Long duration gamma‐ray bursts (GRBs) originate from the core collapse of massive stars, but the identity of the central engine remains elusive. Previous work has shown that rapidly spinning, strongly magnetized protoneutron stars (‘millisecond protomagnetars’) produce outflows with energies, time‐scales and magnetizations σ0 (maximum Lorentz factor) that are consistent with those required to produce long duration GRBs. Here we extend this work in order to construct a self‐consistent model that directly connects the properties of the central engine to the observed prompt emission. Just after the launch of the supernova shock, a wind heated by neutrinos is driven from the protomagnetar. The outflow is collimated into a bipolar jet by its interaction with the progenitor star. As the magnetar cools, the wind becomes ultrarelativistic and Poynting flux dominated (σ0≫ 1) on a time‐scale comparable to that required for the jet to clear a cavity through the star. Although the site and mechanism of the prompt emission are debated, we calculate the emission predicted by two models: magnetic dissipation and shocks. Magnetic reconnection may occur near the photosphere if the outflow develops an alternating field structure due to e.g. magnetic instabilities or a misalignment between the magnetic and rotation axes. Shocks may occur at larger radii because the Lorentz factor of the wind increases with time, such that the faster jet at late times collides with slower material released earlier. Our results favour magnetic dissipation as the prompt emission mechanism, in part because it predicts a relatively constant ‘Band’ spectral peak energy Epeak with time during the GRB. The baryon loading of the jet decreases abruptly when the neutron star becomes transparent to neutrinos at s. Jets with ultrahigh magnetization cannot effectively accelerate and dissipate their energy, which suggests this transition ends the prompt emission. This correspondence may explain both the typical durations of long GRBs and the steep decay phase that follows. Residual rotational or magnetic energy may continue to power late time flaring or afterglow emission, such as the X‐ray plateau. We quantify the emission predicted from protomagnetars with a wide range of physical properties (initial rotation period, surface dipole field strength and magnetic obliquity) and assess a variety of phenomena potentially related to magnetar birth, including low‐luminosity GRBs, very luminous GRBs, thermal‐rich GRBs/X‐ray flashes, very luminous supernovae and short‐duration GRBs with extended emission.

Aims. We present a new numerical code, ECHO, based on a Eulerian conservative high-order scheme for time dependent threedimensional general relativistic magnetohydrodynamics (GRMHD) and magnetodynamics (GRMD). ECHO is aimed at providing a shock-capturing conservative method able to work at an arbitrary level of formal accuracy (for smooth flows), where the other existing GRMHD and GRMD schemes yield an overall second order at most. Moreover, our goal is to present a general framework based on the 3 + 1 Eulerian formalism, allowing for different sets of equations and different algorithms and working in a generic space-time metric, so that ECHO may be easily coupled to any solver for Einstein's equations. Methods. Our finite-difference conservative scheme previously developed for special relativistic hydrodynamics and MHD is extended here to the general relativistic case. Various high-order reconstruction methods are implemented and a two-wave approximate Riemann solver is used. The induction equation is treated by adopting the upwind constrained transport (UCT) procedures, appropriate to preserving the divergence-free condition of the magnetic field in shock-capturing methods. The limiting case of magnetodynamics (also known as force-free degenerate electrodynamics) is implemented by simply replacing the fluid velocity with the electromagnetic drift velocity and by neglecting the contribution of matter to the stress tensor. Results. ECHO is particularly accurate, efficient, versatile, and robust. It has been tested against several astrophysical applications, like magnetized accretion onto black holes and constant angular momentum thick disks threaded by toroidal fields. A novel test of the propagation of large-amplitude, circularly polarized Alfvén waves is proposed, and this allows us to prove the spatial and temporal high-order properties of ECHO very accurately. In particular, we show that reconstruction based on a monotonicity-preserving (MP) filter applied to a fixed 5-point stencil gives highly accurate results for smooth solutions, both in flat and curved metric (up to the nominal fifth order), while at the same time providing sharp profiles in tests involving discontinuities.

Abstract.A third order shock-capturing numerical scheme for three-dimensional special relativistic magnetohydrodynamics (3-D RMHD) is presented and validated against several numerical tests. The simple and efficient central scheme described in Paper I (Del Zanna & Bucciantini 2002) for relativistic hydrodynamics is here extended to the magnetic case by following the strategies prescribed for classical MHD by Londrillo & Del Zanna (2000). The scheme completely avoids spectral decomposition into characteristic waves, which is computationally expensive and subject to many degenerate cases in the magnetic case, while it makes use of a two-speed Riemann solver that just requires the knowledge of the two local fast magnetosonic velocities. Moreover, the onset of spurious magnetic monopoles, which is a typical problem for multi-dimensional MHD upwind codes, is prevented by properly taking into account the solenoidal constraint and the specific antisymmetric nature of the induction equation. Finally, the extension to generalized orthogonal curvilinear coordinate systems is included, thus the scheme is ready to incorporate general relativistic (GRMHD) effects.

We present a new model for the spectral evolution of pulsar wind nebulae (PWNe) inside supernova remnants (SNRs). The model couples the long‐term dynamics of these systems, as derived in the 1D approximation, with a one‐zone description (all quantities are assumed uniform in the nebula) of the spectral evolution of the emitting plasma. Our goal is to provide a simplified theoretical description that can be used as a tool to put constraints on unknown properties of PWN‐SNR systems: a piece of work that is preliminary to any more accurate and sophisticated modelling. In this paper, we apply the newly developed model to a few objects of different ages and luminosities. We find that an injection spectrum in the form of a broken power law gives a satisfactory description of the emission for all the systems we consider. More surprisingly, we also find that the intrinsic spectral break turns out to be at a similar energy for all sources, in spite of the differences mentioned above. We discuss the implications of our findings on the workings of pulsar magnetospheres, pair multiplicity and on the particle acceleration mechanism(s) that might be at work at the pulsar wind termination shock.