J'L' Puget IAS, Orsay

1
J.L. Puget(IAS, Orsay)
Planck CMB and foregrounds
2
Planck vs WMAPCMB work vs other science

• Planck will go to high l (2500)
• Planck will be able to use polarization much
  better because of its higher sensitivity
• This leads to goals for the Planck component
  separation for CMB studies
• get a very good understanding of polarized
  foregrounds
• Special attention to point sources (both radio
  and IR galaxies)
• Planck data will allow many non CMB science
  projects which will need to develop specific
  component separation software
• With a better understanding of the astrophysics
  involved
• With much better ancillary data in some areas of
  the sky giving a better separation and a check of
  all sky methods

3
FOREGROUNDS intensity
12 cirrus
1 CIB
1 cirrus
4
Foregrounds High frequencies
5
Foregrounds low frequencies
6
High frequency component separation

• Temperature at lgt100 CMB and SZ have very
  specific SEDs very different from IS dust and CIB
  (methods of component separation are very
  efficient
• The main problem is interstellar dust/CIB
  separation (similar SED)
• The CIB fluctuations DOMINATES over the galactic
  dust emission in this l range
• Zodiacal cloud emission very smooth empirical
  removal of large scale emission needed (zodi
  bands detected by IRAS)

7
Spectre CIB/cirrus
CIB
Cirrus pour 1020 at/cm2
8
(No Transcript)
9
CIB - Poisson contribution dominates at high
l - Correlated part is strong (max at
l1000) Cirrus dominates at low l
10
High latitude Cirrus emission

• Cirrus fluctuations is well represented by a
  power low
• P0 scales as l2 for low brightness cirrus thus
  s/I is constant for brightness lower than 10
  MJy/sr
• HI 21cm data is a good tracer of a large fraction
  of the diffuse ISM
• Low velocity gas (in the galactic disk) has
  different dust content and temperature than
  intermediate and high velocity clouds (3 spatial
  templates)
• All sky is mapped with high quality data (control
  of far side lobes contributions) but with only 40
  arcmin resolution
• Specific regions will have better data high
  angular resol from interferometers, high
  sensitivity maps (GBT)

11
IRIS map (left) fBm center Modified fBm
right Wavelt ceof of the IRIS map Non goussianity
of the wavelet coef
12
High galactic latitude dust emission intensity
statistical propertiespower spectrum, spectral
index distribution and trend
13
High galactic latitude dust emission intensity
statistical propertiespower spectrum
normalization with respect to intensity I
contrast s / I as a function of I
14
Séparation 60-100
Couleur des cirrus 0.21 à 0.28
Sorel, PhD Thesis
15
DRAO Planck Deep Field

• 21 cm data
• DRAO (Penticton, Canada) interformetric data (1
  resolution)
• Green Bank Telescope (Virginia, USA) 100 m dish
  (9 resolution)
• High resolution 21 cm observations of a 40 square
  degrees region at high Galactic latitude
• Column density range from 1019 to 1021 cm-2
• Well suited to study optically thin and
  uniformely heated interstellar matter.
• Region where the Cosmic Infrared Background (CIB)
  will be a significant fraction of the
  Herschel-Planck emission
• Use of the HI data to separate it from the
  galactic emission
• Significant IVC and HVC emission

16
Specific regions with m

• The Planck deep field project
• HI - 21 cm observations
• DRAO 40? at 1 and 1 km/s resolutions
• GBT 100? at 9 and 0.25 km/s resolutions
• Column density 1019 -1021 cm-2
• We proposed to map the same region with PACS and
  SPIRE

17
21 cm Green Bank Telescope, DRAO Deep Field
18
1 scale
b3
b1.4
NH2.7 1020 cm-2
1.5 1020 cm-2
19
100 µK
10 µK
20
(No Transcript)
21
(No Transcript)
22
Foregrounds low frequencies
23
Low frequency component separation

• The temperature component separation is more
  complicated that the high frequency one (more
  component at similar levels)
• Synchrotron, free free, and anomalous dust
  emission are difficult to separate. The physics
  of the last is still not well established
• Synchrotron is the only polarized one
• See many talks on this very active topic showing
  that ones needs to use all informations and needs
  a model of the galactic magnetic field
• Radio sources will be a significant high l
  foreground. Ground based radio will be an
  important ancillary information

24
Ancillary dataRegions to be used for
understanding the foreground ohysics and test
component separations

• For point sources regions with deep or ultra
  deep surveys
• in the radio 5 GHz,..
• mid infrared range Spitzer
• In submillimeter-millimeter SCUBA (850 µm), IRAM
  (1.2mm)
• For interstellar foregrounds associated with gaz
  maps of tracers (HI, CO, FIR, mid IR) regions
  with
• high dynamical range (requires very good control
  of far side lobes and zero level determination)
• wide spatial frequency coverage (requires single
  dish and interferometer data at 21 cm)

25
Anomalous dust emmision(M.A. Miville-Descênes et
al)
Anomalous dust emmision dominates at relatively
high column density
26
POLARIZED FOREGROUND SEPARATION

• More complicated physics
• Only 4 components to be dealt with
• CMB
• Thermal dust emission
• Synchrotron
• Radio sources
• The 3 diffuse components have very different SEDs
• The radio sources has a large dispersion in SEDs
  and are variable

27
Planck
W. Hu
28
Polarized foregrounds (at 1 scale)
Puissance
Tucci et al, astro-ph/0411567
29
Goal less than 10 residuals of the two galactic
diffuse foregrounds
1µK
30
The turbulent component of the galactic magnetic
field

• Galactic magnetic fields can be separated into
• a large scale component due to the amplification
  of a seed field (for which there is still no good
  model) by the dynamo mechanism
• a smaller scale component created by interstellar
  turbulence
• The dynamical structure and physics of
  interstellar clouds (which controls the formation
  and contraction of gravity bound structures
  leading to star formation) is still not well
  understood.
• A number of indirect evaluation of the turbulent
  part of the field lead to an equipartition with
  some other energy densities in the ISM
• Indirect determination have been done from
  syncrotron observations (Beck et al)
• The component separation analysis gets a similar
  value (Bturb/ltBgt 0.6)

31
How to probe the turbulent galactic B field ?

• Pulsar rotation measures gives a picture of the
  large scales galactic magnetic field
• The polarization of interstellar grains seen in
  absorption on star lines of sight has been the
  first measurement of dust polarization on
  intermediate scales
• The synchrotron emission maps also the large
  scale galactic magnetic field
• Its degree of polarization depends critically on
  the ratio of the turbulent to average field (see
  M.A. Miville-Deschênes at this worshop)
• The Archeops balloon borne experiment was the
  first to measure the polarization of the diffuse
  galactic microwave dust emission
• The galactic disc shows a polarization fraction
  of about 5 except in lines of sight along
  galactic arms in agreement with optical
  polarization observed in absorption
• Some interstellar clouds were shown to have a
  high degree of polarization (10-20)

32
How to probe the turbulent galactic B field with
Planck data?

• Planck will improve on the analysis done on the
  WMAP data (better sensitivity and angular
  resolution)
• Planck can probe the turbulent field by
  statistical analysis of the polarization degree
  and orientation in nearby interstellar clouds
  (dust thermal emission)
• The alignment mechanism can be probed by
  observing the central part of dark clouds (role
  of radiation)
• The self consistency of the different tracers of
  the B field will be essential

33
The galactic plane dust polarization seen in
optical absorption
coherent alignment of grains
Polarization a few ...?
Stein 1966, ApJ, 144, 318
Mesurements needed
(Heiles)
Fosalba et al 2002, ApJ, 564, 762
34
Dust emission Polarization maps from Archeops
(353 GHz)
35
The galactic plane dust emission latitude
profiles

• Galactic profiles I, Q, U
• Fit on both profiles simultaneously

105 lt l lt 110
P () 4.6 0.7 ?(deg) 64 4
36
The galactic plane dust emission longitude
profiles
Orientation roughly orthogonal to the galactic
plane P about 3-5 (average)
37
Individual interstellar clouds dust emission
polarization
38
bgt5 Degree of polar 5 to 10
39
(No Transcript)
40

• The degree of polarization is highly variable
• The orientation is variable as well
• The use of the 353 GHz data as a tracer of the
  dust polarization will be essential (very
  different SED from the CMB)

41
POINT SOUCES

• Point sources removal/masking wand
  characterization of residuals will be essential
• Deep surveys with radio telescopes, SCUBA,
  Herschel and Spitzer will be a critical tool to
  validate all sky work or do cosmology requiring
  high accuracy removal
• All sky surveys (IRAS, AKARI, WISE) will be used
  for the all sky removal (not very deep !)

42
(No Transcript)
43
Predictions of high galactic latitude polarized
dust power spectra
44
EBL and CIB
Stacking w/ S24gt0uJy
HESS Spitzer EBL well measured
Dole et al., 2006, AA, 451, 417
45
Cosmic Infrared Background

• Galaxies Making-up the CIB
• MIR galaxies S24gt60mJy contribute to 80 of
  the FIR CIB
• Measured, completely model-independent
• Confirms Elbaz et al (2002) model-dependent
  result
• MIR galaxies are thus good tracers of galaxies
  making-up the bulk of the CIB
• We can also probe the CIB deeper for the
  S24lt60mJy galaxies
• New Estimate of the FIR CIB
• Using Stacking Analysis S24gt60mJy
• Using unresolved bkg at 24mm S24lt 60mJy
• Using 70/24 and 160/24 observed colors

Dole et al., 2006, AA, 451, 417