Particle Acceleration in SNRs

The main production mechanisms responsible for high-energyγ-ray emission in SNRs are synchrotron emission of electrons in the SNR magnetic field,

Monoceros [8] 120 0.17 (>800 GeV) Cas A [6] 232 0.058 (>1 TeV) Tycho [7] 64.6 <0.058 (>1 TeV)

γ-Cygni [139] 47 <1.1 (>500 GeV)

Durham Group

SN1006 [42] 41 <1.7 (>300 GeV) Whipple

Monoceros [115] 13.1 <4.8 (>500 GeV) Cas A [115] 6.9 <0.66 (>500 GeV)

Tycho [39] 14.5 <0.8 (>300 GeV)

γ-Cygni [39] 9.3 <2.2 (>300 GeV)

W44 [39] 6.0 <3.0 (>300 GeV) W51 [39] 7.8 <3.6 (>300 GeV) W63 [39] 2.3 <6.4 (>300 GeV)

CAT

Cas A [74] 24.4 <0.74 (>400 GeV)

IC-scattering, electron bremsstrahlung, and π^◦ production. The first three mechanisms involve electrons. The fourth mechanism involves hadrons. In all of these processes the VHE γ-ray production is only possible, when the seed particles whether hadrons or electrons have very high energies. For this reason, it is important to understand the possible acceleration mechanisms of charged particles. There are three fundamental processes, through which particles can be accelerated, [128, 120]. These are given below:

• Shock Acceleration:Shock waves are smooth bulk motions of plasma produced in SN explosions and these waves propagate into the inter-stellar space sweeping up interinter-stellar matter. Particles passing through shock fronts of SNRs can be accelerated by the first-order Fermi ac-celeration mechanism. In this mechanism, a particle passing though a shock will be scattered by magnetic inhomogeneities behind the shock.

Tabelle 2.2: Past Observations of the Plerions in GeV - TeV γ-ray range.

Information taken from [63].

Plerion Exposure Flux/ Pulsation

Name Time Upper Limit in VHEγ-ray

[hrs] [×10⁻¹¹cm⁻²s⁻¹] Signal ALL ground-based detectors in Section 1.3.3

Crab Nebula → ∞ 7.0 (>400 GeV) No

The particle gains energy from this interaction and scatters back and crosses the shock front again. It can again be scattered by the magnetic inhomogeneities this time ahead of the shock, and so bounces back and forth many times gaining energy each time. The energy that a particle gains is proportional to v/c, where v is the velocity of the shock front relative to the un-shocked interstellar space and cis the speed of light.

After n crossings the particle has an energy of E = E₀ (v/c)ⁿ. After cer-tain time the particles will be carried away by the downstream shock.

If the probability of remaining in the shock region after each crossing is P, then after n crossings, the remaining number of particles, N, is given as N = N₀Pⁿ, where N₀ is the initial number of particles before the interaction with the shock front. The resulting energy spectrum of these particles is approximately dN/dE = E^−α, where α ∼2.0 - 2.5.

This mechanism is also known as diffusive shock acceleration (DSA) mechanism, which was introduced by Blanford and Ostriker, [29], and Bell, [20] in 1978. The observational evidence that cosmic-ray electrons are accelerated in SNRs is given in Sections 2.3.1 and 2.3.2.

• Stochastic Acceleration: In this scenario, particles are immersed in a turbulent medium and change their energy randomly due to many interactions with moving interstellar magnetic field and are eventually, on average, accelerated. This process is called the second-order Fermi acceleration mechanism. Through this mechanism particles gain energy proportional to v²/c².

Abbildung 2.8: The Vela supernova remnant as a good example for composite type SNRs. The picture shows an X-ray image taken by the ROSAT and Chandra satellites.

• Direct Acceleration:The most direct way to accelerate charged par-ticles is through DC electric fields. This type of mechanism is used to explain the acceleration of particles in neutron stars. One problem in this scenario is that oppositely charged particles are accelerated in opposite directions, and a large scale charge separation occurs. While particles gain energy from the electric field, they are exposed to the drag force from the oppositely charged particles. Therefore, it is the inter-play between these two forces that determines whether or not electrons or ions can be accelerated out of a bulk particle distribution.

Because SNRs are regarded as the most probable production and acce-leration sites of cosmic rays, the TeV γ-ray observations of SNRs play an important role in the debate of probing the origin of the cosmic rays. These debates were first proposed by Ginzburg and Syrovatskii, [71].

Detection of the signature of aπ⁰-bump at MeV energies and a spectrum extending to tens of TeV would be a clear indication that cosmic-ray accele-ration does take place in SNRs (DAV model proposed by Drury, Aharonian, and Völk, [61]). The γ-ray emission from π⁰ decay peaks at the beginning of the Sedov phase (Section 2.3) and then slowly decays with the SNR evolu-tion in time. The expected γ-ray spectrum is very hard, which enables TeV γ-ray observations. However, the γ-ray luminosity from π⁰ decay compared

to the luminosity from inverse-Compton scattering of relativistic electrons and CMBR is be too low to be detected above 100 MeV in most of the SNRs.

From a selected set of shell-type SNRs observed by EGRET, the ones having an interaction with a nearby molecular cloud (which provides a high-density target for π^◦ production) can be selected for further observations in TeV range.

The energy spectrum of most of the shell-type SNRs listed in Table 2.1 and Table 2.2 like Vela and SN1006 can be explained as a composite of a synchrotron and an inverse-Compton component emitted by a population of accelerated electrons. Detected TeV fluxes from two other SNRs, Cas A and PSR 1706-44, are not strong enough to constrain the emission mechanisms.

However, the energy spectrum of RX J1713-3946 is claimed by CANGAROO collaboration (Section 1.3.3) to fit best to the models, which produce γ rays via π⁰ decay. This claim has been disputed by Reimer and Pohl [143]. They used the complete data set of EGRET measurements to show that the GeV flux required by π⁰ decay models significantly exceeds the EGRET measure-ments. From the recent results of the H.E.S.S. experiment on RX J1713-3946 observations, it can be concluded that because this SNR is also interacting with some molecular clouds, it needs multi-wavelength observations to disen-tangle the relative contributions of various processes, [48].

2.5 Model of Gamma-ray Emission from the