaxies:
visible vs infrared output
• the heavy elements (and associated dust) abundance in the
universe is increasing with time
• a naïve view would thus predict a decreasing FIR/visible ratio
with redshift
• a few galaxies had been shown to have a large ratio but the
bulk of the galaxies were expected to have rather small one
(a few percent)
• some marginal measurements (rocket from M. Harwit’s group
and balloon observations by Frank Low near the galactic
center and Olthof and van Duinen along the galactic ridge of
the Milky Way had shown signs of the strong diffuse FIR
emission which were not well accepted
01/04/2016
balloon experiment (AGLAE) using 2K cooled
bolometers with NEP of 2 10-14 W/ rt Hz
- 40% of the luminosity of our galaxy comes out in FIR
- IRAS showed that 1/3 of the energy output of galaxies in the local
universe comes out in the FIR (Soifer,Neugebauer,Houck 1987,ARAA 25,187)
-IRAS also found a class of luminous star burst IR dominated galaxies
(IRAS 10214 was at the time the second most luminous object in the universe)
Serra, et al, 1978, (ApJ 222, L21,
),
1979 Serra et al (A&A 50,247),
Gispert, et al 1982, (A&A 106,293)
Caux et al 1984 ( A&A 137,1)
01/04/2016
SED and luminosity function of infrared galaxies
• the mid infrared emission from the ISM in our galaxy and in
other galaxies was interpreted as due to out of equilibrium
emission by very small grains and PAHs (Puget, Leger 1989 ARAA,
27,161)
• the PAHs molecules emit mostly at 6-9 and 12-15 µm
• ISO detected a large number of galaxies at 15 µm with a peak
around 1 mJy which could only be due to the redshifted 6-9 µm
emission of a strongly evolving population of luminous galaxies
(1011 Lo) at z about 1
• this was implying a surprising result that the IR/visible output
ratio was incresing with redshift
Lagache et al 2003 (MNRAS, 338,555), Lagache, Puget, Dole 2005 (ARAA 43 727)
• later SPITZER was shown to be affected by a similar
phenomenon at 24 µm (E. Le Floc’h, H. Dole)
01/04/2016
IRAS/COBE/ISO
• High |b| average: Td=17.5 K, b=2, (t/NH)250mm=10-25 cm2 (Boulanger et al 1996 A&A 312, 256)
• Interstellar Cirrus spectrum for a column density of NH=1020 cm-2
ISO
FIRAS
IRAS
DIRBE
CIB
01/04/2016
Galaxies at z>1 : SPITZER (24 mm)
– Detect LIRGs and ULIRGs up
to z~3
 PAH features passing through
the filter
First results:
- SWIRE, Guaranteed time obs.,
FLS
- L= 5 1011-1014 L @ z=1-3
- zmed ~1 (~30% z>1.5)
- Starburst. At high z,
SFR>500M /yr
01/04/2016
(e.g. Lonsdale
Caputi et al. 2005
et al. 04, Le Floc’h et al. 04, Yan et al. 04, 05, Bell et al. 05)
Predictions for Redshifts
bimodal contribution:
For S24 < 0.2 mJy: ~30% of z
> 2 galaxies
• 0.3 < z < 1 (11 to 13 µm
features) peaks at 0.5 mJy
•1.6 < z < 2.5 (6 to 9 µm
features) peaks at 0.2 mJy
•min contribution 1.1 < z <
1.6
z < 0.8
z < 0.3
z<1
z < 1.3
z<2
01/04/2016
Lagache, Dole, Puget et al, 2004, ApJS, 154
star formation history and CIB:
implications of this evolution of our understanding of galaxy evolution
• normal galaxies emit typically one half of their energy in the mid
and FIR (large dust grains + VSG-PAHs)
• the FIR/visible ratio increases from z=0 to 1
• the L* of the galaxies dominating the integrated emission increases
with redshift (by one order of magnitude from 0 to 1)
• the visible light traces the old low mass stars, the FIR traces the
UV of the massive stars absorbed by dust thus the star formation
rate
• the FIR and submm CIB traces the star formation rate and should
peak around 150 µm and the best wavelength to detect it is near the
minimum between the ISM emission and the CMB (300 µm).
• the near IR CIB could trace the very early galaxies near
reionization (Ly alpha at z=10-15) which could be detected near
01/04/2016
the minimum between the zodi scattering and emission (2-3 µm)
the detection of the CIB with FIRAS
• detecting an isotropic background is notoriously difficult;
astronomers like to do differential measurements not absolute
ones
• although on COBE the DIRBE experiment was planned for the
detection of the CIB and the FIRAS was planned to measure
the CMB and its main foreground spectra (ISM dust) it seemed
in the mid 90s the best instrument for the detection for the CIB
• absolute measurements means black bodies filling the beam to
calibrate and excellent control of straylight from far side lobes
•
01/04/2016
• the problem was not the sensitivity of FIRAS but the
component separation problem, remember that the first CMB
spectrum with 104 signal to noise shown at the ASS in
Washington was with 7 minutes of data !
• the NEP of the 4 bolometers was about 10-14 W/rt Hz
• but FIRAS was a double difference polarizing FTS
spectrometer with a very large throughput
• two schools of thinking were opposed on this question:
– use of the COBE data alone because the astrophysical data were not
controlled well enough
– use of a tracer of the diffuse ISM at high latitude (HI 21 cm survey with
very good correction for far side lobes contribution:
the Burton & Hartmann Dwingeloo survey)
- for the near IR the main problems were the faint stars and the zodiacam
scattering and emission
01/04/2016
-
• The dust/gas correlation at high Galactic latitude, F. Boulanger et al 1996,
(A&A 312, 256).
• Tentative detection of a cosmic far-infrared back-ground with COBE,
J.L. Puget et al : 1996, (A&A 308, L5), Fixsen et al 1998 (ApJ 508 523)
01/04/2016
CIB: Les fluctuations
• Distribution spectrale des galaxies IR:
z= 0.1, 0.5, 1, 3, 5, 10
• Redshift: différentes contributions aux différentes l
=> Décorrélation des fluctuations d’une l à une autre
01/04/2016
Béthermin et al., 2011, to be submitted
modèle: CIB SED et contributions
01/04/2016
H. Dole
Zodiacal scattered and emitted spectrum
Interstellar at high latitude
COBE DIRBE and FIRAS residuals after foregrounds removal
01/04/2016
quelles galaxies contribuent à l’EBL ?
x3
spectre de galaxie
infrarouge redshifté
01/04/2016
Lagache, Puget, Dole, 2005, ARAA
Cosmic Infrared Background
Dole et al 2006, A&A 451, 417
Stacking w/
S24>0uJy
From M. Griffin
01/04/2016
Extragalactic Background Light
COB/CIB
~1.0
COB~
20 to 35
COB
CIB
CIB~
24 to 27
01/04/2016
Dole et al., 2006, A&A, 451, 417
01/04/2016
The 6 fields studied by Planck for CIB
note that Planck HFI has 52 bolometers with NEP 10-17 W/rt Hz
01/04/2016
01/04/2016
evolution of galaxies:
luminosity function vs mass function
• The luminosity function shows rapidly increasing L* with redshift
(Bettermin, 2010)
• at z> 0.5 the FIR luminosity per comoving volume becomes larger
than the optical-UV one
• the FIR traces the star formation not the stellar mass already
formed
• so far up to z around 3 the FIR still dominates the radiation output
01/04/2016
Open questions
• what is the redshift at which the FIR per comoving volume start to
decrease sharply ?
• Why the star formation rate is not high within the baryonic matter
in collased DM halos of 109 Mo near the reionization redshift ?
• says differently why Kennicut’s law breaks down at high z ?
• these questions are difficult to answer because summ observations
lack sensitivity and resolution to find the very distant FIR galaxies
(except ALMA !)
• the CIB anisotropies might give us the answers before these
ALMA deep surveys are carried out: Planck, ACT, SPT,….
01/04/2016
01/04/2016
CIB opacity to TeV gamma rays
Photon-photon interaction can produce pairs if
E . e > 2 (me c2 )2
1 TeV  1eV or 1 µm
Flares have been observed in 6 Blazars by HESS.
The analysis of the absorption in their spectra gives
constraints on the CB in the optical and IR
Can be done to get an absolute upper limit on the CB
at a given frequency or assume a shape and give a
limit on its normalization.
This method is very useful when deep counts are
already available. The Gamma rays give a tool to
check if there is or not a missed component
01/04/2016
Béthermin et al., 2011, to be submitted
modèle: CIB SED et contributions
01/04/2016