Inference on inner galaxy structure via gravitational waves from supermassive binaries

Technical Summary: Inference on Inner Galaxy Structure via Gravitational Waves from Supermassive Binaries

Problem Statement
The detection of a stochastic gravitational wave background (SGWB) by pulsar-timing arrays (PTAs), specifically the NANOGrav 15-year dataset, indicates a population of supermassive black hole binaries (SMBHBs). While the observed spectrum generally aligns with predictions for circular binaries driven by gravitational wave (GW) emission ($h_c \propto f^{-2/3}$), the data exhibit a mild preference for a spectral turnover at the lowest observed frequencies (around 4 nHz). This feature suggests an acceleration in orbital hardening prior to the GW-dominated regime, potentially resolving the "final parsec problem." The physical origin of this turnover is hypothesized to be environmental interactions—specifically, the ejection of stars and dark matter particles via gravitational three-body slingshots—which extract orbital energy more efficiently than two-body dynamical friction. However, the specific properties of the galactic-center matter distribution (density and slope) that drive this process, and their interplay with initial binary eccentricity, remain unconstrained.

Methodology
The authors model the co-evolution of SMBHB orbital parameters (semimajor axis $a$ and eccentricity $e$) and the surrounding matter density profile. The analysis integrates two primary mechanisms:

1. Gravitational Wave Emission: Governs orbital decay and circularization, particularly at high frequencies.
2. Three-Body Scattering: Describes the hardening of the binary orbit due to interactions with background stars and dark matter. The rate of hardening is parameterized by the local matter density $\rho$ and velocity dispersion $\sigma$.

The study assumes a power-law density profile for the galactic center prior to scouring: $\rho(r) = \rho_{pc} (r/1 \text{ pc})^{-\gamma}$, where $\rho_{pc}$ is the density normalization and $\gamma$ is the radial slope. The influence radius $r_i$ is defined where the enclosed mass equals twice the binary mass. The authors calculate the turnover frequency $f_t$ where three-body scattering transitions to GW dominance, which depends on $\rho_i$, $\sigma_i$, the binary mass $M$, and mass ratio $q$.

To infer constraints, the authors:

• Adopted the NANOGrav 15-year dataset, focusing on the lowest five frequency bins (2–10 nHz) where the signal-to-noise ratio is robust.
• Modeled the SMBHB population using a fiducial astrophysical distribution (mass, redshift, mass ratio) derived from previous NANOGrav analyses.
• Performed a Bayesian inference to constrain the joint parameter space of initial eccentricity ($e_0$), density normalization ($\rho_{pc}$), and slope ($\gamma$).
• Tested alternative parameterizations, including power-law distributions for initial eccentricity and Gaussian distributions for density variance, to ensure robustness.

Key Results

• Density Constraints: The analysis favors a parsec-scale galactic-center matter density of approximately $\rho_{pc} \sim 10^6 M_\odot \text{pc}^{-3}$ across most of the parameter space. The posterior distribution for $\log_{10}(\rho_{pc}/10^5 M_\odot \text{pc}^{-3})$ is estimated at $0.6^{+1.4}_{-2.6}$ (1$\sigma$).
• Eccentricity-Density Degeneracy: A degeneracy exists between initial eccentricity ($e_0$) and density ($\rho_{pc}$). A lower density requires an extremely high initial eccentricity ($e_0 > 0.999$) to reproduce the observed low-frequency turnover, as lower densities imply a larger influence radius where GW emission would otherwise circularize the orbit before reaching the observed frequencies.
• Profile Slope: Flatter density profiles (smaller $\gamma$) are preferred over steeper ones. Steeper profiles lead to a broader distribution of turnover frequencies across the SMBHB population, requiring a normalization factor inconsistent with the fiducial model.
• Benchmark Comparisons: The favored density range naturally includes core-like stellar distributions (e.g., the Milky Way core and M87 stellar distribution) but disfavors steep dark matter spikes (e.g., adiabatic spikes in M87), as the latter result in lower densities at the influence radius and thus fail to produce the observed turnover without extreme eccentricities.
• Robustness: Alternative parameterizations allowing for distributions in eccentricity and density do not alter the primary conclusion that the parsec-scale density is of the order $10^6 M_\odot \text{pc}^{-3}$.

Significance and Claims
The paper claims that current PTA data, despite marginal sensitivity, already encode measurable information regarding the matter distribution on parsec scales in galactic centers. By jointly modeling eccentricity and environmental hardening, the authors demonstrate that the low-frequency turnover in the SGWB spectrum can be used to infer the pre-scouring density profile of galactic nuclei.

The study highlights that:

1. Three-body scattering is a dominant mechanism for orbital hardening in the PTA frequency band, particularly for comparable-mass binaries.
2. The observed spectral features provide a constraint on the combined matter density (stars and dark matter) near galaxy centers, suggesting densities higher than some previous expectations for certain environments.
3. If dark matter is the dominant component, these findings offer potential insights into the particle nature of dark matter and its density profile at galactic centers, though specific non-cold dark matter models require further evaluation.
4. Future observations from next-generation PTAs (e.g., FAST, DSA-2000, SKA) and astrometry missions are expected to provide more precise spectral measurements, potentially resolving the degeneracy between eccentricity and density to precisely determine the galactic-center environment.

The authors emphasize that while they set a stringent upper limit on the galactic-center density distribution, precise determination of the density profile requires resolving the degeneracy with initial eccentricity, potentially through individual binary resolution or correlation analysis across frequency bins.