Are the accretion rates super-Eddington in NLS1s?

Suzy Collin (Observatoire de Paris - Meudon)

Thanks to the release of large AGN samples, the determination of the black hole (BH) masses is now possible for several hundreds of AGN, using the empirical relation of Kaspi et al. (2000). One finds thus a strong correlation between the BH masses and the FWHMs of the Hbeta line, implying a luminosity close to the Eddington limit for NLS1s. This could be due to a systematic underestimation of the BH mass in a fraction of objects, but I will show that it is most probably not the case in the majority of them. With T. Kawaguchi we have assumed that the optical luminosity is provided by the accretion disk and we have shown that NLS1s accrete at strongly super-critical rates, even when taking into account the possibility of a non-viscous energy release. This last case leads to a redder optical continuum than a standard disc which agrees better with the observations. The fact that the bolometric luminosities saturate at a few times the Eddington limit for strongly super-critical accretion rates is predicted by slim disc models. The accretion rates stay however always of the order of a few M$_{\odot}$/yr, indicating that the growing of the BHs is mass supply limited (linked with a starburst?). Since the BH masses would increase by one order of magnitude in a few 10$^7$ years, NLS1s should not follow the same BH - bulge relation as other galaxies, and they should play an important role in shaping the mass function of BHs. I will also discuss the alternative possibility that the optical emission in NLS1s is not mainly due to the accretion disk, as for instance to a strong wind, but I will show that it is not likely.

  • PDF