COSMIC RAYS IN THE MILKY WAY &
OTHER GALAXIES
IGOR V MOSKALENKO − STANFORD
The discovery of cosmic rays
! Victor Hess, an Austrian scientist, took a radiation counter (a simple electroscope) on a balloon flight
! He rose to 5200 m (without oxygen) and measured that the amount of radiation increases as the balloon climbed
! Hess correctly concluded that the ionization was caused by highly penetrating radiation coming from outside the atmosphere
! The results by Hess were later confirmed by the
Victor Hess flight on August 7, 1912
Nobel Prize: 1936
Hess Kolhörster
High energy gamma-ray emission processes
! pp −> π
0(2γ)+X − neutral pion production and decay
! Inverse Compton scattering
! Bremsstrahlung π
0p p p
p
π
0γ γ
γ
γ
e
e p
e γ
γ
hν
Fermi-LAT skymap >1 GeV, 48 months
4-year sky map, >1 GeV, front converting (best psf) (4.52M events)
! LAT: ~275B triggers, 225M Source class events
! GBM: >1000 GRBs
! shows where accelerated particles meet targets (gas, photons)
! ~80% of the emission is diffuse
! our Galaxy provides the best opportunity to study CRs: direct and indirect measurements with
excellent resolution
Cosmic Scales
parsec 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102 103 104 105 106 107 108 109 1010
3 000 km
30 000 km
300 000 km
3✕106 km
3✕107 km
2 AU
20 AU
200 AU
2 000 AU
0.1 pc
1 pc
10 pc
100 pc
1 kpc
10 kpc
100 kpc
1 Mpc
10 Mpc
100 Mpc
1 Gpc
10 Gpc
Proxima Centauri Sirius
Vega
Antares Milky Way
Andromeda Galaxy (@0.8 Mpc)
Local Group
Fermi-LAT skymaps, 48 months
E>1 GeV
! Fewer sources at high energies
! Less diffuse emission
E>10 GeV
Fermi-LAT skymaps, 48 months
! Fewer sources at high energies
! Less diffuse emission
! Interesting physics
Cosmic rays from supernovae (1934)
! “…production of
cosmic rays is related to some sporadic
process, such as the flare-up of a super- nova…”
“If cosmic rays come from super-novae their intensity in points far away from any individual super-nova will be
essentially independent of time...”
“We have tentatively suggested that the super-nova process represents the
transition of an ordinary star into a
Local emissivities
! Local gamma-ray emissivities derived from observations of the local gas clouds are consistent with the direct CR measurements
! Show intensity variations due to errors in gas mass estimates, gas
composition, or true CR intensity variations
! Morphology of γ-ray emission coincides well with shocked H
2Fermi-LAT observations of SNRs IC443 & W44
IC443
5 − 50 GeV
W44
2 − 10 GeV
clouds
Abdo+2010
• Low-energy cut off at half the π
0mass
• Clear evidence for SNR as hadronic acceleration sites
proton spectra
Fermi-LAT spectra of IC443 & W44:
π 0 decay γ-rays
e
±X,γ
gas
gas ISRF
e
+- π
+- ISM
π
0IC bremss
helio-modulation ACE
42 sigma (2003+2004 data)
HESS
PSF
B
F lux
20 GeV/n
CR species:
" Only 1 location
" modulation
π
+-
PAMELA BESS
Fermi HESS
Chandra
X,γ
synchrot ron
e
±e
±GALPROP solar modulation
CRs in the interstellar medium
AMS-02
Direct probes of CR propagation
50 kpc
p, 10 GeV e C Fe, TeV e
Effective propagation distance:
<X> ~ √6Dτ ~ 4.5×10
21R
1/4(A/12)
-1/3cm ~ 1.5 kpc R
1/4(A/12)
-1/3Helium: ~ 2.1 kpc R
1/4Carbon: ~ 1.5 kpc R
1/4- 0.36% of the surface area Iron: ~ 0.9 kpc R
1/4- 0.16%
(anti-) protons:~ 6 kpc R
1/4- 5.76%
Electrons ~ 1 kpc E
12−1/4γ-rays: probe CR p (pbar) and e± spectra in the whole Galaxy ~50 kpc across
! Direct measurements probe a very small volume of the Galaxy
! Light & heavy nuclei probe different propagation volume
! The propagation distances are
shown for nuclei for rigidity ~1
GV, and for electrons ~1 TeV
Original motivation
! Pre-GALPROP (before ~1997)
# Many models, each with a purpose to reproduce data of a single instrument
# Many different simplifying assumptions – hard to compare
# Leaky-box type models: simple, but not physical
# No or few attempts to make a self-consistent model
! GALPROP model – >17 years of development
! One Galaxy – self-consistent modeling
# The key concept underlying the GALPROP code is that various kinds of data, e.g., direct CR measurements including primary and secondary nuclei, electrons and positrons, γ-rays, synchrotron radiation, and so forth, are all related to the same astrophysical components of the Galaxy and hence have to be modeled self-consistently
! As realistic as possible
# The goal is for GALPROP-based models to be as realistic as possible and to
make use of available astronomical information, nuclear and particle data,
with a minimum of simplifying assumptions
CR Propagation: Milky Way Galaxy
Halo 0.1-0.01/ccm
Intergalactic space
1 kpc ~ 3×10
21cm
Optical image: Cheng et al. 1992, Brinkman et al. 1993 Radio contours: Condon et al. 1998 AJ 115, 1693
NGC891
Sun
“Flat halo” model (Ginzburg & Ptuskin 1976)
Transport Equations ~90 (no. of CR species)
Secondary/primary nuclei ratio & CR propagation
Using secondary/primary nuclei ratio (B/C) & radioactive isotopes (e.g. Be
10):
! Diffusion coefficient and its index
! Galactic halo size Z
h! Propagation mode and its parameters (e.g., reacceleration V
A, convection V
z)
! Propagation parameters are model-dependent
Typical parameters (model-dependent):
D ~ 10
28(ρ/1 GV)
αcm
2/s α ≈ 0.3-0.6
Z
h~ 4-6 kpc; V
A~ 30 km/s
Zh increase
Be
10/Be
9Components of GALPROP
! Detailed gas distribution from HI and CO gas surveys (energy losses from ionization, bremsstrahlung; secondary production; γ- rays from π 0 -decay, bremsstrahlung)
! Interstellar radiation field (inverse Compton losses/γ-rays for e ± )
! B-field model
! Nuclear & particle production cross sections + the reaction network (cross section database + LANL nuclear codes + phenomenological codes)
! Cosmic ray source distribution(s)
! Propagation, diffusive acceleration, convection
! Numerically solve transport equations for all cosmic ray species (stable + long-lived isotopes + pbars + leptons ~90) in 2D or 3D
! Derive the propagation parameters corresponding to the assumed
transport phenomenology and source distribution
Computing cluster
http://galprop.stanford.edu
500+ cores:
! Intel Xeon 4×32 cores
# 384 GB shared memory each
! Intel Xeon Phi coprocessor 4×2 cards
! AMD Opteron 3×64 cores
# 256 GB shared memory each
! 6174 AMD Opteron 4×48 cores
# 128 GB shared memory each
! Web server
! Data storage (12 TB RAID0)
! Infiniband link (10 Gbit/s)
Large scale study of the diffuse emission
IC
π
0Brems
! “Conventional model”: CR spectra are consistent with local measurements (CR nuclei, Fermi
electrons)
! GALPROP code with diffusion-reacceleration model for CR propagation
! Propagation parameters - fixed from CR data
! Grid of 128 models covering plausible
confinement volume, CR source distributions, etc.
! Corresponding model sky maps compared with data using maximum likelihood
! Iterative process since the model parameters (Xco, H I) depend on outcome of the fit
! A massive Fermi-LAT study – ApJ 750 (2012) 3
uniform
AMS-02 ! CERN ! Apr 15, 2015 :: IVM 21
Diffuse emission skymap
! Observed Fermi-LAT
counts in the energy range 200 MeV to 100 GeV
! Predicted counts calculated using GALPROP
reacceleration model tuned to CR data
! Residuals (Obs-Pred)/Obs
~ % level, ~10% in some places
– 55 –
Fig. 5.— Upper panel: observed Fermi–LAT counts in the energy range 200 MeV to 100 GeV used in this paper. Lower panel: predicted counts for model
SS
Z4
R20
T150
C5 in the same energy range.
To improve contrast we have used a logarithmic scale and clipped the counts/pixel scale at 3000.
The maps are in Galactic coordinates in Mollweide projection with longitudes increasing to the left and the Galactic centre in the middle.
Observed
Predicted
Spectrum and profiles
! Components of the model
# Neutral pion emission from gas H
2, H I , H II
# Inverse Compton
# Bremsstrahlung
# Detected sources
# Isotropic emission
Large scale study: residuals
! Agreement for models is overall good, but features are visible in residuals at
~% level
! Difference between illustrative models shown in right maps : structure due to
variations of model parameters
! Models details:
2: SNR
Z4
R20
T150
C5 44: Lorimer
Z6
R20
T∞
C5 93: Yusifov
Z10
R30
T150
C2 119: OB
Z8
R30
T∞
C2
Model 2
93
44
119
NASA press release
! Models reproduce the main features of the diffuse emission quite well
! Discrepancies between the physical model and high-resolution data (residuals) are the gold mines of new phenomena!
! Every extended source and/or process that is not included into the model pops up and
exposes itself as a residual
Fermi-LAT: emissivity gradient in the Galaxy
! Radial profile with Galactocentric radius of the emissivity integrated between 200 MeV and 10 GeV. Black dots/horizontal bars mark the ranges in kinematic distance encompassing the Gould Belt, the main part of the local arm, the
Perseus and outer arms.
Abdo+’2010
GALPROP
Milky Way in the global picture
! The Milky Way is the nearest example of a spiral galaxy
! It provides the best opportunity to study ongoing star
formation, cosmic rays, and related processes in the ISM
! Important reference point
! Baryonic content of the Universe is dominated by ``normal'’
galaxies
! ~70% are spiral galaxies
Milky Way as an electron calorimeter
! Calculations for Z
halo= 4 kpc
! Leptons lose ~60% of their energy
! γ-rays: 50-50 by nucleons and by leptons
Total gamma rays 1.6%
Neutral pions 0.85%
Synchrotron 0.35%
Bremsstrahlung 0.15%
Inverse Compton 0.58%
Primary electrons
1.41%
Primary nucleons 98.6%
Cosmic rays
7.90×10
40erg/s
Secondary leptons e
+: 0.33%
e
−: 0.10%
0.06% (13.5 %) 0.09% (6.6%)
* The percentages in brackets show the values relative to the luminosity of their respective lepton populations
Strong+’2011
Starforming Galaxies
Cosmic Rays as a Universal Phenomenon
! γ-ray luminosity vs. IR
luminosity for normal galaxies detected with Fermi-LAT
! The γ-ray luminosity scales linearly (index ~1.1) with the total emission of hot stars reprocessed by dust − a tracer of star formation
! The ratio approaches the
calorimetric limit in star-burst galaxies
! An evidence of the SNR-CR connection in normal star- forming galaxies
Thin target Thick target
calorimetric limit
Galactic cosmic rays in the solar system
! Allows to reconcile direct & indirect observations
# Test models of interactions
# Calibration of the instrument
! Detected sources:
# The Earth (PRD 80, 122004, 2009)
• The limb
• Terrestrial γ-ray flashes
# The Moon (ApJ 758, 140, 2012)
# The steady Sun (ApJ 734, 116, 2011)
# Solar flares
Direct measurements
Break in the CR p and He absolute fluxes
! Preliminary AMS-02 data do nor show any spectral feature
Adriani+’2011
He p
! Data from several experiments (BESS,
AMS-01, ATIC’2009, CREAM’2010,
PAMELA’2011) are all consistent and
indicate spectral hardening above ~100
GeV/nucleon
Fermi-LAT observations of the Earth’s limb
p
γ
Column density
Emission profile
>30 GeV
PSF
thin target:
68.4°≤θ≤70.0°
D if fus e em is si on dom ina te s
! Due to its proximity, the Earth is the brightest γ-ray source on the sky
! The emission is produced by the CR cascades in the atmosphere
! Most energetic γ-rays are produced by CRs hitting the top of the atmosphere at tangential directions (thin target)
γ-ray flux
>3.6 GeV
2 g cm
-2γ-ray flux
Inferred CR Proton Spectrum from
pp Model by Kachelrieβ & Ostapchenko (2012)
Index = 2.68 ± 0.04
Index
1= 2.81 ± 0.11 E
break= 302 ± 96 GeV Index
2= 2.61 ± 0.08
SPL
BPL
March 28, 2014 W. Mitthumsiri Page 10 / 12
R
H L
I P
! All scenarios are tuned to the data, except the Reference scenario
! Scenarios L and H: the local source
component is calculated by the subtraction of the propagated Galactic spectrum from the data
! The local source is assumed to be close to us, so no propagation;
only primary CR species
10-1 1
Flux2.7 F(),GV1.7 cm-2 s-1 sr-1
1 10 102 103 104 105
R igidity , G V
Calculation R CR EAM ATIC-2 PAM ELA
p
He
10-1 1
Flux2.7 F(),GV1.7 cm-2 s-1 sr-1
1 10 102 103 104 105
R igidity , G V
Calculation P CR EAM ATIC-2 PAM ELA
p
He
10-1 1
Flux2.7 F(),GV1.7 cm-2 s-1 sr-1
1 10 102 103 104 105
R igidity , G V
Calculation I CR EAM ATIC-2 PAM ELA
p
He
10-1 1
Flux2.7 F(),GV1.7 cm-2 s-1 sr-1
Calculation L CR EAM ATIC-2 PAM ELA
p=Galactic+local
Galactic p local p
He=Galactic+local Galactic He
local He 10-1
1
Flux2.7 F(),GV1.7 cm-2 s-1 sr-1
Calculation H CR EAM ATIC-2 PAM ELA
p=Galactic+local
Galactic p local p He=Galactic+local
Galactic He local He
Propagation
Reference
Injection
Local LE Local HE
P and He spectra
Vladimirov+’2012, ApJ 752, 68
2 5
10
p/Hefluxratio
1 10 102 103 104 105
R igidity ,G V C alculation R C alculation P
C alculation I C alculation L C alculation H
PA M ELA A TIC -2 C R EA M Plain D iff.M odel
10-3 10-2 10-1 1
p/p10-3
10-1 1 10 102 103 104
E,G eV /nucleon C alculation R C alculation P
C alculation I C alculation L C alculation H
PA M ELA
D iff.-R eacc.M odel
10-4 10-3 10-2 10-1
Anisotropy,
103 104 105 106
E,G eV Calculations
R, I, L and H Calculation P Diff.-Reacc. Model
More predictions
! P/He ratio is tuned in all scenarios except Reference scenario
! Predicted antiproton/proton ratio agrees with the existing data, but exhibits different behavior at >100 GeV
! Only scenario P agrees with the data on CR anisotropy
! Only scenario L can explain the sharp break in the p, He spectra
! Await for more accurate data
B/C ratio
! P-scenario is predicting an upturn in the B/C ratio at
~100 GeV/n (~200 GV)
1 TeV
New measurements of B/C ratio
PAMELA fit to the B/C ratio
! Use GALPROP
! Parameters from Vladimirov+’2012
! Best fit index of the diffusion coefficient: δ = 0.4
! No significant spectral features
<400 GeV/n (AMS-02)
E(GeV/n)
0.4 1 2 3 4 5 6 7 8 10 20 30 40 100
B/C
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
PAMELA AMS-02 (preliminary) Galprop
CREAM TRACER ATIC-2 HEAO AMS-01
E(GeV/n)
0.4 1 2 3 4 5 6 7 8 910 20 30 40 50 100
1.7 (GeV/n)-1 s sr)2 (m2.7 E×Flux
10-1
1
10 C
B
PAMELA Galprop CREAM TRACER ATIC-2 HEAO CRN
B & C spectra PAMELA
B/C ratio PAMELA
?
PAMELA data: rising positron fraction
! PAMELA team reported a rise in the positron fraction compared to the “standard”
model predictions
! “Standard” model:
# Secondary production
# Steady state
# Smooth CR source distribution
Nature 458, 607 (2009)
IVM&Strong’1998
PAMELA
Adriani+’2009
AMS-02: Rise in the positron fraction
! Asymptotically approaches a const ~0.15-0.2 ?
! A cut off at HE ?
Nature 458, 607 (2009)
IVM&Strong’1998
Positron fraction: Astrophysical papers
! Blasi 2009 “Origin of the positron excess in cosmic rays”
! Blasi & Serpico 2009 “High-energy antiprotons from old supernova remnants”
! Mertsch & Sarkar 2009 “Testing astrophysical models for the PAMELA positron excess with cosmic ray nuclei”
! Shaviv+ 2009 “Inhomogeneity in cosmic ray sources as the origin of the electron spectrum and the PAMELA anomaly”
! Delahaye+2010 “Galactic electrons and positrons at the Earth: new estimate of the proimary and secondary fluxes”
! Stawarz+2010 “on the energy spectra of GeV/TeV cosmic ray leptons”
! Lee+ 2011 “Explaining the cosmic ray e
+/(e
-+e
+) and pbar/p ratios using a steady-state injection model”
! Kachelriess+2011 “Antimatter production in supernova remnants”
! Kachelriess & Ostapchenko 2013 “B/C ratio and the PAMELA positron excess”
! Blum+ 2013 “AMS-02 results support the secondary origin of cosmic ray positrons”
! Cholis & Hooper 2013 “Dark matter and pulsar origin of the rising cosmic ray positron fraction in light of new data from the AMS”
! Erlykin & Wolfendale 2013 “Cosmic ray positrons from a local, middle-aged supernova remnant”
! Berezhko & Ksenofontov 2013 “Energy spectra of electrons and positrons produced in supernova remnants”
! Berezhko & Ksenofontov 2013 “Antiprotons produced in supernova remnants”
! Cholis & Hooper 2014 “Constraining the origin of the rising cosmic ray positron fraction with the boron- to-carbon ratio”
! Di Mauro+2014 “Interpretation of AMS-02 electrons and positrons data”
! Mertsch & Sarkar 2014 “AMS-02 data confronts acceleration of cosmic ray secondaries in nearby sources”
! Cowsik+2014 “The origin of the spectral intensities of cosmic ray positrons”
! …
! + Dark Matter paper >1000
Interpretations
! Dark matter annihilation/decay
! SNR shocks:
# Galactic SNRs
# Local SNR(s)
! “Nested Leaky-Box” (SNRs)
! Inhomogeneity of CR sources (SNRs)
! “Model-independent estimates”
! Photoproduction
! Pulsars & Pulsar Wind Nebulae
ISM
Old friends − pulsars
! Arons 1981 “Particle acceleration by
pulsars”
! Harding & Ramaty 1987 “The pulsar contribution to
Galactic cosmic ray positrons”
! Boulares 1989 “The nature of the cosmic- ray electron
spectrum, and
supernova remnant contributions”
“Therefore, the only role observed pulsars might play as direct cosmic ray sources is in providing positrons and electrons…”
3 components:
! Secondary e
+/-! Primary e
-from SNR
! Primary e
+/-from
pulsars
Reinvention of the Nested Leaky-Box − SNRs
! Cowsik & Wilson 1974 “The nested Leaky-Box model for Galactic cosmic rays”
! Berezkho+2003
“Cosmic ray production in supernova
remnants including reacceleration: The secondary to
primary ratio”
n = 1 cm
-3n = 0.3 cm
-3n = 0.003 cm
-3“In this paper we shall in addition take the effect of nuclear spallation inside the sources into account.
The energy spectrum of these source secondaries is harder than that of
reaccelerated secondaries.
Therefore it plays a dominant role at high
energies for a high-density ISM...”
“The ‘inner box’ of cosmic ray confinement,
corresponding to the region immediately surrounding the source, is assumed to have energy-dependent life time...”
B/C
Secondary production in SNR shock
E
2F lux
E
Primary
shock γ~2-2.4
Secondary δ
γ~3-3.4 δ
! Gas in the shock – target for p, A
! Flatter spectrum of p, A – flatter spectrum of secondaries
! Assume no energy losses
! δ~0.3-0.7 – effect of IS propagation (no losses)
! Same effect should be observed for any secondaries (pbars, B, e
+/–)
! Energy losses will modify the spectra of e
+/–at low and high energies - depend on the environment
shock γ~2-2.4
δ
E
2F lux
E
Sec
SNR
ISM Prim
Positron production in SNR shock
! Old SNR ~10
4yr
! Leaky-box type propagation model:
# Halo size ~3 kpc
# D~10
28(E/GeV)
0.6cm
2s
-1# IC+sync energy losses
Blasi 2009
Antiproton production in SNR shock
! Same model as in Blasi’2009
! pbars/p ratio does not contradict to positron fraction
Blasi & Serpico 2009
Boron production in SNR shock - I
! A model similar to Blasi & Serpico 2009
! Proposed to distinguish between pulsar & SNR origin of the positron excess
! Leaky-Box for Galactic propagation
! Cross sections from Silberberg &
Tsao
! Nearby SNR(s)
! Data: HEAO-3, ATIC-2, CREAM
! Predicts an upturn in the B/C ratio near ~100 GeV/nucleon
! Ti/Fe is much less reliable than B/C:
# Charge resolution at HE
# Uncertain production cross sections
• Model prediction
• Contamination: production in the atmosphere & instrument