Remove duplicate of potential vorticity parameter

[jsmretschnig]

ACCF.met_variables included the same parameter twice: pv (potential_vorticity)

[jsmretschnig]

Also: the grib1_id for relative_humidity (r) should be 52 instead of None. See here: NCEP Grib v1 Code Table Can't find it in the code though. The unit seems correct (%), but it's not consistent with ACCF.met_variables, there it says 1.

To reproduce the issue: try to download ERA5 data with ACCF.met_variables as parameter

[zebengberg]

Also: the grib1_id for relative_humidity (r) should be 52 instead of None. See here: NCEP Grib v1 Code Table Can't find it in the code though. The unit seems correct (%), but it's not consistent with ACCF.met_variables, there it says 1.

To reproduce the issue: try to download ERA5 data with ACCF.met_variables as parameter

Thanks @jsmretschnig ! This is certainly confusing.

pycontrails has two different RelativeHumidity variables:

• A "core" RH with grib1_id=52. This gives the humidity as a proportion.
• An "ecmwf" flavored RH with grid1_id=None. This gives the humidity as a percentage.

Can you confirm that ACCFs expect the first flavor of RH? If so, I'll figure out how to fix this.

[jsmretschnig]

So apparently CLIMaCCF is fine with both options: https://github.com/dlr-pa/climaccf/blob/907c5f0cc84db8be675337d6a213c8706b024409/climaccf/weather_store.py#L277

And uses this parameter https://codes.ecmwf.int/grib/param-db/157