This is a collection of questions and references for the radiative transfer discussion. We'll start with a quick taxonomy of radiation-hydro issues, then move on to these questions.


  1. Most people working in multiple dimensions are using flux-limited diffusion. How bad is it, and how do we check? For what types of problems is it OK, and for which do we need something better?
  2. For problems where we do need to do better, what sort of approach should we pursue? Should we try combining FLD with ray-tracing or Monte-Carlo, go to variable Eddington factor methods, or something else?
  3. Do we need to worry about operator-splitting radiation and hydro, or do we need to combine them without splitting to obtain high accuracy?
  4. Should we be focusing on mixed-frame or comoving-frame formulations of the radiation-hydro equations? Comoving-frame formulations are probably simpler in at least some types of codes, but by construction they cannot be conservative.
  5. Are there any problems in linking sink particles with radiative transfer, i.e. having sinks become "star particles"?

References (in no particular order):