2004ApJ...616..346A

Most of the proposed associations between magnetars and supernova remnants suffer from age problems. Usually, supernova remnant ages are determined using some approximation for the Sedov-Taylor supernova phase, which yields a relation between radius and age for a fixed energy of the explosion (generally assumed to be ∼10⁵¹ ergs). Those ages do not generally agree with the characteristic ages of the (proposed) associated magnetars. We show in this work that a faster expansion results when the energy injected into the supernova remnant by magnetar spin-down is taken into account, thus helping to improve the matches between characteristic ages and supernova remnant ages. However, the magnetar velocities inferred from observations would make some associations inviable if correct. Since characteristic ages may not be good age estimators after all, their influence on the likelihood of the association may not be as important. In this work, we perform simple numerical simulations of supernova remnant expansion with internal magnetars and apply them to the observed sample of objects. A short initial spin period, thought to be important for the very generation of the magnetic field, is shown to be quite relevant to the modified expansion of the remnant. We finally analyze all proposed associations on a case-by-case basis, addressing the likelihood of each one, according to this perspective. We consider a larger explosion energy and reassess the characteristic age issue, and conclude that ∼50% of the associations can be real, provided that soft gamma repeaters and anomalous X-ray pulsars are magnetars.