General relativistic magnetohydrodynamical simulations of the jet in M87

Context. The connection between black hole, accretion disk, and radio jet can be constrained best by fitting models to observations of nearby low-luminosity galactic nuclei, in particular the well-studied sources Sgr A* and M87. There has been considerable progress in modeling the central engine of active galactic nuclei by an accreting supermassive black hole coupled to a relativistic plasma jet. However, can a single model be applied to a range of black hole masses and accretion rates? Aims. Here we want to compare the latest three-dimensional numerical model, originally developed for Sgr A* in the center of the Milky Way, to radio observations of the much more powerful and more massive black hole in M87. Methods. We postprocess three-dimensional GRMHD models of a jet-producing radiatively inefficient accretion flow around a spinning black hole using relativistic radiative transfer and ray-tracing to produce model spectra and images. As a key new ingredient in these models, we allow the proton-electron coupling in these simulations depend on the magnetic properties of the plasma. Results. We find that the radio emission in M87 is described well by a combination of a two-temperature accretion flow and a hot single-temperature jet. Most of the radio emission in our simulations comes from the jet sheath. The model fits the basic observed characteristics of the M87 radio core: it is "edge-brightened", starts subluminally, has a flat spectrum, and increases in size with wavelength. The best fit model has a mass-accretion rate of (M)over dot similar to 9 x 10(-3) M-circle dot yr(-1) and a total jet power of P-j similar to 10(43) erg s(-1). Emission at lambda = 1.3mm is produced by the counter-jet close to the event horizon. Its characteristic crescent shape surrounding the black hole shadow could be resolved by future millimeter-wave VLBI experiments. Conclusions. The model was successfully derived from one for the supermassive black hole in the center of the Milky Way by appropriately scaling mass and accretion rate. This suggests the possibility that this model could also apply to a wider range of low-luminosity black holes.