The iron line in MCG-6-30-15 from XMM-Newton:: evidence for gravitational light bending?

The lack of variability of the broad iron line seen in the Seyfert 1 galaxy MCG-6-30-15 is studied using the EPIC pn data obtained from a 2001 observation of 325 ks, during which the count rate from the source varied by a factor of 5. The spectrum of 80 ks of data from when the source was bright, using the lowest 10 ks as background, is well fitted over the 3-10 keV band with a simple power law of photon index Gamma= 2.11. The source spectrum can therefore be decomposed into an approximately constant component, containing a strong iron emission line, and a variable power law component. Assuming that the power-law continuum extends down to 0.5 keV enables us to deduce the absorption acting on the nucleus. A simple model involving Galactic absorption and two photoelectric edges (at 0.72 and 0.86 keV) is adequate for the EPIC spectrum above 1 keV. It still gives a fair representation of the spectrum at lower energies where absorption lines, Unresolved Transition Arrays (UTAs), etc. are expected. Applying the absorption model to the low flux spectrum enables us to characterize it as reflection-dominated with iron about three times Solar. Fitting this model, with an additional power-law continuum, to 32 x 10 ks spectra of the whole observation reveals that Gamma lies mostly between 2 and 2.3 and the normalization of the reflection-dominated component varies by up to 25 per cent. The breadth of the iron line suggests that most of the illumination originates from about 2 gravitational radii. Gravitational light bending will be very strong there, causing the observed continuum to be a strong function of the source height and much of the continuum radiation to return to the disc. Together these effects can explain the otherwise puzzling disconnectedness of the continuum and reflection components of MCG-6-30-15.