Plotting jada increments with AVD tools

[RickRawlins]

This is a placeholder for an issue with plotting analysis increments generated by jada. The AVD team have developed visualisation software (not necessarily fully released?) for examining lfric-based files using geovista modules (which will eventually be a welcome replacement for our much-used xconv utility for UM model files). The new tool works on lfric diagnostic model files and on increment files (as created from VAR by lfric2um route) but NOT for jada-created increment files. Both sources of increment files are routinely read by IAU within lfric, but presumably there remain differences in metadata that affect the plotting interface.

[RickRawlins]

VERY minor addendum. The gmsh facility - which was the interactive facility for viewing fields before increment files became available - may well have some restrictions on file naming, or somesuch, which made it difficult to use in 1 case I examined. [Since this is (probably) a deprecated method, not worth pursuing unless it is re-introduced for lfric-jedi.] Following https://github.com/JCSDA-internal/lfric-lite-jedi/pull/302, gmsh files can be optionally generated. My legacy example had
gmsh file path: /data/d03/frrr/gmsh_temp/sonde_mob2023.0704
but this produced files, eg
sonde_mob2023.0704_specific_humidity.msh
....
sonde_mob2023.0704_mass_content_of_cloud_liquid_water_in_atmosphere_layer.msh
sonde_mob2023.msh.p9 (etc, per processor)
so the truncated .p? filenames all referred to last generated ie mass_content, so only this field could be looked at.
Maybe something to do with having a period . in the file name reference? Maybe fixed at a later version (although -from a quick look - I could see no later updates ).

But really just to note - it will be much better to have a full file interactive visualisation acting on the increment file.

[phlndrwd]

Hi @RickRawlins ,

Recent conversations with @mo-ginverarity have highlighted that this issue almost certainly extends beyond MONIO. However, we're beginning to think of ways to produce at least a temporary fix. With a bit of luck we should be able to close this issue when that fix is implemented.

Cheers,
Phil

[phlndrwd]

Hi @RickRawlins. Is this issue still valid?