Detecting gravitational-wave bursts from black hole binaries in the Galactic Center with LISA

TL;DR

This paper investigates how highly eccentric stellar-mass black hole binaries near the galactic center emit gravitational-wave bursts detectable by LISA, and proposes an unmodeled wavelet-based method to analyze their signals and infer binary eccentricity.

Contribution

It introduces a wavelet decomposition approach for detecting and characterizing GW bursts from eccentric BHBs near Sgr A* in LISA data, enabling eccentricity estimation without templates.

Findings

Wavelet parameters can infer binary eccentricity with 20% error.

Unmodeled wavelet approach effectively captures burst time-frequency properties.

Method constrains parameter space for Bayesian signal reconstruction.

Abstract

Stellar-mass black hole binaries (BHBs) in galactic nuclei are gravitationally perturbed by the central supermassive black hole (SMBH) of the host galaxy, potentially inducing strong eccentricity oscillations through the eccentric Kozai-Lidov (EKL) mechanism. These highly eccentric binaries emit a train of gravitational-wave (GW) bursts detectable by the Laser Interferometer Space Antenna (LISA) -- a planned space-based GW detector -- with signal-to-noise ratios (SNRs) up to ${\sim}100$ per burst. In this work, we study the GW signature of BHBs orbiting our galaxy's SMBH, Sgr A$^*$, which are consequently driven to very high eccentricities. We demonstrate that an unmodeled approach using a wavelet decomposition of the data effectively yields the time-frequency properties of each burst, provided that the GW frequency peaks between $10^{-3}\,\,\mathrm{Hz}$--$10^{-1}\,\,\mathrm{Hz}$. The wavelet parameters may be used to infer the eccentricity of the binary, measuring $\log_{10}(1-e)$ within an error of $20\%$. Our proposed search method can thus constrain the parameter space to be sampled by complementary Bayesian inference methods, which use waveform templates or orthogonal wavelets to reconstruct and subtract the signal from LISA data.