The supermassive black hole masses of reverberation-mapped active galactic nuclei

TL;DR

This study calibrates supermassive black hole masses in active galactic nuclei using reverberation mapping data, finding that line dispersion in the mean spectrum provides the most accurate estimates, and refining virial factors improves mass calculations.

Contribution

It introduces a calibration of virial factors using multivariate regression and compares different velocity tracers for more accurate SMBH mass estimates in reverberation-mapped AGNs.

Findings

Line dispersion in mean spectrum yields smallest offset rms.

Calibrated virial factors reduce mass estimation errors.

Preferred velocity tracer for single-epoch spectra is $\sigma_{Heta,mean}$.

Abstract

Using different kinds of velocity tracers derived from the broad H$\beta$ profile (in the mean or rms spectrum) and the corresponding virial factors $f$, the central supermassive black hole (SMBH) masses ($M_{\rm BH}$) are calculated for a compiled sample of 120 reverberation-mapped (RM) AGNs. For its subsample of RM AGNs with measured stellar velocity dispersion ($\sigma_{\rm \ast}$), the multivariate linear regression technique is used to calibrate the mean value $f$, as well as the variable FWHM-based $f$. It is found that, whether excluding the pseudo-bulges or not, $M_{\rm BH}$ from the H$\beta$ line dispersion in the mean spectrum ($\sigma_{\rm H\beta,mean}$) has the smallest offset rms with respect to the $M_{\rm BH}-\sigma_{\ast}$ relation. For the total sample excluding SDSS-RM AGNs, with respect to $M_{\rm BH}$ from $\sigma_{\rm \ast}$ or that from the H$\beta$ line dispersion in the rms spectrum ($\sigma_{\rm H\beta,rms}$), it is found that we can obtain $M_{\rm BH}$ from the $\sigma_{\rm H\beta,mean}$ with the smallest offset rms of 0.38 dex or 0.23 dex, respectively. It implies that, with respect to the H$\beta$ FWHM, we prefer $\sigma_{\rm H\beta,mean}$ to calculate $M_{\rm BH}$ from the single-epoch spectrum. Using the FWHM-based $f$, we can improve $M_{\rm BH}$ calculation from FWHM(H$\beta$) and the mean $f$, with a decreased offset rms from 0.52 dex to 0.39 dex with respect to $M_{\rm BH}$ from $\sigma_{\rm \ast}$ for the subsample of 36 AGNs with $\sigma_{\rm \ast}$. The value of 0.39 dex is almost the same as that from $\sigma_{\rm H\beta,mean}$ and the mean $f$.