2010ApJ...714..860K

The Suzaku observation of LMC X-3 gives the best data to date on the shape of the accretion disk spectrum. This is due to the combination of very low absorbing column density along this line of sight, which allows the shape of the disk emission to be constrained at low energies by the CCDs while the tail can be simultaneously determined up to 30 keV by the high-energy detectors. These data clearly demonstrate that the observed disk spectrum is broader than a simple "sum of blackbodies," and relativistic smearing of the emission is strongly required. However, the intrinsic emission should be more complex than a (color-corrected) sum of blackbodies as it should also contain photoelectric absorption edges from the partially ionized disk photosphere. These are broadened by the relativistic smearing, but the models predict ∼3%-5% deviations for 1/3-1 solar abundance around the edge energies, significantly stronger than observed. This indicates that the models need to include more physical processes such as self-irradiation, bound-bound (line) absorption, and/or emission from recombination continua and/or lines. Alternatively, if none of these match the data, it may instead require that the accretion disk density and/or emissivity profile with height is different to that assumed. Thus, these data demonstrate the feasibility of observational tests of our fundamental understanding of the vertical structure of accretion disks.