The black hole mass of the $z=2.805$ multiply imaged quasar SDSS J2222+2745 from velocity-resolved time lags of the CIV emission line

TL;DR

This study measures the black hole mass of a high-redshift quasar using velocity-resolved reverberation mapping over 4.5 years, providing new insights into black hole properties and calibration at early cosmic times.

Contribution

First velocity-resolved reverberation mapping of a $z=2.805$ quasar, establishing the $r_{ m BLR}-L$ relation and black hole mass at high redshift.

Findings

Measured CIV line lag of 36.5 days in the quasar.

Velocity-resolved lags consistent with circular Keplerian orbits.

Derived black hole mass of approximately 4.3 billion solar masses.

Abstract

We present the first results of a 4.5 year monitoring campaign of the three bright images of multiply imaged $z=2.805$ quasar SDSS J2222+2745 using the Gemini North Multi-Object Spectrograph (GMOS-N) and the Nordic Optical Telescope (NOT). We take advantage of gravitational time delays to construct light curves surpassing 6 years in duration and achieve average spectroscopic cadence of 10 days during the 8 months of visibility per season. Using multiple secondary calibrators and advanced reduction techniques, we achieve percent-level spectrophotometric precision and carry out an unprecedented reverberation mapping analysis, measuring both integrated and velocity-resolved time lags for CIV. The full line lags the continuum by $\tau_{\rm cen} = 36.5^{+2.9}_{-3.9}$ rest-frame days. We combine our measurement with published CIV lags and derive the $r_{\rm BLR}-L$ relationship $\log_{10}( \tau / {\rm day}) = (1.00\pm 0.08) + (0.48\pm 0.04) \log_{10}[\lambda L_\lambda(1350{\r{A}})/10^{44}~{\rm erg ~s}^{-1}]$ with 0.32$\pm$0.06 dex intrinsic scatter. The velocity-resolved lags are consistent with circular Keplerian orbits, with $\tau_{\rm cen} = 86.2^{+4.5}_{-5.0}$, $25^{+11}_{-15}$, and $7.5^{+4.2}_{-3.5}$ rest-frame days for the core, blue wing, and red wing, respectively. Using $\sigma_{\rm line}$ with the mean spectrum and assuming $\log_{10} (f_{{\rm mean},\sigma}) = 0.52 \pm 0.26$, we derive $\log_{10}(M_{\rm BH}/M_{\odot}) = 8.63 \pm 0.27$. Given the quality of the data, this system represents a unique benchmark for calibration of $M_{\rm BH}$ estimators at high redshift. Future work will present dynamical modeling of the data to constrain the virial factor $f$ and $M_{\rm BH}$.