enable asgs to handle version differences in adcirc and adcirc+swan

[jasonfleming]

Different versions of ADCIRC and SWAN have different requirements for input. Some things that were required in older versions are now forbidden, and vice versa. These changes include the following:

1. wave_refraction_in_swan nodal attribute
2. initial_river_elevation nodal attribute
3. courant limiter on spectral propagation velocities in swan
4. meaning of IM value in v53 vs v54, specifically regarding Smagorinsky lateral viscosity

These differences need to be catalogged and then systematically accounted for in the scripts that generate ASGS input files.

[wwlwpd]

Ok, sounds like an important capability. Where would you place this in terms of priority wrt this season's operations?

[wwlwpd]

The basic proposal for this is to enhance the initadcirc tool to allow for easier switching and adding Environmental variables that will communicate to ASGS what sorts of ADCIRC related features are enabled or not enabled.

• initadcirc may become adcircctl that includes an install option
• load adcirc will remain (used in combination with save profile)
• additional meta data about the "loaded" ADCIRC will be exposed via environmental variables
• adcircctl can assist with communicating meta data about an ADCIRC installation as a tool that can report this meta data (e.g., adcircctl --get-features "ADCIRC-INSTALL-NAME, etc

The "hard" part will be making ASGS aware of these feature options in a clean and extensible way.

Longer term, adcircctl could also be used as a wrapper around generating the fort.15, etc. Hiding all of that complexity (in an sane way) behind adcircctl would allow core ASGS scripts to be much cleaner.