Add Cloud_Optical_Thickness to viirs_l2.yaml

[martin-rdz]

Add the cloud optical thickness derived from the two channel retrieval to the CLDPROP_L2 reader of VIIRS.

• Adds to Add Viirs L2 Reader + Enhancments #2729
• Tests added
• Fully documented
• Add your name to AUTHORS.md if not there already

[djhoese]

If this value is in the file I don't think it should be needed here if the python code pulls it from the file and puts in in the DataArray's .attrs. If it is needed here, then it should be surrounded by double quotes:

Suggested change

	long_name: Cloud Optical Thickness two-channel retrieval using 2.2 um and either 0.65 um, 0.86 um or 1.24um (specified in Quality_Assurance) from best points: not failed in any way, not marked for clear sky restoral
	long_name: "Cloud Optical Thickness two-channel retrieval using 2.2 um and either 0.65 um, 0.86 um or 1.24um (specified in Quality_Assurance) from best points: not failed in any way, not marked for clear sky restoral"

[martin-rdz]

True point, the long_name is covered in the netcdf. Shortened ncdump below:

short Cloud_Optical_Thickness(number_of_lines, number_of_pixels) ; 
    Cloud_Optical_Thickness:long_name = "Cloud Optical Thickness two-channel retrieval using 2.2 um and either 0.65 um, 0.86 um or 1.24um (specified in Quality_Assurance) from best points: not failed in any way, not marked for clear sky restoral" ; 
    Cloud_Optical_Thickness:_FillValue = -9999s ;
    Cloud_Optical_Thickness:valid_min = 0s ;
    Cloud_Optical_Thickness:valid_max = 15000s ;
    Cloud_Optical_Thickness:scale_factor = 0.01 ;
    Cloud_Optical_Thickness:add_offset = 0. ; 
    Cloud_Optical_Thickness:units = "none" ; 

My very first thought was to match it to Cloud_Top_Height and the variables in aerdb_l2_viirs, but I see the point in not having these information redundant. Can we just leave out the long_name tag or does it need to be specified, that this value should be taken from the .attrs?

Thanks for the review!
Martin

[djhoese]

You'd have to test it to be sure, but these lines in the python seem to do what we'd want:

satpy/satpy/readers/viirs_l2.py

Lines 102 to 103 in 5d2b1fd

	i = getattr(self[var_path], "attrs", {})
	i.update(ds_info)

So if you leave long_name out of the YAML then the reader should pull it from the file's variable's attributes.

[djhoese]

I left some comments that should hopefully make pre-commit other tests happier.

[djhoese]

Tests are failing and it seems to be from the long_name in the YAML needing the quotes I suggested.

[martin-rdz]

Ah, sorry for triggering the CI chain. I tried to update my fork and the merge got pushed automatically.
Thanks for pointing out the issue with the quotes again. I hope find some time to fix the issue the next days.
Again sorry for causing hassle.