Diffuse Galactic ?-ray emission model

1
Diffuse Galactic ?-ray emission model
Igor V. Moskalenko (Stanford) with S. Digel
(SLAC) T. Porter (UCSC) O. Reimer (Stanford)
A. W. Strong (MPE)
2
Diffuse Galactic Gamma-ray Emission

• 80 of total Milky Way luminosity at HE !!!
• Tracer of CR (p, e-) interactions in the ISM
  (p0,IC,bremss)
• Study of CR species in distant locations (spectra
  intensities)
• CR acceleration (SNRs, pulsars etc.) and
  propagation
• Emission from local clouds ? local CR spectra
• CR variations, Solar modulation
• May contain signatures of exotic physics (dark
  matter etc.)
• Cosmology, SUSY, hints for accelerator
  experiments
• Background for point sources (positions, low
  latitude sources)

• Besides
• Diffuse emission from other normal galaxies
  (M31, LMC, SMC)
• Cosmic rays in other galaxies !
• Foreground in studies of the extragalactic
  diffuse emission
• Extragalactic diffuse emission (blazars ?) may
  contain signatures of exotic physics (dark
  matter, BH evaporation etc.)

Calculation requires knowledge of CR (p,e)
spectra in the entire Galaxy
3
Conventional model vs EGRET data
?0 IC Bremss EG
Conventional model consistent with local p,e
spectra exhibits the GeV excess a factor 2
We must have at least 2 diffuse emission models
with/without the excess
4a-f
4
DC2 diffuse emission model

• galprop ID 6002029RB
• Based on
• Strong,Moskalenko,Reimer, 2004, ApJ 613,962
• Strong,Moskalenko,Reimer,Digel,Diehl, 2004, AA
  422, L47
• Optimized to fit EGRET data (GeV excess CR
  spectra)
• Includes secondary electrons positrons
• Pulsar/SNR source distribution
• Gradient in X-factor (H2/CO)
• Improvements
• new HI, CO data (Digel)
• new interstellar radiation field (Porter)
• fine adjustments to reflect these new inputs
• Examples of model unconvolved and convolved with
  EGRET PSF

5
GeV excess Optimized/Reaccleration model

• Uses all sky and antiprotons gammas
• to fix the nucleon and electron spectra
• Uses antiprotons to fix
• the intensity of CR nucleons _at_ HE
• Uses gammas to adjust
• the nucleon spectrum at LE
• the intensity of the CR electrons
• (uses also synchrotron index)
• Uses EGRET data up to 100 GeV

antiprotons

• pbars
• e -flux
• ?-rays

Ek, GeV
protons
electrons
Strong etal 2004
x4
x1.8
Ek, GeV
Ek, GeV
6
Secondary e are seen in ?-rays !
Lots of new effects !
electrons
Heliosphere e/e0.2
sec.
IC
positrons
brems
Improves an agreement at LE
7
(No Transcript)
8
Distribution of interstellar gas

• Neutral interstellar medium most of the
  interstellar gas mass
• 21-cm H I 2.6-mm CO (standing for H2)
• Differential rotation of the Milky Way plus
  random motions, streaming, and internal velocity
  dispersions is largely responsible for the
  spectrum
• This is the best but far from perfect
  distance measure available

(25, 0)
CO
25
Dame et al. (1987)
G.C.
H I
Hartmann Burton (1997)
W. Keel
9
New H2 maps (S.Digel)
10
New HI maps (S.Digel)
11
Interstellar Radiation Field
Local ISRF (PS05)
Old model

• Target for CR leptons (IC)
• Energy losses

Optical
IR

• Model components
• Geometrical disk, ring, halo, bar, triaxial
  bulge, arms
• 87 stellar types (main sequence), AGB exotics
• Dust silicate, graphite, PAH (5Å few ?m)
• Absorbed light gives mid-IR (small grains PAH)
  and FIR (0.1-1 ?m grains)

PAH
Scatt.opt.
SMR00 PS05
R0
4 kpc

• Systematic errors
• Star distribution star counts
• Grain properties lab measurements
• Gas/dust proportion extinction curve
• Reasonable parameters
• Compare with ISRF data only at R?

12 kpc
16kpc
12
Distribution of CR Sources Gradient in the CO/H2
CR distribution from diffuse gammas (Strong
Mattox 1996) SNR distribution (Case
Bhattacharya 1998)
Pulsar distribution Lorimer 2004
sun
XCON(H2)/WCO Histo This work, Strong et
al.04 ----- -Sodroski et al.95,97 1.9x1020
-Strong Mattox96 Z-1 Boselli et
al.02 Z-2.5 -Israel97,00, O/H0.04,0.07
dex/kpc
13
Inner Galaxy region
Comparison with EGRET COMPTEL spectral data
Other regions demonstrate equally good agreement
14
Model comparison with data

• Convolution with EGRET PSF
• Important below 1 GeV
• A large effect at low energies especially in
  latitude affecting the overall spectral shape
• Convolution itself is model dependent -depends on
  spectrum, not fully accounted for

15
Effect of Convolution 70-100 MeV
Longitude profile blt5?
Unconvolved Convolved
16
Effect of Convolution 70-100 MeV
Latitude profile llt30?
Unconvolved Convolved
17
Effect of Convolution 0.5-1 GeV
Longitude profile blt5?
Unconvolved Convolved
18
Effect of Convolution 0.5-1 GeV
Latitude profile llt30?
Unconvolved Convolved
19
1-2 GeV
1000 2000 MeV
Convolution effect is negligible
20
Effect of De-Convolution Spectrum llt30? blt5?
Convolved data De-convolved
NB here the spatial convolution correction is
applied to the DATA based on the model. Hence the
DATA changes, not the model (procedure
appropriate for spectra)