Statistical evidence for massive black hole recoils in active galactic nuclei
Abstract
We search for a population-level signature of gravitational-wave recoiling supermassive black holes: a positive correlation between dust obscuration and the magnitude of the line-of-sight velocity offset of broad emission lines relative to the host. Using the SDSS DR16 quasar catalogue, we estimate the velocity offset, Δv, as the difference between the broad Hβ redshift and a noise-weighted redshift from narrow lines ([O III] 5007, [O II] 3728, and Ca II 3934). We adopt the redshift-relative colour excess Δ(g−i) as a proxy for dust column density. Analysing ∼105 quasars that meet basic spectral quality requirements, we find a modest but highly significant positive correlation between |Δv| and Δ(g−i) (Spearman r≃0.12 and Pearson r≃0.13, with p≪10−10 in both cases). The fraction of highly obscured quasars increases with |Δv|, indicating that the correlation is driven by a dust-reddened subpopulation. The result is robust to the choice of minimum |Δv| threshold and to the line redshift estimator (peak vs. centroid). As expected, the correlation is largely absent when velocity offsets are computed between narrow emission lines. We find systematic differences between redshifted and blueshifted subsamples, which may point to residual velocity biases or additional physical effects (e.g. winds, inflows, orientation-dependent obscuration, or asymmetric broad-line regions). Recoiling massive black holes provide a natural explanation for the observed correlation, but alternative scenarios should be explored. If confirmed, this would enable population-level constraints on massive black hole merger rates, recoil dynamics, and active galactic nuclei disc properties.
Versions
Related
- A Break in the X-Ray Loudness of Markarian 590: Evidence for an Active Galactic Nucleus Spectral State Transition?
- Contributions of massive stars and classical novae to Galactic Al 26
- Constraints on the Metallicity-dependent Explodability of Massive Stars from Galactic Chemical Evolution: Toward Alleviating the Red Supergiant Problem
- GALACTIC CENTERS AS TRACERS OF GALAXY EVOLUTION: Zooming from galaxy large-scale properties into their central massive black holes
- Machine learning-based classification of type 1 active galactic nuclei
BibTeX
@misc{becsy2026statistical,
title = {{Statistical evidence for massive black hole recoils in active galactic nuclei}},
author = {Bence Bécsy and Peter Raffai and Zoltán Haiman and Andor Budai and Zsolt Frei},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2605.04781},
eprint = {2605.04781},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.04781},
}
RIS
TY - UNPB
TI - Statistical evidence for massive black hole recoils in active galactic nuclei
AU - Bence Bécsy
AU - Peter Raffai
AU - Zoltán Haiman
AU - Andor Budai
AU - Zsolt Frei
PY - 2026
DA - 2026/09/10
PB - arXiv
DO - 10.48550/arxiv.2605.04781
UR - https://arxiv.org/abs/2605.04781
AB - We search for a population-level signature of gravitational-wave recoiling supermassive black holes: a positive correlation between dust obscuration and the magnitude of the line-of-sight velocity offset of broad emission lines relative to the host. Using the SDSS DR16 quasar catalogue, we estimate the velocity offset, Δv, as the difference between the broad Hβ redshift and a noise-weighted redshift from narrow lines ([O III] 5007, [O II] 3728, and Ca II 3934). We adopt the redshift-relative colour excess Δ(g−i) as a proxy for dust column density. Analysing ∼10^5 quasars that meet basic spectral quality requirements, we find a modest but highly significant positive correlation between |Δv| and Δ(g−i) (Spearman r≃0.12 and Pearson r≃0.13, with p≪10^−10 in both cases). The fraction of highly obscured quasars increases with |Δv|, indicating that the correlation is driven by a dust-reddened subpopulation. The result is robust to the choice of minimum |Δv| threshold and to the line redshift estimator (peak vs. centroid). As expected, the correlation is largely absent when velocity offsets are computed between narrow emission lines. We find systematic differences between redshifted and blueshifted subsamples, which may point to residual velocity biases or additional physical effects (e.g. winds, inflows, orientation-dependent obscuration, or asymmetric broad-line regions). Recoiling massive black holes provide a natural explanation for the observed correlation, but alternative scenarios should be explored. If confirmed, this would enable population-level constraints on massive black hole merger rates, recoil dynamics, and active galactic nuclei disc properties.
ER -