Randy review proposed netCDF enhancements with technical team, determine best course of action, break into tasks and assign.



John's input:

For now, I agree that using the ncl_convert2nc tool is a good solution.

I'd be in favor of doing the following:

(1) Enhance MET to read the NetCDF output of ncl_convert2nc

  - that should cover the HDF5 satellite stuff

(2) Enhance MET to read the NetCDF output of the pinterp WRF-ARW tool.

  - that should satisfy a lot of WRF users

(3) Enhance MET to read NetCDF data that follows the CF-convention.

  - including the convention for how to specify projections

(4) Fix MET so that all of the NetCDF files we write (output of

PCP-Combine, Grid-Stat, MODE, and Wavelet-Stat) are CF-compliant and can

be displayed in IDV.



Some good news is that the NetCDF-4 library is backwards compatible and

can be used to read/write NetCDF-3 files, referred to as the "classic"

data model as opposed to the "enhanced" data model available in NetCDF-4.

Based on the issues discussed in the class, I think we should stick with

the "classic" data model for the foreseeable future. It's much, much more

widely used than the enhanced data model.
[MET-4] created by tressa