Use Python API to convert ADPSFC prepbufr data to IODA

[nicholasesposito]

Configuration

• Test Date is 2021080100
• Model Background: FV-3 6-hour forecast on the cubic-sphere grid
• BUFR Converter :
  • convert Longitude from [0,360] to [-180,180] (it was unable to do this with yaml)
  • convert obstimeMinusCycleTime to dateTime (in units of seconds since 1970-01-01 00z)
  • grab and convert (if needed): latitude, stationElevation, airTemperature, virtualTemperature, windNorthward, windEastward, specificHumidity, stationPressure
  • The files used for this (the JSON and run*py file are in feature/NickE_adpsfc_prepbufr on GDASApp

Test using fv3jedi_hofx_nomodel.x with the IODA files converted from BUFR converter

• Convert prepBUFR to IODA format using bufr2ioda.x with python API
• Run UFO HofX

Check List

Investigate diag output from UFO

• [] ADPSFC HofX

[nicholasesposito]

The full conversion works. I had to put both MetaData/pressure and ObsValue/stationPressure into the IODA file because various obs operators require different inputs. I was able to run fv3jedi_hofx_nomodel.x on the IODA file to produce these results:

[nicholasesposito]

Here are the results of running the GSI (the files in /parm/atm/obs/testing/sfc.yaml

[nicholasesposito]

Emily and I noticed that the stationPressure values for GSI were much further away from the 1:1 line than for UFO. We plotted hofx for the GSI diag file, and found data points much closer to 1:1 (ignoring the data counts). This revealed that the UFO model has a better surface pressure model.