New candidate for no vertical interp

[rdPatmore]

@jdha I can see you have created a branch under #158 for adding a feature for by-passing vertical interpolation. I have created separate version in parallel and I wanted to create branch that didn't interfere with yours. This enhancement appears to function well. Instead of placing it in the profile code I have rewritten the zgrv code to handle each case and have added a namelist option to control the feature.

[rdPatmore]

I have tested this code. Bounday files and coordinates are created fine, but the model doesn't seem to run past timestep 4 when using these in combination with interpolation on the fly. I am yet to map whether this issue steps from the pybdy code or my application of these bdys to nemo.

[rdPatmore]

I have plotted temperature. The fields look sensible but it appears the pybdy has chopped off the bottom - depths now have a maximum of ~1300 m instead of 5000+ m. Also, e3u has been set to nan everywhere, whereas this is not the case for e3t.

The way I have coded this up we set the depth to the same as the src. However, the code still runs through the interpolation. It would be cleaner to also bypass the vertical interpolation. I have been attempting to understand the details of the interpolation in order to do this but I’m not quite there with the understanding.

[rdPatmore]

The fix has been made into a pull request #163.