Reading the EUMETSAT compact viirs format returns wrong platform name (J01 instead of NOAA-20)

[adybbroe]

Describe the bug

When reading the VIIRS compact format data the platform_name returned from the reader is J01 while it should be NOAA-20

To Reproduce

from satpy.scene import Scene
from satpy import find_files_and_readers
from datetime import datetime
from satpy.utils import debug_on
debug_on()

DATADIR = "/home/a000680/data/ears-viirs/noaa_new"

my_files = find_files_and_readers(base_dir=DATADIR,
                                  reader='viirs_compact',
                                  start_time=datetime(2019, 4, 9, 13, 1, 0),
                                  end_time=datetime(2019, 4, 9, 13, 18, 0))
scn = Scene(filenames=my_files)
scn.load(['true_color'])
scn.save_dataset('true_color', './true_color_NOAA20.png')

Expected behavior
No Warnings or errors from Pyspectral
and a nice image in the end

Actual results

Pyspectral looks for a file that doesn't exist, and then the true color either fails (pyspectral >=v0.8.5) or it just doesn't use the RSR data to derive the atm corrected VIIRS bands.

Screenshots
If applicable, add screenshots to help explain your problem.

Environment Info:

• OS: Linux
• Satpy Version: [0.14.0]
• Pyspectral Version: 0.8.4 and 0.8.7
• Readers and writers dependencies (when relevant): [run from satpy.config import check_satpy; check_satpy()]

Additional context
Add any other context about the problem here.

[djhoese]

On slack you mentioned https://github.com/pytroll/pyspectral/blob/4cd92661e6a5bfde3526a832e1b8eab9b741592d/pyspectral/rayleigh.py#L174

This was changed as part of pytroll/pyspectral#56 but I'm not really sure why. I don't remember at least. I don't/wouldn't mind if pyspectral went back to raising an exception but I think it should do it in a sane way. That is, get_effective_wavelength can raise an exception, but I think it should be caught (IOError, OSError) by the Rayleigh class which should then raise a KeyError or IOError. I say KeyError/IOError because that's what SatPy is currently trying to catch. If SatPy catches these then it will try using the wavelength instead of the platform/sensor/band.

Most importantly, the viirs_l1b reader also needs to be fixed to return NOAA-20 too.

[adybbroe]

@djhoese I can only agree.