Using three-dimensional magnetohydrodynamics simulations, we show that the efficiency of cosmic-ray (CR) production at supernova remnants (SNRs) is over-predicted if it could be estimated based on proper motion measurements of H$alpha$ filaments in c
ombination with shock-jump conditions. Density fluctuations of upstream medium make shock waves rippled and oblique almost everywhere. The kinetic energy of the shock wave is transferred into that of downstream turbulence as well as thermal energy which is related to the shock velocity component normal to the shock surface. Our synthetic observation shows that the CR acceleration efficiency as estimated from a lower downstream plasma temperature, is overestimated by 10-40%, because rippled shock does not immediately dissipate all upstream kinetic energy.
Synchrotron X-rays can be a useful tool to investigate electron acceleration at young supernova remnants (SNRs). At present, since the magnetic field configuration around the shocks of SNRs is uncertain, it is not clear whether electron acceleration
is limited by SNR age, synchrotron cooling, or even escape from the acceleration region. We study whether the acceleration mechanism can be constrained by the cutoff shape of the electron spectrum around the maximum energy. We derive analytical formulae of the cutoff shape in each case where the maximum electron energy is determined by SNR age, synchrotron cooling and escape from the shock. They are related to the energy dependence of the electron diffusion coefficient. Next, we discuss whether information on the cutoff shape can be provided by observations in the near future which will simply give the photon indices and the flux ratios in the soft and hard X-ray bands. We find that if the power-law index of the electron spectrum is independently determined by other observations, then we can constrain the cutoff shape by comparing theoretical predictions of the photon indices and/or the flux ratios with observed data which will be measured by NuSTAR and/or ASTRO-H. Such study is helpful in understanding the acceleration mechanism. In particular, it will supply another independent constraint on the magnetic field strength around the shocks of SNRs.
An extreme case of electron shock drift acceleration in low Mach number collisionless shocks is investigated as a plausible mechanism of initial acceleration of relativistic electrons in large-scale shocks in galaxy clusters where upstream plasma tem
perature is of the order of 10 keV and a degree of magnetization is not too small. One-dimensional electromagnetic full particle simulations reveal that, even though a shock is rather moderate, a part of thermal incoming electrons are accelerated and reflected through relativistic shock drift acceleration and form a local nonthermal population just upstream of the shock. The accelerated electrons can self-generate local coherent waves and further be back-scattered toward the shock by those waves. This may be a scenario for the first stage of the electron shock acceleration occurring at the large-scale shocks in galaxy clusters such as CIZA J2242.8+5301 which has well defined radio relics.
