Constraining Binary Neutron Star Populations using Short Gamma-Ray Burst Observations

Abstract

The landmark multi-messenger observations of the binary neutron star (BNS) merger GW170817 provided firm evidence that such mergers can produce short gamma-ray bursts (sGRBs). However, the limited number of BNS detections by current gravitational-wave (GW) observatories raises the question of whether BNS mergers alone can account for the full observed sGRB population. We analyze a comprehensive set of 64 BNS population synthesis models with a Monte Carlo-based framework to reproduce the properties of sGRBs detected by Fermi-GBM over the past 16 years. We consider three jet geometry scenarios: a universal structured jet calibrated to GW170817, a universal top-hat jet, and a non-universal top-hat jet with distributions of core opening angles. Our results show that models characterized by low local BNS merger rates ($R_{BNS}(0) \lesssim 50$ Gpc$^{-3}$ yr$^{-1}$) predict too few observable sGRBs to reproduce the Fermi-GBM population, effectively disfavoring them as sole progenitors. Even when relaxing assumptions on jet geometry, low-rate models remain viable only for wide jets ($θ_c \ge 15^\circ$), in tension with the narrow jet cores ($θ_c \approx 6^\circ$) inferred from sGRB afterglow observations. In contrast, models with local merger rates of order $R_{BNS}(0) \approx 100$ Gpc$^{-3}$ yr$^{-1}$ successfully reproduce the observed sGRB population, assuming a plausible fraction of BNS mergers launch relativistic jets and realistic jet geometries. This analysis highlights the power of combining GW observations of BNS mergers with electromagnetic observations of sGRBs to place robust constraints on the BNS merger population and to assess their role as progenitors of sGRBs.

Figures (20)

• Figure 1: 1D marginalized posterior probability distributions of the jet fraction $f_j$ obtained for each considered BNS population synthesis model, assuming a universal structured jet calibrated to GW170817/GRB 170817A. In each plot, the median is indicated by a black line, with its value and 90% C.I. reported below. The common envelope efficiency $\alpha_{CE}$ is given on the y-axis. The x-axis shows the different models. The color bar indicates the local merger rate of each model.
• Figure 2: Jet fraction statistics for the universal structured jet model assuming the GW170817/GRB170817A structure. Top row: Median jet fraction $f_j$ versus the predicted local BNS merger rate density $R_\text{BNS}(0)$ (left) and total integrated rate $\Lambda$ (right). Bottom row: Quantile of the posterior distribution of $f_j$ at $f_j=1$, as a function of local BNS rate density (left) and total integrated rate $\Lambda$ (right). Symbols are colored according to the $\alpha_{CE}$ parameter, and each symbol denotes a different population. The shaded vertical region indicates the 90% credible interval for the BNS merger rate from GWTC-4 GWTC4. The cited fractions from the works of Ronchini2022 and Loffredo_2025 are given as R22 and L25 for comparison.
• Figure 3: Same as Fig. \ref{['fig:fj_ridge']}, but with the 1D marginalized distributions of $\epsilon = f_j (1 - \cos(\theta_c))$ for all the BNS populations, considering the universal top-hat jet model. For clarity $\log(\epsilon)$ is shown.
• Figure 4: Same as Fig. \ref{['fig:fj_ridge']} however assuming a universal top-hat jet with three different opening angles $\theta_c = 5^\circ$, $10^\circ$, and $20^\circ$. These posteriors are obtained from the conditioned distribution $P(f_j\mid \theta_c)$ (see Fig. \ref{['fig:epsilon_posteriors']}). The model LK with $\alpha_{CE}=0.5$ does not have any samples below $f_j \leq 10$ for $\theta_c = 5^\circ$.
• Figure 5: Same as the bottom row of Fig. \ref{['fig:universal_jet_combined']}, assuming a universal top-hat with three different aperture angles.