The Galaxy Activity, Torus, and Outflow Survey (GATOS) - VI. Black hole mass estimation using machine learning

Q: What is the main goal of the GATOS study in this work?

To more precisely and accurately determine supermassive black hole masses in the circum-nuclear regions of GATOS galaxies than estimates from scaling relations.

Q: How is ALMA used to support black hole mass estimation here?

ALMA high spatial resolution data resolve CO(3–2) emission within <100 pc around the central SMBH, enabling analysis when sufficient nuclear gas is detected.

Q: What method does the paper introduce for estimating black hole masses?

A supervised machine-learning approach that estimates BH mass from ALMA-derived position–velocity diagrams or first-moment maps, trained and validated using numerical simulations with a wide parameter range.

Aug 4, 2026

Upload by Theodore

The feeding and feedback mechanisms of active galactic nuclei (AGNs) remain insufficiently understood, particularly for low-luminosity and obscured systems where precise central black hole (BH) masses are hard to obtain. This work uses the GATOS sample to refine BH mass estimates in the circum-nuclear regions, leveraging ALMA’s high spatial resolution to resolve CO(3–2) emission within <100 pc. Seven nearby galaxies with adequate nuclear gas enable supervised machine-learning estimates, supported by training with numerical simulations and comparisons to prior scaling relations and activity-plane methods.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages18

Reading Time45 min

AccessFree to download

Abstract

Introduction

Active galactic nuclei and black hole scaling relations

Motivation from galaxy evolution and feedback modes

Radiative and kinetic AGN feedback

A&A, 693, A311 (2025) [https:](https://doi.org/10.1051/0004-6361/202347566)[//](https://doi.org/10.1051/0004-6361/202347566)[doi.org](https://doi.org/10.1051/0004-6361/202347566)[/](https://doi.org/10.1051/0004-6361/202347566)[10.1051](https://doi.org/10.1051/0004-6361/202347566)[/](https://doi.org/10.1051/0004-6361/202347566)[0004-6361](https://doi.org/10.1051/0004-6361/202347566)[/](https://doi.org/10.1051/0004-6361/202347566)[202347566](https://doi.org/10.1051/0004-6361/202347566) The Authors 2025 &~~t~~ronomy~~t~~rop~~h~~ys~~i~~cs The Galaxy Activity, Torus, and Outﬂow Survey (GATOS) VI. Black hole mass estimation using machine learning R. Poitevineau 1,2 ; ? , F. Combes 1,3, S. Garcia-Burillo4, D. Cornu 1, A. Alonso Herrero5 , C. Ramos Almeida6 ,7 , A. Audibert6 ,7 , E. Bellocchi8 ,9 , P. G. Boorman 10 , A. J. Bunker 11 , R. Davies 12 , T. Díaz-Santos 13 , 14 , I. García-Bernete 11 , B. García-Lorenzo6 ,7 , O. González-Martín 15 , E. K. S. Hicks 16 , S. F. Hönig 17 , L. K. Hunt 18 , M. Imanishi 19 ,20 , M. Pereira-Santaella21 , C. Ricci22 ,23 , D. Rigopoulou 11 , 14 , D. J. Rosario24 , D. Rouan25 , M. Villar Martin4 , and M. Ward26 (A􀀎liations can be found after the references) Received 26 July 2023 / Accepted 23 November 2024 ABSTRACT The detailed feeding and feedback mechanisms of active galactic nuclei (AGNs) are not yet well known. For low-luminosity AGNs, obscured AGNs, and late-type galaxies, the masses of their central black holes (BH) are di􀀎cult to determine precisely. Our goal with the GATOS sample is to study the circum-nuclear regions and, in the present work, to better determine their BH mass, with more precise and accurate estimations than those obtained from scaling relations. We used the high spatial resolution of ALMA to resolve the CO(3–2) emission within 􀀘 100 pc around the supermassive black hole (SMBH) of seven GATOS galaxies and try to estimate their BH mass when enough gas is present in the nuclear regions. We studied the seven bright (LAGN (14􀀀150 keV) 􀀕 1042 erg=s) and nearby (<28 Mpc) galaxies from the GATOS core sample. For the sake of comparison, we ﬁrst searched the literature for previous BH mass estimations. We also made additional estimations using the MBH–􀀛 relation and the fundamental plane of BH activity. We developed a new method using supervised machine learning to estimate the BH mass either from position-velocity diagrams or from ﬁrst-moment maps computed from ALMA CO(3–2) observations. We used numerical simulations with a large range of parameters to create the training, validation, and test sets. Seven galaxies had su􀀎cient gas detected, thus, we were able to make a BH estimation from the ALMA data: NGC 4388, NGC 5506, NGC 5643, NGC 6300, NGC 7314, NGC 7465, and NGC 7582 . Our BH masses range from 6.39 to 7.18 log(MBH =M􀀌 ) and are consistent with the previous estimations. In addition, our machine learning method has the advantage of providing a robust estimation of errors with conﬁdence intervals. The method has also more growth potential than scaling relations. This work represents the ﬁrst step toward an automatized method for estimating MBH using machine learning. Key words. galaxies: active – galaxies: ISM – galaxies: kinematics and dynamics – galaxies: nuclei – galaxies: spiral 1. Introduction The centers of massive galaxies host supermassive black holes (SMBHs) and black holes (BHs) with masses ranging from 106 M􀀌 to 10 10 M􀀌 . Those BHs have grown together with their host galaxies over time by accreting matter. The release of this considerable gravitational energy produces shocks and emits powerful radiation in the central region of the galaxies, a phenomenon known as active galactic nuclei (AGNs) . There is an empirical relation between the mass of the SMBH (MBH ) and the bulge mass of their galaxy host, the latter often measured by its central velocity dispersion, the MBH–􀀛 relation (e.g., Kormendy & Ho 2013 ; Shankar et al. 2016, 2019 ; Mar

FAQs

What is the main goal of the GATOS study in this work?

To more precisely and accurately determine supermassive black hole masses in the circum-nuclear regions of GATOS galaxies than estimates from scaling relations.

How is ALMA used to support black hole mass estimation here?

ALMA high spatial resolution data resolve CO(3–2) emission within <100 pc around the central SMBH, enabling analysis when sufficient nuclear gas is detected.

What method does the paper introduce for estimating black hole masses?

A supervised machine-learning approach that estimates BH mass from ALMA-derived position–velocity diagrams or first-moment maps, trained and validated using numerical simulations with a wide parameter range.

Aug 4, 2026

Upload by Theodore