galactic source populations of vhe gamma rays felix aharonian dublin institute for advanced studies,...
Post on 22-Dec-2015
217 views
TRANSCRIPT
Galactic Source Populations of VHE Gamma Rays
Felix AharonianDublin Institute for Advanced Studies, DublinMax-Planck-Institut f. Kernphysik, Heidelberg
Lecture 4
VHE gamma-ray observations:
“Universe is full of extreme accelerators on all astronomical scales” Extended Galactic Objects Shell Type SNRs Giant Molecular Clouds Star formation regions Pulsar Wind Nebulae
Compact Galactic Sources Binary pulsar PRB 1259-63 LS5039, LSI 61 303 – microquasars? Cyg X-1 ! (?) - a BH candidate
Galactic Center Extragalactic objects M87 - a radiogalaxy TeV Blazars – with redshift from 0.03 to 0.18 or even 0.5 ? (3C 279)
and a large number of yet unidentified TeV sources …
VH
E gam
ma-ray source populations
Potential Gamma Ray Sources
Major Scientific Topics
G-CRs Relativistic Outflows
Compact Objects Cosmology
ISM SNRsSFRs Pulsars Binaries
Galactic SourcesExtragalactic Sources
GRBs AGN GLX CLUST
IGM
GMCsM
ag
neto
sph
ere M
icro
qu
asa
rs C
old
Win
d
Pu
lsar
Neb
ula
Bin
ary
P
uls
ars
Rad
iog
ala
xie
s
B
laza
rs
N
orm
al
Sta
rbu
rst
EXG-CRs
EB
L
GeV GeVGeV GeV GeV GeV
Microquasars ?
Pulsars/Plerions ?
SNRs ?
Galactic Center ?
. . .
Gaisser 2001
OB, W-R Stars ?
* the source population responsible for the bulk of GCRs are PeVatrons ?
Galactic TeVatrons and PeVatrons - particle acceleratorsresponsible for cosmic rays up to the “knee” around 1 PeV
Visibility of SNRs in high energy gamma-rays
F(>E)=10-11 A (E/1TeV)-1 ph/cm2s
A=(Wcr/1050erg)(n/1cm-3
)(d/1kpc) -2
for CR spectrum with =2
if electron spectrum >> 10 TeV synchrotron X-rays and IC TeV ’s
main target photon field 2.7 K: F,IC/Fx,sinch=0.1 (B/10G)-2
Detectability ? compromise between angle (r/d) and flux F (1/d2) typically A: 0.1-0.01 :
0.1o - 1o
1000 yr old SNRs (in Sedov phase)
o component dominates if A > 0.1 (Sx/10 J)(B/10 G ) -2
TeV -rays – detectable if A > 0.1
nucleonic component of CRs - “visible” through TeV (and GeV) gamma-rays !
Inverse Compton
0 –decay (A=1)
TeV -rays and shell type morphology: acceleration of p or e in the shell toenergies exceeding 100TeV
2003-2005 data
can be explained by -rays from pp ->o ->2
but IC canot be immediately excluded…
RXJ1713.7-4639
and with just ”right” energetics
Wp=1050 (n/1cm-3)-1 erg/cm3
leptonic versus hadronic
IC origin ? – very small B-field, B < 10 G, and very large E, Emax > 100 TeV
two assumptions hardly can co-exists within standard DSA models, bad fit of gamma-ray spectrum below a few TeV, nevertheless …
arguments against hadronic models:
nice X-TeV correlaton well, in fact this is more natural for
hadronic rather than leptonic models
relatively weak radio emission problems are exaggerated
lack of thermal X-ray emission => very low density plasma or low Te ? we do not (yet) know the mechanism(s) of electron heating in supernova remnants so comparison with other SNRs is not justified at all
Suzaku measurements => electron spectrum 10 to 100 TeV
Variability of X-rays on year timescales - witnessing particle acceleration in real time
flux increase - particle acceleration
flux decrease - synchrotron cooling *)
both require B-field of order 1 mG in hot spots and, most likely, 100G outside
Uchiyama, FA, Tanaka, Maeda, Takahashi, Nature 2007
*) explanation by variation of B-field does’t work as demonstrated for Cas A (Uciyama&FA, 2008)
strong support of the idea of amplification of B-field by in strong nonlinear shocks through non-resonant
streaming instability of charged energetic particles (T. Bell; see also recent detailed theoretical treatment of the problem by Zirakashvili, Ptuskin Voelk 2007)
acceleration in Bohm diffusion regime
Strong support for Bohm diffusion - from the synchrotron cutoffgiven the upper limit on the shock speed of order of 4000 km/s !
with h=0.67 +/- 0.02keV
energy spectrum of synchrotron radiation of electrons in the
framework of DSA (Zirakashvili&FA 2007)
B=100 G + Bohm diffusion - acceleration of particles to 1 PeV
(Tanaka et al. 2008)
protons:
dN/dE=K E- exp[-(E/Ecut)]
-rays:
dN/dE v E- exp[-(E/E0)]
=+, 0.1, =/2, E0 = Ecut/20
Wp(>1 TeV) ~ 0.5x1050 (n/1cm-3)-1 (d/1kpc)2
RXJ 1713.7-3946
neutrinos: marginally detectable by KM3NeT
Probing PeV protons with X-rays
SNRs shocks can accelerate CRs to <100 TeV unless magnetic field significantly exceeds 10 G
recent theoretical developments: amplification of the B-field up to >100 mG is possible through plasma waves generated by CRs
>1015 eV protons result in >1014 eV gamma-rays and electrons “prompt“ synchrotron X-rays
t() = 1.5 (/1keV) -1/2 (B/1mG) -3/2 yr << tSNR
typically in the range between 1 and 100 keV with the ratio Lx/L larger than 20% (for E-2 type spectra)
“hadronic“ hard X-rays and (multi)TeV -rays – similar morphologies !
three channels of information
about cosmic PeVatrons:
10-1000 TeV gamma-rays
10-1000 TeV neutrinos
10 -100 keV hard X-rays
-rays: difficult, but possible with future “10km2“ area multi-TeV IACT arrays
neutrinos: marginally detectable by IceCube, Km3NeT - don’t expect spectrometry, morphology; uniqueness - unambiguous signatute! “prompt“ synchrotron X-rays: smooth spectrum a very promising channel - quality! (NexT, NuSTAR, SIMBOL-X)
protons
broad-band
GeV-TeV-PeV s
synch. hard X-rays
broad-band emision initiated by pp interactiosn : Wp=1050 erg, n=1cm-3
no competing X-ray radiation mechanisms above 30 keV
probing hadrons with secondary
X-rays with sub-arcmin resolution! Simbol-X
new technology focusing telescopes NuSTAR (USA), Simbol-X (France-Italy), NeXT (Japan) will provide X-ray imaging and spectroscopy in the 0.5-100 keV band with angular resolution 10-20 arcsec
and sensitivity as good as 10-14 erg/cm2s!
complementary to gamma-ray and neutrino telescopes
advantage - (a) better performance, deeper probes (b) compensates lack of neutrinos and gamma-rays at “right energies”
disadvantage - ambiguity of origin of X-rays
Searching for Galactic PeVatrons
gamma-rays from surrounding regions add much to our knowledge about highest energy protons which quickly escape the accelerator and therefotr do not signifi-cantly contribute to gamma-ray production inside the proton accelerator-PeVatron
the existence of a powerful accelerator is not
yet sufficenrt for -radiation; an additional component – a dense gas target - is required
older source – steeper -ray spectrum
tesc=4x105(E/1 TeV) -1 -1 yr (R=1pc); =1 – Bohm Difussion
Qp = k E-2.1 exp(-E/1PeV) Lp=1038(1+t/1kyr) -1 erg/s
Gamma-rays and neutrinos inside and outside of SNRs
neutrinosgamma-rays
SNR: W51=n1=u9=1
ISM: D(E)=3x1028(E/10TeV)1/2 cm2/s
GMC: M=104 Mo d=100pcd=1 kpc
1 - 400yr, 2 - 2000yr, 3 - 8000yr, 4 - 32,000 yr
[S. Gabici, FA 2007]
MGRO J1908+06 - a PeVatron?
HESSpreliminary
Milagro
gamma-ray emitting clouds in GC region
HESS J1745-303
diffuse emission along the plane!
(1) indirect discovery of the site of particle acceleration
(2) measurements of the CR diffusion coefficient
Pulsar Winds and Pulsar Wind Nebulae (Plerions)
Crab Nebula – a perfect PeVatron of electrons (and protons ?)
Crab Nebula – a very powerful W=Lrot=5x1038 erg/s
and extreme accelerator: Ee > 1000 TeV
Emax=60 (B/1G) -1/2 -1/2 TeV and hcut=(0.7-2) f-1mc2 -1 = 50-150 -1 MeV
=1 – minimum value allowed by classical electrodynamics Crab: hcut= 10MeV: acceleration at ~10 % of the maximum rate ( 10)
maximum energy of electrons: E=100 TeV => Ee > 100 (1000) TeV B=0.1-1 mG
– very close the value independently derived from the MHD treatment of the wind
1-10MeV
100TeV
* for comparison, in shell type SNRs DSA theory gives =10(c/v)2=104-105
Standard MHD theory
cold ultrarelativistc pulsar wind terminates by a reverse shock resulting in acceleration
with an unprecedented rate: tacc=rL/c, < 100
*)
synchrotron radiation => nonthermal optical/X-ray nebulaInverse Compton => high energy gamma-ray nebula
.MAGIC (?)
HEGRA
TeV gamm-rays from other Plerions (Pulsar Wind Nebulae)
Crab Nebula is a very effective accelerator but not an effective IC -ray emitter
we see TeV gamma-rays from the Crab Nebula because of large spin-down flux
gamma-ray flux << “spin-down flux“ because of large magnetic field but the strength of B-field also depends on
less powerful pulsar weaker magnetic field higher gamma-ray efficiency detectable gamma-ray fluxes from other plerions
HESS confirms this prediction ! – many famous PWNe are already detected in TeV gamma-rays - MSH 15-52, PSR 1825, Vela X, ...
HESS J1825 (PSR J1826-1334)
Luminosities: spin-down: Lrot= 3 x 1036 erg/s
X: 1-10 keV Lx=3 x 1033 erg/s (< 5 arcmin)
: 0.2-40TeV L=3 x 1035 erg/s (< 1
degree)the -ray luminosity is comparable to the TeV luminosity of the Crab Nebula, while the spindown luminosity is two orders of magnitude less ! Implications ? (a) magnetic field should be significantly less than 10G.
but even for Le=Lrot this condition alone is not sufficient to achieve 10 % -ray production
efficiency (Comton cooling time of electrons on 2.7K CMBR exceeds the age of the source) (b) the spin-down luminosity in the past was much higher.
red – below 0.8 TeVyellow – 0.8TeV -2.5 TeVblue – above 2.5 TeVPulsar‘s period: 110 ms, age: 21.4 kyr,
distance: 3.9 +/- 0.4 kpc
energy-dependent image - electrons!
Gamma-ray Binaries
Mirabel 2006
PSR1259-63 - a unique high energy laboratory
binary pulsars - a special case with strong effects associated with the optical star on both the dynamics of the pulsar wind and the radiation before and after its termination
the same 3 components - Pulsar/Pulsar Wind/Synch.Nebula - as in plerionsbut with characteristics radiation and dynamical timescales less than hours
both the cold ultrarelativistic wind and shocke-accelerated electrons are illuminated by optical radiation from the companion star => detectable IC gamma-ray emission
on-line watch of creation/termination of the pulsar wind accompanied with formation of a shock and effective acceleration of electrons
time evolution of fluxes and energy spectra of X- and gamm-rays contain unique information about the shock dynamics, electron acceleration, B(r), plus … a unique probe of the Lorentz factor of the cold pulsar wind
HESS: detection of TeV gamma-rays from PSR1259-63 several days before the periastron and 3 weeks after the peristron
the target photon field is function of time, thus the only unknown parameter is B-field? Easily/robustly predictable X and gamma-ray fluxes ?
unfortunately more unknown parameters - adiabatic losses, Doppler boosting, etc. One needs deep theoretical (especially MHD) studies to understand this source
Probing the wind Lorentz factor with comptonizied radiation
Loretz factors exceeding 106 are excluded
the effect is not negligible, but notsufficient to explain the lightcurve
GLAST
HE
SS
Khangulyan et al. 2008
TeV Gamma Rays From microquasars?
HESS, 2005
MAGIC, 2006microqusars or binary pulsars?
independent of the answer – particle acceleration is linked to (sub) relativistic outflows
scenarios? -ray production region within and outside the
binary system cannot be excluded periodicity expected? yes – because of periodic variation of the geometry
(interaction angle) and density of optical photons – as target photons for IC scattering
and absorption, as a regulator of the electron cut-off energy; also because of
variation of the B-field, density of the ambient plasma (stellar wind), ...
periodicity detected ! is everything OK ?
may be OK, but a lot of problems and puzzles with interpretation of the data …
LS5039 and LS I +61 303 as TeV gamma-ray emitters
LS 5039 as a perfect TeV clock
and an extreme TeVatron
close to inferior conjuction - maximum
close to superior conjuction – minimum one needs a factor of 3 or better sensitivity compared to HESS to detect signals within different phase of width 0.1 and measure energy spectra (phase dependent!)
can electrons be accelerated to > 20 TeV in presence of radiation? yes, but accelerator should not be located deep inside the binary system, and even at the edge of the system < 10
does this excludes the model of “binary pulsar” yes, unless the interaction of the pulsar and stellar winds create a relativistic bulk motion of the shocked material (it is quite possible)
can we explain the energy dependent modulation by absorption ? yes, taking into account the anysotropic character of IC scattering ? can the gamma-ray producton region be located very deep inside the system
no, unless magnetic field is less than 10(R/R*)-1 G (or perhaps not at all)
TeV observations with a sensitivity a factor of 3 (or so) better than HESS, to measure, in particular, the fluxes and spectra within narrow phases , very import are both 10 TeV (maximum electron energy and no absorption)
and 0.1 TeV regions (maximum absorption, maximum anysotropy effect, etc.)
GeV observation (GLAST) to measure the cascade component
X-ray observations - synchrotron radiation of primary and secondary electrons
neutrinos - if -ray are of hadronic origin, and less than several percent of the original flux escapes the source, one may expect neutrino flux marginally detectable by km3 volume detectors (current limit from X-ray observations), could be higher If GLAST detects high (cascade) fluxes
future key observations