Beyond the Background: Gravitational-wave Anisotropy and Continuous Waves from Supermassive Black Hole Binaries

Abstract

Pulsar timing arrays have found evidence for a low-frequency gravitational-wave background (GWB). Assuming that the GWB is produced by supermassive black hole binaries (SMBHBs), the next gravitational-wave (GW) signals astronomers anticipate are continuous waves (CWs) from single SMBHBs and their associated GWB anisotropy. The prospects for detecting CWs and anisotropy are highly dependent on the astrophysics of SMBHB populations. Thus, information from single sources can break degeneracies in astrophysical models and place much more stringent constraints than the GWB alone. We simulate and evolve SMBHB populations, model their GWs, and calculate their anisotropy and detectability. We investigate how varying components of our semianalytic model, including the galaxy stellar mass function, the SMBH–host galaxy relation (M BH–M bulge), and the binary evolution prescription, impact the expected detections. The CW occurrence rate is greatest for few total binaries, high SMBHB masses, large scatter in M BH–M bulge, and long hardening times. The occurrence rate depends most on the binary evolution parameters, implying that CWs offer a novel avenue to probe binary evolution. The most detectable CW sources are in the lowest frequency bin for a 16.03 yr PTA, have masses from ∼109 to 1010 M ⊙, and are ∼1 Gpc away. The level of anisotropy increases with frequency, with the angular power spectrum over multipole modes ℓ varying in low-frequency C ℓ>0/C 0 from ∼5 × 10−3 to ∼2 × 10−1, depending on the model; typical values are near current upper limits. Observing this anisotropy would support SMBHB models for the GWB over cosmological models, which tend to be isotropic.

Spectral Variance in a Stochastic Gravitational-wave Background from a Binary Population

“…Recent work has attempted to account for binary population variance in the spectrum and to distinguish between astrophysical and cosmological signal origins. This has included comparing the spectra of cosmological models against a power law (Kaiser et al 2022;Afzal et al 2023), fitting Gaussian processes or neural networks that are trained on many realizations of simulated astrophysical GWBs to PTA data (Taylor et al 2017;Afzal et al 2023;Bonetti et al 2024), searching for anisotropy (Agazie et al 2023c) that may be frequency dependent (Gardiner et al 2024), quantifying the discreteness of the GWB at high frequencies (Agazie et al 2024), and testing the Gaussian ensemble model given a finite population of GW sources (Allen & Valtolina 2024).…”