The radiated power profile is customarily read into SNAP as a function of minor radius from BR waveforms or 2D UFILES. A separate code maintained by John Schivell performs the actual Abel-like inversion of chord-integrated radiated power intensity from tangentially-viewing or vertically-viewing bolometer arrays. Unfortunately, the poloidal symmetry assumed by the inversion routine is rarely satisfied by the plasma. In addition to strong poloidally-localized radiated power near the bumper limiter, there can be other asymmetries such as the BORS (bottom outside radiation streak) during beam injection.
These difficulties affect both the edge and core measurements of radiated power density. There is a considerable uncertainty (factor ) in the magnitude of the flux-surface-averaged radiated power density near the plasma edge, where it is large. The uncertainty in the core radiated power density is even larger, because it is small compared to the edge density and so is severely affected by relatively small errors in determining the edge radiation. So for practical purposes in most beam-heated plasmas, we have no information on the central radiated power density.
In many cases, the profile read in from the BR waveform or UFILE will be poorly behaved in the plasma core (extending out even to )--it will diverge to a very large number, or else it will go negative. The standard practice to deal with this problem is to apply a cut to the profile, which sets the radiated power density flat inside a user-defined radius. This generally results in a radiated power density that is negligible in the core, compared to other power sources and sinks (as expected).