Add table of central wavenumber to iasi converter

[smaticka]

Description

Rather than pulling central wavenumber per experiment in the skylab yaml, we are adding a static table of central wavenumber values to be exported during the ioda convertion.

Need to add an 'assign method' in ioda-converter yaml for iasi that looks up a table of values.

Issue(s) addressed

Resolves #1313

[rmclaren]

If you use gettab on this BUFR file you get the following mnemonics:

gettab test/testinput/gdas.t12z.mtiasi.tm00.bufr_d | grep WAVE

| WVNM     | 006029 | WAVE NUMBER                                              |
| LOGRCW   | 025076 | LOG-10 OF (TEMPERATURE-RADIANCE CENTRAL WAVENUMBER)      |

but when you run print_queries.x test/testinput/gdas.t12z.mtiasi.tm00.bufr_d you find out these mnemonics are not available in the test dataset...

You can add a global like you tried, but I'm not sure this this is the right answer as I'm guessing these wavenumbers will be different for different IASI files (?). For more flexibility you could read the data using the python API.

[smaticka]

If you use gettab on this BUFR file you get the following mnemonics:

gettab test/testinput/gdas.t12z.mtiasi.tm00.bufr_d | grep WAVE

| WVNM     | 006029 | WAVE NUMBER                                              |
| LOGRCW   | 025076 | LOG-10 OF (TEMPERATURE-RADIANCE CENTRAL WAVENUMBER)      |

but when you run print_queries.x test/testinput/gdas.t12z.mtiasi.tm00.bufr_d you find out these mnemonics are not available in the test dataset...

You can add a global like you tried, but I'm not sure this this is the right answer as I'm guessing these wavenumbers will be different for different IASI files (?). For more flexibility you could read the data using the python API.

Thanks for checking this @rmclaren. The input test data file hasn't been updated yet, but was the next planned task. Just wanted to get the draft created for conversation on the table entry to begin. I'll work on the new file next.

[rmclaren]

Cool. I should probably update the print_queries.x output to show the description information so that you don't need to switch back and forth... Please be aware that the gettab output often does not really reflect the data in the file (lists subsets that don't exists, often does not list subsets actually in the file). print_queries.x will show you whats actually available in the file.

[smaticka]

Cool. I should probably update the print_queries.x output to show the description information so that you don't need to switch back and forth... Please be aware that the gettab output often does not really reflect the data in the file (lists subsets that don't exists, often does not list subsets actually in the file). print_queries.x will show you whats actually available in the file.

Gotcha. that makes sense then b/c WVNM was still in the output of gettab, but it's not in the print_queries.x output.

[fabiolrdiniz]

Thanks for working on it, @smaticka. I suspect your code will have a global attribute with the sensorCentralWavenumber values. However, I believe what you are looking for is a MetaData channel-dimensioned variable with those values (MetaData/sensorCentralWavenumber(Channel)).

@rmclaren, is there a way of passing a float vector straight to an IODA variable? We know she can do that through a filter when using this file (UFO level), but I believe people would like to have it added at this stage. The solution via YAML at runtime opens the door to have different sensorCentralWavenumber though (which was a concern above).

[fabiolrdiniz]

Do you know why the original file doesn't work? Please, notice that your added file (test/testinput/gdas.mtiasi.t00z.20220216.bufr) has 779 MB, it's too large. I'd revert it to use the original file or reduce the new one.

Suggested change

	obsdatain: "./testinput/gdas.mtiasi.t00z.20220216.bufr"
	obsdatain: "./testinput/gdas.t12z.mtiasi.tm00.bufr_d"

[BenjaminRuston]

@smaticka this likely will be moved to future work as updates to the converter are needed to make this happen. Should plan this either with Ron as part for bufr2ioda.x,,, or with SteveH so yaml can be used to write MetaData

[BenjaminRuston]

in short we may want to close this

[smaticka]

Closing to reopen later once 2d ability is allowed for MetaData

[BenjaminRuston]

@smaticka yes this is so confusing, but yes this ioda-converter PR can likely close again

we'll use the table of wavenumbers in a ufo instrumentTests ctest with a reduced ioda file