Speaker
Description
It is generally accepted that radio relics are the result of synchrotron emission from shock-accelerated electrons. Current models, however, are still unable to explain several aspects of their formation. In this paper, we focus on three outstanding problems: i) Mach number estimates derived from radio data do not agree with those derived from X-ray data, ii) cooling length arguments imply a magnetic field that is at least an order of magnitude larger than the surrounding intracluster medium (ICM), and iii) spectral index variations do not agree with standard cooling models. We use a hybrid approach to solve these problems; first identifying typical shock conditions in cosmological simulations and then using these to inform idealized shock-tube simulations, which can be run with substantially higher resolution. We post-process our simulations with the cosmic ray electron spectra code CREST and the emission code CRAYON+, allowing us to generate mock observables ab-initio. We identify that upon running into an accretion shock, merger shocks generate a dense, shock-compressed sheet, which, in turn, runs into upstream density fluctuations. This mechanism directly gives rise to solutions to the three aforementioned problems: density fluctuations lead to a distribution of Mach numbers forming at the shock-front. This flattens cosmic ray electron spectra, thereby biasing radio-derived Mach number estimates to higher values. We show that such estimates are particularly inaccurate in weaker shocks ($\mathcal{M} \lesssim 2$). Secondly, the density sheet becomes Rayleigh-Taylor unstable at the contact discontinuity, causing turbulence and additional compression downstream. This amplifies the magnetic field from ICM-like conditions up to $\upmu$G levels. We show that synchrotron-based measurements are strongly biased by the tail of the distribution here too. Finally, the same instability also breaks the common assumption that matter is advected at the post-shock velocity downstream, thus invalidating laminar-flow based cooling models.
