[Feature Request] default to prefer netCDF for all input format types

[BenjaminTJohnson]

Currently, In various IO routines, filenames of the various coefficient files are constructed based on SENSOR_ID and *_Format choice between 'binary' and 'netCDF', wherein it selects the extension appropriate to the *_Format variable. binary is presently the default format. The default format or user-defined format occurs in CRTM_Init. If it fails to find the specified input format file (e.g., SpcCoeff_Format='binary' would look for SENSOR_ID//".SpcCoeff.bin"), it doesn't attempt to look for a netCDF file as a backup.

However, as we transition to netCDF only inputs, we're running into a situation where we want to prefer netCDF files first, but not fail if it doesn't exist.

This "check" should occur in the IO routines of the various input filetypes, where we start with the default (netCDF), then if it fails, try to find the binary file, if that fails, report an error. If it fails to find the netCDF file, but finds the binary as the fallback, this should be reported to the user as a warning. I don't think this should be an over-engineered solution, since ultimately our goal is to completely eliminate the need for binary format files.

We'll aim for CRTM v3.2 for this.

[BenjaminTJohnson]

I think this also eliminates the need to have format specifier inputs in CRTM_Init.

[chengdang]

Across aerosol, cloud, surface emissivity coefficient LUTs, the only tables that still need to be converted into the NetCDF format are MW water emissivities.

Something to be discussed: given the incoming plan for surface module upgrades, and the FASTEM algorithm/structure of these tables, I'd recommend using the existing hardcoded parameterization directly without converting them into the NetCDF format.