search

JCSDA-internal / ioda-converters

Various converters for getting obs data in and out of IODA
9 stars 5 forks source link

add ctest for ground-based gnss ipw product #1539

Closed BenjaminRuston closed 2 months ago

BenjaminRuston commented 3 months ago

add a ctest for the ground-based gnss integrated precipitable water (IPW) product

Description

add a ctest for an NCEP BUFR file for IPW from ground-based GNSS

Issue(s) addressed

Resolves #

Impact

Potential to add IPW product to assimilation

Checklist

BenjaminRuston commented 3 months ago

@ibanos90 added this to see if it properly passes regression testing

the input file, and resulting output file are too large.. @gthompsnJCSDA may be able to help edit the input BUFR to trim it down to well under a MegaByte

for this file at current resolution got a IODArange.py output of:

file: 20240806T00Z_ipw_gnssgb_ncep.nc
range(/MetaData/dateTime): [2024-08-05 23:59:00, 2024-08-06 00:02:00]
range(/MetaData/heightOfStation): [90.0, 90.0]
range(/MetaData/latitude): [-77.83836364746094, 79.98895263671875]
range(/MetaData/longitude): [-170.72244262695312, 179.19654846191406]
range(/MetaData/stationName): [0001-JMA, ZZON-ROBH]
range(/ObsValue/precipitableWater): [2, 72]
BenjaminRuston commented 2 months ago

@ibanos90 is the variable that should be a float but is reported as an integer the Observation value itself: range(/ObsValue/precipitableWater): [2, 72]

will pick this up and try to drag across the finish line

ibanos90 commented 2 months ago

@ibanos90 is the variable that should be a float but is reported as an integer the Observation value itself: range(/ObsValue/precipitableWater): [2, 72]

will pick this up and try to drag across the finish line

Yes, that's correct, but @mrixlam can provide more info

mrixlam commented 2 months ago

@ibanos90 is the variable that should be a float but is reported as an integer the Observation value itself: range(/ObsValue/precipitableWater): [2, 72] will pick this up and try to drag across the finish line

Yes, that's correct, but @mrixlam can provide more info

The variable we discussed about is /ObsValue/precipitableWater.

I have some decoded bufr messages here: /glade/derecho/scratch/mrislam/work/singv_ng_mpas/data/gpsipw in case you have time to take a look at them. Decoded bufr messages have floating point values for precipitableWater but in the converted NetCDF file, it becomes integer and we have no idea why.

BenjaminRuston commented 2 months ago

@mrixlam and @ibanos90 think I've wrapped this one up,, I'm getting floats for the output and I've reduced the sample file size down to just a few observations to make the ctest dataset much smaller

please let me know if this works for you.. thx