Supermassive black holes and their host spheroids I. Disassembling galaxies

TL;DR

This study performs detailed galaxy decompositions using Spitzer imaging to better understand the relationships between supermassive black holes and their host spheroids, addressing previous inconsistencies in models.

Contribution

It introduces a comprehensive, multi-component decomposition method for 66 galaxies, including spiral types, with careful analysis and a new prescription for estimating uncertainties.

Findings

Largest sample with 66 galaxies analyzed

1D decomposition techniques are more advantageous for large galaxies

Developed a new method for uncertainty estimation in galaxy modeling

Abstract

Several recent studies have performed galaxy decompositions to investigate correlations between the black hole mass and various properties of the host spheroid, but they have not converged on the same conclusions. This is because their models for the same galaxy were often significantly different and not consistent with each other in terms of fitted components. Using $3.6 \rm ~\mu m$ $Spitzer$ imagery, which is a superb tracer of the stellar mass (superior to the $K$-band), we have performed state-of-the-art multicomponent decompositions for 66 galaxies with directly measured black hole masses. Our sample is the largest to date and, unlike previous studies, contains a large number (17) of spiral galaxies with low black hole masses. We paid careful attention to the image mosaicking, sky subtraction and masking of contaminating sources. After a scrupulous inspection of the galaxy photometry (through isophotal analysis and unsharp masking) and - for the first time - 2D kinematics, we were able to account for spheroids, large-scale, intermediate-scale and nuclear disks, bars, rings, spiral arms, halos, extended or unresolved nuclear sources and partially depleted cores. For each individual galaxy, we compared our best-fit model with previous studies, explained the discrepancies and identified the optimal decomposition. Moreover, we have independently performed 1D and 2D decompositions, and concluded that, at least when modelling large, nearby galaxies, 1D techniques have more advantages than 2D techniques. Finally, we developed a prescription to estimate the uncertainties on the 1D best-fit parameters for the 66 spheroids that takes into account systematic errors, unlike popular 2D codes that only consider statistical errors.