Open issue: The relative humidity is provided above water and ice depending on temperature.
I am not sure what the effect is on downstream data analysis. Quoting from the ncdump
float rh(time, level) ;
rh:_FillValue = -999.9f ;
rh:long_name = "Relative humidity" ;
rh:units = "1" ;
rh:missing_value = -999.9f ;
rh:comment = "With respect to liquid above 0 degrees C and with respect to ice below 0 degrees C.
Calculated using Goff-Gratch formula." ;
rh:standard_name = "relative_humidity" ;
However, also the specific humidity is available in the file, so we could calculate on our own.
The cloudnet datacenter provides profile extracts from ECMWF IFS for stations on the extraction list (https://cloudnetpy.readthedocs.io/en/latest/quickstart.html#model-data). A daily netcdf file contains hourly profiles with 136 height levels (https://cloudnetpy.readthedocs.io/en/latest/fileformat.html#model-file). This data should be superior to the GDAS files.
The datasource can be either configured in the config files or in matlab directly (hinting to https://github.com/PollyNET/picasso-nb-interface ;) )
Example for arielle at Neumayer:
Open issue: The relative humidity is provided above water and ice depending on temperature. I am not sure what the effect is on downstream data analysis. Quoting from the ncdump
However, also the specific humidity is available in the file, so we could calculate on our own.