parent sigma to child sigma

[micdom]

pyBDY does not work with ln_sco = .true. ! s- or hybrid z-s-coordinate. We have to consider keeping this option in some form if we are going to include an option for parent sigma to child sigma.

[jdha]

Initial thoughts:

Create a generic Depth Class to handle source and destination 3D vertical coordinates.

unit tests to include:

source data with: zt, zw, zu, zv, e3t, e3w, e3u, e3v this is reduced incrementally to input with only: zt (making some crude assumptions to derive other vertical variables)

[both in 1D and 3D formats]

In all cases a vertical weighting is generated to allow horizontal interpolation on a common depth level [in the simplest case of ZCO this will default to 1 on the common depth level].

ln_sco flag to be removed as this was used in NEMO <=3.6 when we made use of hc and sigma coefs

The destination 3D vertical coordinate Class will assume NEMO >=4.0 and access to a current style domain_cfg or mesh_zgr file for simplicity.