Characteristics of the Deep Minimum and Normal states of MCG$-$6-30-15

The study described in Sections 3 and 4 represents the most detailed study of the Deep Minimum state of MCG$-$6-30-15 to date. We have come to two important conclusions.

1. We have demonstrated the robustness of the extremely broadened disk reflection features reported in Paper I. We employ the generalized thin disk model of Agol & Krolik (2000), which includes a torque applied at the radius of marginal stability $r=r_{\rm ms}$, and find that this inner torque is dominating the energetics of the system. In other words, the Deep Minimum disk is shining via the extraction of spin-energy from the central black hole and not through accretion.
2. Examination of both the difference spectra and direct spectral fits to the 10ksec segments of data shows that the intensity of the broad disk feature is consistent with being proportional to the 2-10keV flux. In other words, the equivalent width of the disk feature is roughly constant as the source undergoes its large amplitude variability. This is expected from the simplest X-ray reflection model.

It is important to compare and contrast these results with studies of MCG$-$6-30-15 in its normal state. In their paper that originally identified the Deep Minimum state, Iwasawa et al. (1996) used ASCA to show that the iron line profile was substantially broader in the Deep Minimum than at other times. More recently, Fabian et al. (2002; hereafter F02) examined an independent and long (350ksec) XMM-Newton observation of MCG$-$6-30-15 which mostly caught it in its normal flux state. In agreement with the expectation from Iwasawa et al. (1996), F02 found the iron line profile to be generally narrower than in the Deep Minimum state of Paper I, although they clearly noted an extreme red-tail extending down to $\sim 3$keV. Fitting the iron line with a near-extreme Kerr black hole model ($a=0.998$) using a broken-powerlaw emissivity profile indicated a rather flat emissivity profile ($\beta\sim 2.5$) for $r>6r_{\mathrm g}$, breaking to a steep profile ($\beta\sim 5$) within this radius. Thus, the principal difference in the shape of the emissivity profile between the Deep Minimum and normal states of MCG$-$6-30-15 appears to lie beyond some radius $r\sim 6r_{\mathrm g}$. While it is beyond the scope of this paper to fit our physical accretion disk models to the long XMM-Newton data set, it is clear that a torque-dominated disk around a rapidly spinning black hole cannot reproduce the normal state emissivity profile.

There are also interesting differences in the spectral variability properties of the two states. Careful analysis of the RXTE-PCA data for MCG$-$6-30-15 during the normal state clearly showed that the iron line flux underwent significant variations but was not correlated with the continuum flux (Lee et al. 2000; Reynolds 2000; Vaughan & Edelson 2001). This was confirmed in a rather direct manner by Shih, Iwasawa & Fabian (2002) who used the 910ksec ASCA observation of MCG$-$6-30-15 to show that neither the intensity nor profile of the iron line were functions of the continuum flux. Finally, Fabian et al. (2002) and Fabian & Vaughan (2003) examined the long (350ksec) XMM-Newton/EPIC data of MCG$-$6-30-15 in its normal state and found that the hard-band difference spectra were all well described by power-law forms. Using this fact, these authors decompose the EPIC-pn spectrum into an almost constant reflection dominated component and variable power-law component. This is clearly different to the behaviour that we find during the Deep Minimum state.