Radiative particle-in-cell simulations - How synthetic diagnostics help to understand plasma structure and dynamics

We present recent results of plasma simulations performed with PIConGPU, a fully relativistic 3D particle-in-cell (PIC) code running on GPU clusters. We extended our code to compute the radiation spectra of all particles in the simulation based on classical Liénard-Wiechert potentials including full coherence and polarization properties. We discuss physics tests, scaling and show simulation results of laser-wakefield accelerator and astrophysical plasmas, for which we calculated angularly resolved spectra ranging from infrared to X-ray wavelengths. Such an extensive treatment of plasma radiation across billions of macro particles makes it possible to explore temporally resolved plasma radiation spectra on linear and logarithmic photon energy scales over large solid angles ("sky-maps").
This ability of obtaining quantitative spectral data in plasma simulations poses a unique tool for determining the phase space distribution of electrons. Since spectral information is readily accessible in experiments, our results can serve as a valuable input to new diagnostics.