English

The Galactic population of magnetars : a simulation-based inference study

High Energy Astrophysical Phenomena 2025-06-16 v1

Abstract

Population synthesis modeling of the observed dynamical and physical properties of a population is a highly effective method for constraining the underlying birth parameters and evolutionary tracks. In this work, we apply a population synthesis model to the canonical magnetar population to gain insight into the parent population. We utilize simulation-based inference to reproduce the observed magnetar population with a model which takes into account the secular evolution of the force-free magnetosphere and magnetic field decay simultaneously and self-consistently. Our observational constraints are such that no magnetar is detected through their persistent emission when convolving the simulated populations with the XMM-Newton EPIC-pn Galactic plane observations, and that all of the \sim30 known magnetars are discovered through their bursting activity in the last 50\sim50 years. Under these constraints, we find that, within 95 % credible intervals, the birth rate of magnetars to be 1.80.6+2.61.8^{+2.6}_{-0.6} kyr1^{-1}, and lead to having 10.74.4+18.810.7^{+18.8}_{-4.4} % of neutron stars born as magnetars. We also find a mean magnetic field at birth (μb\mu_b is in T) log(μb)=10.20.2+0.1\log\left(\mu_b\right) = 10.2^{+0.1}_{-0.2}, a magnetic field decay slope αd=1.91.3+0.9\alpha_d = 1.9 ^{+0.9}_{-1.3}, and timescale τd=17.914.5+24.1\tau_d = 17.9^{+24.1}_{-14.5} kyr, in broad agreement with previous estimates. We conclude this study by exploring detection prospects: an all-sky survey with XMM-Newton would potentially allow to get around 7 periodic detections of magnetars, with approximately 150 magnetars exceeding XMM-Newton's flux threshold, and the upcoming AXIS experiment should allow to double these detections.

Cite

@article{arxiv.2503.11875,
  title  = {The Galactic population of magnetars : a simulation-based inference study},
  author = {Mattéo Sautron and Alexander Eli McEwen and George Younes and Jérôme Pétri and Paz Beniamini and Daniela Huppenkothen},
  journal= {arXiv preprint arXiv:2503.11875},
  year   = {2025}
}

Comments

This paper was submitted to ApJ, the two first authors contributed equally

R2 v1 2026-06-28T22:21:24.932Z