The host galaxies and merger environments of short and long gamma-ray bursts producing kilonovae

TL;DR

This study examines the host galaxies of nine kilonova-associated GRBs, revealing diverse morphologies and recent merger activity, challenging traditional progenitor models and emphasizing detailed host analysis.

Contribution

It provides the first detailed morphological analysis of kilonova-associated GRB hosts, highlighting the role of recent galaxy mergers and the importance of host decomposition.

Findings

Five hosts show tidal features indicating recent mergers.

No clear morphological difference between short and long GRB hosts.

Host decomposition affects offset estimates and interpretation.

Abstract

Gamma-ray bursts (GRBs) have traditionally been classified by their prompt emission duration and spectral hardness, with short GRBs (sGRB; $\lesssim2 \ \rm{s}$) originating from compact object mergers and long GRBs (LGRB; $\gtrsim2 \ \rm{s}$) from massive star core-collapse. Recent kilonova (KN) associations with long-duration GRBs have challenged this standard picture. We analyze the host galaxies of nine GRBs with associated kilonova candidates at $z<0.6$, including five sGRB-KNe and four LGRB-KNe. Using both parametric and non-parametric modeling of the host light distributions, we investigate the progenitor environments of these events and test whether their hosts show evidence for recent galaxy interactions that could favor dynamical formation channels or isolated pathways following merger-driven star formation episodes for neutron star binaries. We find that five of the nine hosts display tidal features that show they have likely undergone recent mergers, suggesting that merger-driven, dynamical formation pathways may contribute in some systems. We find no clear morphological distinction between sGRB-KN and LGRB-KN hosts as both populations span a wide range of morphologies, including ellipticals, spirals, and interacting systems with tidal features. Multi-S\'ersic modeling of the host light profiles further shows that host-normalized offsets inferred from single-S\'ersic fits can be overestimated when the transient is associated with a specific subcomponent of a complex host light profile. These results highlight the importance of decomposing host morphology into physically relevant components when interpreting GRB environments and galactocentric offsets.