Improvements for the Feature Model codes

current suggestions

1. Bring Feature Models into HOPS.
   • LAA's Biology
   • LAA's TS-based Gulf Stream
   • AG's velocity based Gulf Stream
   • NQS's shelf-break front
   • HGA's generalized
2. LAA's Biology:
   • Allow selection of physical base field. Don't hardwire for temperature.
3. LAA's TS-based Gulf Stream:
   • Read in a GRIDS netCDF file. Check domain parameters, transfer any mask.
   • Add error fields. Non-Dim errors are 1 outside buffer regions, some small value inside "data" regions and cosine interpolated in buffer regions.
   • Copy logic from CJL Matlab FM to report ring peak velocities.
   • Echo the input parameters to the log file. Include the standard header.
   • Convert the unformatted input files to either netCDF or ASCII formatted.
   • Flesh-out the read-me file.
   • Look at Larry's note on the other needed pieces.

     >   (1) you have a number of programs (as opposed to subroutines) in that
     >       directory
     >       what are they and how do they go together?
     
     mk_clim.F:
     
        This makes a levitus climatology file (input to FM) from the levitus
        data files I have elsewhere.
     
     mk_cross.F:
     
        This makes "cross.section2" from "cross.section" (see below).
     
     mk_rings4.F:
     
        This is for making the ring files (input to FM).
     
     >   (2) you have a couple of hard-wired, unformatted data files
     >       what are they and how are they used/created?
     
     cross.section
     
        This is Gulf Stream cross section data (input to FM). For its format, 
        see "getgs.f". It is made up of ZTS profiles across the Gulf Stream,
        where the variable "xgs" gives the position of the profile (in km).
     
        If you want to see how I made "cross.section" from the raw BioSYNOP
        data, see /home/alboran2/larrya/fmodel9504/d01.data/pro9.f.
        But that is pretty ugly. (It essentially involves taking the stations
        from 8 cross sections, combining them, interpolating to standard depths,
        interpolating horizontally, and putting on end profiles.)
        You might want to use Pegasus or other cross section data instead?
     
     cross.section2
     
        This is a truncated version of "cross.section", useful if you want to
        put Levitus climatology beneath the Gulf Stream cross section (as the
        BioSYNOP observations really only extend to about 800 meters).
           

   • For LAA's cross-section code, add logic to compute the zonal velocity. Create an error map based on u*abs(u). Use this error field when melding in the TS-FM (ie. only meld in Stream data.)