[Dart-users] DART wind_obs_to_table.f90, obs_diag.f90 "print_obs_locations"

[Tim Hoar]

I'm trying to make things a bit simpler and have set my sights
on some obsolete portions of obs_diag.f90 and a one-off program
"wind_obs_to_table.f90".

To summarize: I want to remove some code from obs_diag that prints
the locations, and I want to totally remove the program wind_obs_to_netcdf.
All of these functions are already supported by obs_seq_to_netcdf.

Please give me your thoughts on the issue.

obs_diag  has a 'print_obs_locations' feature that prints the locations
of the incoming observations - a function that is totally duplicated by
the obs_seq_to_netcdf program. Rather than maintain both of them,
I'd like to remove the print_obs_locations functionality. Just convert
the observation sequence file to netCDF and everything is there,
and you can then explore the data with a lot of tools. The location
output from obs_diag lacks the metadata to really make it useful anyway.
It's time to remove that functionality from obs_diag.  Speak now ...

Similarly, wind_obs_to_table.f90 was written as a one-off program that
is now totally duplicated by obs_seq_to_netcdf.  The output of wind_obs_to_table
was plotted by 'DART/diagnostics/matlab/plot_wind_vectors.m'. This function
has been totally rewritten (and extended) to use the output of obs_seq_to_netcdf.
In many ways, wind_obs_to_table made some very scary assumptions about how to
pair up U&V components, the new way is much less scary to me.

Tim Hoar, Associate Scientist
National Center for Atmospheric Research
thoar at ucar.edu
303 497 1708

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/dart-users/attachments/20110422/bfc05944/attachment.html