At the moment we have temperature, salinity, u, v, ssh, ice1, ice2, and ice3. The ice variables sometimes get hijacked for biogeochem. It would be betterto have generic tracer3d and tracer2d. This would make setting up Ocean/Ice/BioGeoChem boundary conditions easier (the true variable name being defined in the YAML file and meta data harvested from the parent netcdf file). We would also like ice with categories to be possible as perhaps tracer_cat. This would be somewhere between tracer3d and tracer2d since it has 3 dimensions but we would not want vertical interpolation just several independent stacked slices.
At the moment we have temperature, salinity, u, v, ssh, ice1, ice2, and ice3. The ice variables sometimes get hijacked for biogeochem. It would be betterto have generic tracer3d and tracer2d. This would make setting up Ocean/Ice/BioGeoChem boundary conditions easier (the true variable name being defined in the YAML file and meta data harvested from the parent netcdf file). We would also like ice with categories to be possible as perhaps tracer_cat. This would be somewhere between tracer3d and tracer2d since it has 3 dimensions but we would not want vertical interpolation just several independent stacked slices.