For the thick disks we use the same model parameters as in paper I,
i.e. the accretion luminosity
,
the viscosity
parameter
and the fraction of the energy released in the
virialized atmosphere f=0.95. The mass of the galactic BH is
and the mass of the AGN is
.
Figure 1 shows the solution at r=10 for a proton energy
flux of
erg cm-2 s-1,
MeV. The cool disk is separated by a sharp
temperature front from a hot part at an optical depth of order
unity. This step like temperature profile is much more pronounced
compared to the results of Paper I. The temperature of the hot part
is
keV. The hot layer acts as an effective
Comptonizing region and a hard spectrum is emitted. The power law
index of the BH spectra in E F(E) is
.
At the soft end
of the spectrum there is an excess of soft photons with respect to a
blackbody with equivalent energy flux. This excess is due to a
"reverse photosphere effect'', as explained in Deufel et al. (2001).
The upper right panel of Fig. 1 shows the heating and
cooling rates. In the narrow transition zone (
)
between the hot top and the cool bottom electron conductivity is not
negligible. Elsewhere conductivity plays no role for the energy
balance.
The geometrical depth of the heated layer is
cm. This is small compared to the
distance from the compact object (
cm). Thus a cool
disk with a proton heated skin on top of it can still be considered a
"thin'' disk.
Figure 2 shows the dependence of the emergent spectrum and the temperature profile on the distance from the compact object in the galactic BH case. We obtain a result comparable to Paper I, i.e. with increasing distance the temperature of the hot part increases somewhat, whereas the optical depth of the heated layer decreases. A discussion of these effects can be found in Paper I.
For distances
from the galactic BH we could not find
heated solutions. With increasing distance from the central object the
proton temperature (
)
as well as the proton energy
flux (
)
decrease. In these conditions we find
solutions in which the energy supplied by the incident protons is
thermalized directly in the cool disk by bremsstrahlung, without
forming a hot surface layer. It is likely that a very thin (
)
hot
layer still forms in these conditions, but we are unable to resolve it with the
present method. This is indicated by the existence of thin hot atmosphere
solutions at low incident energy flux in the case of protons heating a
neutron star surface (Zampieri et al. 1995; Deufel et al. 2001). It probably also occurs
in the present case, but since these thin layers at large distances
from the hole do not contribute much to the overall spectrum, we have
not pursued this further.
Figure 3 shows the dependence of the emergent spectrum and
the temperature profile on the distance from the compact object in the
AGN case. We obtain slightly different solutions compared to our
results in Paper I. The optical depth of the heated layer is
smaller (
), the temperature of the hot part is
keV. Again a steep temperature jump separates
it from the cool disk underneath. Due to the smaller optical depth of
the heated layer Comptonization is not as efficient. A significant fraction
of the incident ions passes through the hot layer and thermalizes in the
cool disk underneath. This increases the soft photon flux relative to
the Comptonized flux, and produces a steeper spectrum. The
index of the AGN spectra in E F(E) is
.
Our model does
not produce heated solutions at distances
from the compact
object, though heated atmospheres might exist at very small optical
depths (see note above).
Copyright ESO 2002