Abstract.We study the evolution of the field on the surface of proto-neutron stars in the immediate aftermath of stellar core collapse by analyzing the results of self-consistent, axisymmetric simulations of the cores of rapidly rotating high-mass stars. To this end, we compare the field topology and the angular spectra of the poloidal and toroidal field components over a time of about one seconds for cores. Both components are characterized by a complex geometry with high power at intermediate angular scales. The structure is mostly the result of the accretion of magnetic flux embedded in the matter falling through the turbulent post-shock layer onto the PNS. Our results may help to guide further studies of the long-term magnetothermal evolution of proto-neutron stars. We find that the accretion of stellar progenitor layers endowed with low or null magnetization bury the magnetic field on the PNS surface very effectively.
IntroductionThe wide range of surface magnetic field strengths found across the population of neutron stars (e.g. [1]) is the result of a combination of processes operating at their formation during stellar core collapse and effects that modify their structure afterwards, such as cooling or accretion. Most observations pertain to relatively old neutron stars, and do not place tight constraints on the very early evolution of the field. Hence, it is difficult to draw conclusions on the fields of nascent proto-neutron stars (PNSs) from observations or, conversely, to connect theoretical results from models of supernova core collapse to evolved neutron stars. Nevertheless, the field is likely to bear at least a significant imprint of these early conditions.A thorough theoretical study of the evolution of the magnetic field of PNSs should optimally be based on three-dimensional long-term simulations including magnetohydrodynamics and neutrino transport, which treat the global dynamics of core collapse and a possible supernova explosion and the generation of the PNS field in a self-consistent manner. The enormous computational costs of spectral neutrino transport mean that thus far only a small number of such simulations exist with the most sophisticated ones [2-4] run for only a fairly limited period.Hence, less expensive axisymmetric models are still of considerable use for approaching this topic (e.g. [5][6][7][8][9]). Depending on the rotational energy and the seed field of the pre-collapse star, but also potentially on the input physics and the numerical method and grid resolution, their results range from minor modifications of non-magnetized versions of the same cores to magnetically driven explosions of a preferentially axial morphology. Depending on the global