Added the Grid_group and Interpretation groups under the sonar group …

[gavinmacaulay]

This version provides a suggestion for placeholder for the Grid_group and Interpretation groups. This is a first draft, and please provide feedback.

Copied from PR 37 on ices-eg/wg_WGFAST by @nilsolav

[lberger29]

Some thoughts about Grid_Group:

• it should avoid duplication of variable between Beam_Group and Grid_Group
• it should contain integrated backscatter on a fixed vertical mesh in MVBS or NASC unit for each beam with Time Latitude and Longitude Coordinate
• a link between a beam in Grid_Group and in Beam_group is needed
• is an estimation of the volume of space integrated needed in the Grid_Group?

Should we allow sample_t backscatter_r to be optionnal if we want to have NetCdf files with only integrated backscatter data for machine learning for instance?