Modelling spin evolution of magnetars

Abstract

ABSTRACT The origin and fate of magnetars [young, extremely magnetized neutron stars (NSs)] remains unsolved. Probing their evolution is therefore crucial for investigating possible links to other species of isolated NSs, such as the X-ray dim NSs (XDINSs) and rotating radio transients (RRATs). Here, we investigate the spin evolution of magnetars. Two avenues of evolution are considered: one with exponentially decaying B-fields, the other with sub- and superexponential decay. Using Monte Carlo methods, we synthesize magnetar populations using different input distributions and physical parameters, such as for the initial spin period, its time derivative, and the B-field decay time-scale. Additionally, we introduce a fade-away procedure that can account for the fading of old magnetars, and we briefly discuss the effect of alignment of the B-field and spin axes. Imposing the Galactic core-collapse supernova rate of ∼20 kyr−1 as a strict upper limit on the magnetar birthrate and comparing the synthetic populations to the observed one using both manual and automatic optimization algorithms for our input parameter study, we find that the B-field must decay exponentially or superexponentially with a characteristic decay time-scale of 0.5−10 kyr (with a best value of ∼4 kyr). In addition, the initial spin period must be less than 2 s. If these constraints are kept, we conclude that there are multiple choices of input physics that can reproduce the observed magnetar population reasonably well. We also conclude that magnetars may well be evolutionary linked to the population of XDINSs, whereas they are in general unlikely to evolve into RRATs.