A Search for Supermassive Black Hole Binary Candidates in 46-Year Radio Light Curves of 83 Blazars

TL;DR

This study analyzes 46-year radio light curves of 83 blazars to search for supermassive black hole binary candidates, identifying one promising candidate with a 17.9-year periodicity, while highlighting the challenges of false positives due to spectral properties.

Contribution

It introduces a comprehensive analysis of long-term radio light curves for SMBHB candidate identification, emphasizing the importance of spectral effects in periodicity detection.

Findings

Identified one new SMBHB candidate with a 17.9-year period.

Most apparent periodicities are due to spectral slope effects, not true binaries.

Approximately 2.4% of the sample are SMBHB candidates.

Abstract

The combined University of Michigan Radio Astronomy Observatory (UMRAO) and Owens Valley Radio Observatory (OVRO) blazar monitoring programs at 14.5/15 GHz provide uninterrupted light curves of $\sim~46-50$ yr duration for 83 blazars, selected from amongst the brightest and most rapidly flaring blazars north of declination $-20^\circ$. In a search for supermassive black hole binary (SMBHB) candidates, we carried out tests for periodic variability using generalized Lomb-Scargle (GLS), weighted wavelet-Z (WWZ), and sine-wave fitting (SWF) analyses of this sample. We used simulations to test the effects of the power law spectrum of the power spectral density (PSD) on our findings, and show that the irregular sampling in the observed light curves has very little effect on the GLS spectra. Apparent periodicities and putative harmonics appear in all 83 of the GLS spectra of the blazars in our sample. We tested the reality of these apparent periodicities and harmonics with simulations, and found that in the overwhelming majority of cases they are due to the steep slope of the PSD, and should therefore be treated with great caution. We find one new SMBHB candidate: PKS 1309+1154, which exhibits a 17.9 year periodicity. The fraction of SMBHB candidates in our sample is $2.4_{-0.8}^{+3.2}\%$.