Binary neutron star merger offsets from their host galaxies

II. Short-duration gamma-ray bursts

¹ Department of Astrophysics/IMAPP, Radboud University, P.O. Box 9010 6500 GL Nijmegen, The Netherlands
² Department of Physics, University of Warwick, Coventry CV4 7AL, UK
³ European Space Agency (ESA), European Space Research and Technology Centre (ESTEC), Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands
⁴ Center for Astrophysics | Harvard & Smithsonian, 60 Garden St., Cambridge, MA 02138, USA
⁵ Center for Interdisciplinary Exploration and Research in Astrophysics (CIERA) and Department of Physics and Astronomy, Northwestern University, Evanston, IL 60208, USA

^⋆ Corresponding author: nicola.gaspari@live.it

Received: 7 April 2025
Accepted: 26 May 2025

Abstract

Context. The mergers of binary neutron stars (BNSs) and neutron star–black hole (NSBH) binaries have long been linked to short-duration gamma-ray bursts (SGRBs). However, despite their stellar progenitors, SGRBs are often found outside the stellar light of the host galaxy. This is commonly attributed to supernova kicks, which displace the SGRB progenitors from the original stellar population.

Aims. Our goal is to use stellar population synthesis models to reproduce and interpret the observed offsets of a statistical sample of SGRBs, using realistic galactic models based on the observed host properties.

Methods. We derived the host galaxy potentials from the observed properties on a case-by-case basis and simulated the galactic trajectories of synthetic BNSs and NSBHs from the BPASS code using three different kick prescriptions. We compared predicted and observed offsets to investigate the impact of velocity kicks, host galaxy types, and host association criteria.

Results. The locations of the SGRB population are consistent with the expectations of kicked BNS or BHNS progenitors, implying that such mergers are the dominant (and perhaps the only) progenitor system. Predictions for NSBHs provide a significantly worse fit compared to BNSs, while we find no significant difference when comparing different kick prescriptions. For late-type hosts, we find the best agreement when including hosts with a probability of chance alignment, P_ch, of up to 20%, while lower P_ch thresholds lead us to overestimate SGRB offsets. We argue that P_ch is biased against viable hosts at the largest offsets and suggest the use of less conservative P_ch thresholds for late-type hosts. For early-type hosts, the predictions underestimate SGRB offsets in a few cases regardless of the P_ch threshold applied. We argue that this is likely due to the models missing galaxy evolution or spurious host associations.