GWKokab: An Implementation to Identify the Properties of Multiple Population of Gravitational Wave Sources

TL;DR

The paper introduces GWKokab, a JAX-based framework for scalable, modular inference of population properties for multiple gravitational-wave source subpopulations with independent rates. It combines hierarchical Bayesian modeling with accelerated sampling via flowMC and compatibility with NumPyro, enabling subpopulation analyses across mass, spin, eccentricity, and redshift distributions. The authors validate GWKokab on synthetic spinning eccentric BBHs, circular multi-source populations, and by reproducing prior GW-population results from GWTC-3, demonstrating both accuracy and substantial computational speedups. The work provides a practical, open-source tool for detailed population analyses and future formation-channel studies in gravitational-wave astrophysics.

Abstract

The rapidly increasing sensitivity of gravitational wave detectors is enabling the detection of a growing number of compact binary mergers. These events are crucial for understanding the population properties of compact binaries. However, many previous studies rely on computationally expensive inference frameworks, limiting their scalability. In this work, we present GWKokab, a JAX-based framework that enables modular model building with independent rate for each subpopulation such as BBH, BNS, and NSBH binaries. It provides accelerated inference using the normalizing flow based sampler called flowMC and is also compatible with NumPyro samplers. To validate our framework, we generated two synthetic populations, one comprising spinning eccentric binaries and the other circular binaries using a multi-source model. We then recovered their injected parameters at significantly reduced computational cost and demonstrated that eccentricity distribution can be recovered even in spinning eccentric populations. We also reproduced results from two prior studies: one on non-spinning eccentric populations, and another on the BBH mass distribution using the third Gravitational Wave Transient Catalog (GWTC-3). We anticipate that GWKokab will not only reduce computational costs but also enable more detailed subpopulation analyses such as their mass, spin, eccentricity, and redshift distributions in gravitational wave events, offering deeper insights into compact binary formation and evolution.

Figures (8)

• Figure 1: Non-Spinning Eccentric BBHs: The corner plot shows the recovery of population parameters on the same dataset used in previous study 2024PhRvD.110f3009Z, and inference made by GWKokab at 2% computational cost as compare to the previous publication.
• Figure 2: Spinning Eccentric Population: The corner plot of the recovered population model hyperparameters $\Lambda$. The two different colors show the different error types in the true values. The blue one is showing the recovery using fake PEs described in appendix \ref{['subsection:syn-pop-uncertainties']} and the orange one is showing the recovery using delta error PEs.
• Figure 3: Multi-Source Population: The left figure shows multi-population injections for BBH power law(blue), BBH peak (orange), BNS (green), and NSBH (red) with their independent rates. The right figure shows schematic diagram of multi-source model, each source type has its separate spin and rate. The neutron stars in NSBH have the same spin distribution as neutron stars in BNS.
• Figure 4: Multi-Source Population: The left corner plot shows the some recovered shape parameters of multi-source model and right corner plot shows independent rate parameters recovery for BBH, BBH Peak, BNS, and NSBH.
• Figure 5: Multi-Source Population: The left PPD plot shows the primary mass distribution, it highlights the recovered features of injected population such as mass gap and peaks produced by BNS, NSBH and BBH(gaussian). The right PPD shows the secondary mass distribution, and we can see the peak around 30 $M_{\odot}$ produced by BBH(gaussian) subpopulation.