The observational appearance of slim accretion disks

We reexamine the hypothesis that the optical/UV/soft X-ray continuum of active galactic nuclei (AGNs) is thermal emission from an accretion disk. Previous studies have shown that fitting the spectra with the standard optically thick and geometrically thin accretion disk models often led to luminosities that contradict the basic assumptions adopted in the standard model. There is no known reason why the accretion rates in AGNs should not be larger than the thin disk limit. In fact, more general, slim accretion disk models are self-consistent even for moderately super-Eddington luminosities. We calculate here spectra from a set of thin and slim, optically thick accretion disks, assuming for simplicity a modified blackbody local emission with no relativistic corrections. We discuss the differences between the thin and slim disk models, stressing the implications of these differences for the interpretation of the observed properties of AGNs. We find that the spectra can be fitted not only by models with a high mass and a low accretion rate (as in the case of thin disk fitting) but also by models with a low mass and a high accretion rate. In the first case, fitting the observed spectra in various redshift categories gives black hole masses of similar to 10(9) M., for a wide range of redshifts and for accretion rates ranging from 0.4 (low redshift) to 8 M. yr(-1) (high redshift). In the second case, the accretion rate is similar to 10(2) M. yr(-1) for all AGNs, and the mass ranges from 3 x 10(6) (low redshift) to 10(8) M. (high redshift). Unlike the disks with a low accretion rate, the spectra of the high accretion rate disks extend into the soft X-ray region. A comparison with observations shows that such disks could produce the soft X-ray excesses claimed for some AGNs. We show also that the sequence of our models with fixed mass and different accretion rates can explain the time evolution of the observed spectra in Fairall 9.