BS in Physics
Vardha N. Bennert
The empirical relation between the stellar velocity dispersion (SVD) of the bulge and the mass of the central supermassive black hole (BH) suggests a link between host galaxy and BH evolution. For active galactic nuclei (AGNs), the BH mass (MBH) can be estimated in a straightforward way from the Doppler broadening of the broad emission lines using the so-called virial method. However, the powerful AGN continuum emission often outshines the underlying stellar absorption lines, making it difficult to measure SVD of the host galaxy. Thus, the MBH - SVD relation is difficult to establish for galaxies containing AGNs. As a remedy, the line-width of the [OIII] 5007 Å emission line has been used as a surrogate for SVD. To probe the validity of this substitution, our team used a sample of ∼100 local AGNs with precise measurements of SVD from spatially-resolved high S/N Keck spectra and fitted the central [OIII] 5007 Å emission lines with single Gaussian, double Gaussian, and Gauss-Hermite profiles, in both latter cases to allow for a blue or red-shifted wing due to e.g., outflows or jet interaction. For the central spectra, the study revealed no clear correlation between SVD and the [OIII] line widths. We here extend this analysis to spectra at greater distances from the nucleus thought to be less affected by outflows or jet interactions and instead dominated by the bulge potential. The results of this analysis show a loose correlation between SVD and the [OIII] line width at larger distances, albeit a large amount of scatter, suggesting that making a substitution between these two quantities should be done only for high S/N data and when there is significant confidence in the accuracy of the measurements and in the absence of interfering factors such as outflows or jet interactions, which have been shown to perturb the relation even when accounting for their effects as we have done. It was also determined that the preferred fitting method to measure the line width of the [OIII] emission is a double Gaussian, but only when there is a clear asymmetry in the profile, otherwise, a single Gaussian fit should be used. Even under ideal conditions, however, it was determined that the [OIII] line width tends to overestimate the SVD by roughly 30 ± 10%.