Improvements for the Plotting codes
Current suggestions
- An interesting misbehavior is occurring when
plotting difference fields. Is the area map failing to reset between
windows?
- Unify plotting field codes (No QG vs PE input card).
- Determine the type of volume data for each variable from
its "positions" attribute.
- When plotting multiple fields, need to allow separate max/min limits
for each field.
- 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.
- First determine which variables are directly present in the
netCDF file. Then determine which can be computed.
- Need to enable clipping to allow "zooms".
- Include plot version number in plots (the tiny text area).
- 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.
- 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.
- Vertical Codes:
- Allow the tangential velocity to be a vector, comprised
of both the horizontal tangential velocity and the vertical velocity.
- Alex Warn-Varnas would like the ability to plot shears and bulk
Richardson numbers.
- When overlaying fields, the max/min reported are for the overlay.
Generally would prefer the statistics from the underlying field.
- 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.
- 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
- 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.