B-pop is a modular, Python-based software package designed to generate realistic populations of binary compact objects—such as binary neutron stars (BNS), neutron star–black hole binaries (NSBH), and binary black holes (BBH)—for use in gravitational-wave (GW) simulations, data analysis, and astrophysical inference. Its goal is to provide users with a flexible and transparent way to model the source populations expected to be observed by ground- and space-based GW detectors such as LIGO, Virgo, KAGRA, Einstein Telescope (ET), and LISA.
In the context of gravitational-wave astronomy, population models are essential for tasks like:
- Forecasting detection rates
- Evaluating selection biases
- Performing mock data analyses
- Interpreting observed distributions in terms of astrophysical or cosmological parameters
B-pop was developed with these goals in mind, offering both astrophysically motivated models and the ability to define custom parametric populations. It is designed to interface smoothly with forecasting and inference tools such as GWFish, Bilby, and GWToolbox, making it a key component in simulation pipelines for the GW community.
Key Features
Astrophysical Fidelity
- B-pop includes built-in models for mass, spin, and redshift distributions of binary systems, based on current theoretical predictions and observational constraints.
- Users can choose between phenomenological models (e.g. power-law + peak mass distributions) and population synthesis–informed templates.
- Support for varying source classes: BNS, NSBH, BBH, and mixed populations.
Modular Design
- The code is highly modular: each component (mass, spin, redshift, inclination, sky position, SNR thresholding) can be customized independently.
- Users can plug in their own distributions, such as metallicity-dependent star formation histories or exotic compact object mass spectra.
Redshift & Cosmology Integration
- Includes standard cosmological models (e.g., flat ΛCDM) and can be configured to compute source-frame vs. detector-frame properties.
- Properly handles volumetric rates and redshift evolution, enabling realistic mock catalog creation across cosmic time.
Output & Compatibility
- B-pop can generate mock catalogs of GW sources with parameters including:
- Component masses, spins, redshift, luminosity distance
- Inclination, polarization, sky location
- Source type and detectability flags
- Outputs are exportable in widely used formats (e.g., HDF5, CSV), ready for injection into pipelines like GWFish (for forecasting) or Bilby (for Bayesian inference).
- SNR thresholds and detector sensitivity curves can be incorporated to simulate selection effects.
Example Use Cases
Forecasting Detection Rates
B-pop can be used to simulate detection campaigns under various network configurations (e.g. LIGO-Virgo-KAGRA vs. ET+CE) and source population assumptions. Combined with GWFish, it enables users to calculate expected event counts, sky localization distributions, and parameter estimation performance as a function of astrophysical model and detector sensitivity.
Bias Studies & Selection Effects
It allows detailed study of selection biases that affect the inference of population parameters from observed samples. For instance, B-pop helps model how an underlying spin distribution might be distorted by detector sensitivity, noise, or selection thresholds.
EM Follow-up Optimization
By generating realistic mock populations, B-pop supports the planning of EM follow-up strategies, particularly for BNS and NSBH systems (from 2026, before we will provide public catalogs of GW sources produced by MOBSE, SEVN, COSMOrate), where EM emission is expected. When coupled with visibility tools (e.g., GWsky) and parameter estimation pipelines (e.g., GRANITE), it helps define what kinds of sources are most likely to yield counterparts.
Astrophysical Inference
Populations generated with B-pop can be used as priors or injected data for hierarchical Bayesian inference on merger rates, mass and spin distributions, redshift evolution, and even the neutron star equation of state (EoS), when used in conjunction with Bayesian samplers like Bilby or dynesty.
B-pop is a versatile, customizable, and user-friendly tool for generating populations of binary compact objects tailored to a wide range of gravitational-wave science goals. Whether used for forecasting, selection bias studies, or data-driven astrophysical inference, it offers an essential building block for realistic simulations in both current and future GW observatories. Its seamless integration with tools like GWFish and Bilby makes it a natural choice for researchers working on the frontiers of GW astrophysics and multi-messenger astronomy.