Extending the Frontier of Spatially-Resolved Supermassive Black Hole Mass Measurements to at $1\lesssim z\lesssim2$: Simulations with ELT/MICADO High-Resolution Mass Models and HARMONI Integral-Field Stellar Kinematics

TL;DR

This study addresses the challenge of obtaining direct dynamical SMBH masses beyond the local universe by simulating ELT observations with MICADO and HARMONI for five bright, massive, quiescent galaxies at $1\lesssim z\lesssim 2$. The authors construct intrinsic Sérsic models from HST data, generate realistic MICADO images and HARMONI IFS datacubes using SimCADO and HSIM, and extract kinematics with pPXF, feeding JAM dynamical models fitted via MCMC to recover $M_{\rm BH}$ with about 10% accuracy. They show MICADO can yield high-fidelity stellar mass models in ~1 h, while HARMONI requires longer integrations (roughly 5–7.5 h at $z\approx1$ and ~5 h at $1<z\lesssim 2$) to obtain robust kinematics, with BH signatures detectable within ~50 mas. The results demonstrate ELT-level capabilities to extend direct BH mass measurements to $z\approx 2$, enabling crucial tests of SMBH–galaxy co-evolution at the high-mass end and guiding future high-redshift SMBH demographics studies.

Abstract

Current spatially resolved kinematic measurements of supermassive black hole (SMBH) masses are largely confined to the local Universe (distances $\lesssim100$ Mpc). We investigate the potential of the Extremely Large Telescope's (ELT) first-light instruments, MICADO and HARMONI, to extend these dynamical measurements to galaxies at redshift $1\lesssim z\lesssim2$. We select a sample of five bright, massive, quiescent galaxies at these redshifts, adopting their Sérsic profiles from HST photometry as their intrinsic surface brightness distributions. Based on these intrinsic models, we generate mock MICADO images using SimCADO and mock HARMONI integral-field spectroscopic data cubes using HSIM. The HARMONI simulations utilize input stellar kinematics derived from Jeans Anisotropic Models (JAM). We then process these mock observations: the simulated MICADO images are fitted with Multi-Gaussian Expansion to derive stellar mass models, and stellar kinematics are extracted from mock HARMONI cubes with pPXF. Finally, these derived stellar mass models and kinematics are used to constrain JAM dynamical models within a Bayesian framework. Our analysis demonstrates that SMBH masses can be recovered with an accuracy of $\sim$10%. We find that MICADO can provide detailed stellar mass models with $\sim$1 hour of on-source exposure. HARMONI requires longer minimum integrations for reliable stellar kinematic measurements of SMBHs. The required on-source time scales with apparent brightness, ranging from 5-7.5 hours for galaxies at $z\approx1$ (F814W, 20-20.5 mag) to 5 hours for galaxies at $1<z\lesssim2$ (F160W, 20.8 mag). These findings highlight the ELT's capability to push the frontier of SMBH mass measurements to $z\approx2$, enabling crucial tests of SMBH-galaxy co-evolution at the top end of the galaxy mass function.

Figures (14)

• Figure 1: Angular-diameter distance ($D_A$) versus redshift ($z$) in a standard flat concordance $\Lambda$CDM cosmology, showing $D_A$ peaking at $\approx 1750$ Mpc around $z\approx1.6$.
• Figure 2: Our selected sample of five quiescent, massive, red-sequence galaxies (large colored stars with red name tags; \ref{['tab:sampledata']}) is shown in the mass–size diagram, separated into three redshift panels: $z \approx 1$ (left), $z \approx 1.5$ (middle), and $z \approx 2$ (right). These galaxies are overlaid on the 3D-HST CANDELS sample vanderWel14, which is divided into corresponding redshift intervals ($0.75 < z \leq 1.25$, $1.25 < z \leq 1.75$, and $1.75 < z \leq 2.25$) and spans a stellar mass range of $2 \times 10^9 <$$M_{\star}$$< 2 \times 10^{12}$$\mathrm{M}_\odot$. Red dashed inclined lines represent contours of constant stellar velocity dispersion from 50 to 350 km s$^{-1}$, estimated using the virial relation Cappellari2006Cappellari23Krajnovic18a.
• Figure 3: HST/WFC3 F814W and F160W images for five galaxies in our simulated sample are displayed in grayscale.
• Figure 4: Grayscale mock MICADO images of the five simulated galaxies, produced using SimCADO.
• Figure 5: The comparison between the MICADO images produced by SimCADO and their best-fit MGE models for five simulated galaxies is presented in terms of 2D surface brightness density. Black contours represent the data, while red contours depict the model, illustrating the alignment between data and model at corresponding radii and contour levels which are spaced decreasing by 0.5 mag arcsec$^{-1}$ outward.