Mystery currents in RDCON response separate from resonant layer with decreasing Lundquist number S

[matt-pharr]

As we increase eta or increase massden (rho), Lundquist number and thus overall resistive effects grow. With a growing lundquist number, some sort of current/shielding behavior emerges from the RDCON resonant layer (res cells shown in red, extensions in blue green and purple). This is shown here in the form of the flux-surface-normal magnetic field of the resonant eigenfunction nearing zero.

An example plot is here:

Following, here, is a logarithmic version where we instead vary massden:

When we vary the amplitude of the perturbation, the example here being the DIII-D C-coil current, the shifted current is shown to not move, just to scale almost linearly:

Does anyone know what these currents could be from? Why do they shift outwards with increasing S? Is this the true location of the resonant shielding current? If so, is it problematic that it leaves the resonant layer?

Hoping one of you all can help understand this @parkjk @aglasser17 @zrwang2015

[matt-pharr]

@AaronFroese You all use RDCON, correct? Are you familiar with currents like these? Alternatively, do you have an understanding you could share about how someone running this code goes about choosing a width for the inner layer? It is set in the examples to $\psi = 10^{-3}$ by default.

[logan-nc]

@parkjk: just reminding you of where this phenomenon we discussed today is recorded