Smoothing functionality applied to specific transport models

[AlexSaperstein]

The smoothing functionality is currently applied globally. This introduces issues when using a "combined" transport model, as the same smoothing settings must be applied to the separate transport components, some of which may not be compatible with the smoothing needs of the others.

The example motivating this issue is the incompatibility between a QLKNN model being used throughout most of the plasma, with a fixed transport coefficients model being used only within the small pedestal region. The smoothing_width is set globally to normally be 0.05-0.1, but this width encompasses the entire ~0.04 long pedestal region, smoothing over it completely. Smaller smoothing_width's, however, are incompatible with the rest of the plasma, as they significantly increase the run time of the simulation.

It would be useful if the smoothing functionality could be extended to be local to specific transport models within a combined set.

[theo-brown]

I've encountered a similar thing in some of my TGLF sims. The counterpoint to consider is that it's often useful to smooth over discontinuous junctions between two transport models - iirc this is why we added global smoothing in the first place. Maybe the best thing to do would be to support both global and per-model smoothing?