Improvements for the Plotting codes

Current suggestions

1. An interesting misbehavior is occurring when plotting difference fields. Is the area map failing to reset between windows?
2. Unify plotting field codes (No QG vs PE input card).
   • Determine the type of volume data for each variable from its "positions" attribute.
3. When plotting multiple fields, need to allow separate max/min limits for each field.
4. Just revised the "derived quantities from transport streamfunction" section in get_cdf2dat. Needed to ensure the correct data was in array f1
   • Check the derivations from QG streamfunction.
   • Check get_cdf3dat.
   • Revised the computations in external for spherical grids.
5. First determine which variables are directly present in the netCDF file. Then determine which can be computed.
6. Need to enable clipping to allow "zooms".
7. Include plot version number in plots (the tiny text area).
8. SACLANT icon code has 2 problems
   • The routine sac.F should be broken up and have its own source listing in the GNUmakefile.
   • The icon is done with "raster" algorithm rather than using NCAR to fill shapes. This takes up too much room in the gmeta file.
9. Horizontal Codes: Examine the overlay option.
   • Allow the overlay field to be interpolated to flat level, even if different from primary level.
   • The overlay of contours onto a color map does not seem to work.
10. Vertical Codes:
    • Allow the tangential velocity to be a vector, comprised of both the horizontal tangential velocity and the vertical velocity.
11. Alex Warn-Varnas would like the ability to plot shears and bulk Richardson numbers.
12. When overlaying fields, the max/min reported are for the overlay. Generally would prefer the statistics from the underlying field.
13. Determine whether the codes can plot the following:
    • Error fields in an OI assimilation
    • Depths of PE levels (both T & V grids)
    • any other field that might be in a HOPS netCDF file.
14. Labeling issues:
    • The titles change depending on whether a single field or multiple fields are requested.
    • Depths are rounded to nearest integer value.
    • Get plotting labels from netCDF file to avoid errors like this

Blasts from the past

1. From PFJL:

   In the title, sometime it writes the depths, some times the level numbers. I haven't been able to understand the logic, but there might be one. That seem to occur when one ask for more than one variable per page.

   • Most likely the codes write depths when dealing with flat levels and level numbers when dealing with terrain following levels. Double check this.