Correct units for displacements in netcdf

[logan-nc]

The netcdf records the units of xigradpsi and m. Just looking at the definition of $\xi\cdot\nabla \psi_N you would expect it to be dimensionless, and indeed Brent Ford from University of Wisconsin found this to be the appropriate interpretation when coupling to ELITE3D.

The units should be corrected if the above understanding is correct.

Can @matt-pharr or @parkjk think of any internal normalizations that would make the xigradpsi (which is the xspout variable in the code) have units of m? If not, do you agree the netcdf units should be changed to dimensionless?

[matt-pharr]

@logan-nc I was looking at this yesterday and I agree, xigradpsi should be unitless but I think xi_n should have units meters; similarly, Jbgradpsi should have units flux (T * 1/m * m^3) and b_n should have units T, based on my understanding.

[parkjk]

I agree xi_n in the units of meters and xigradpsi should be unitless.