missing term to compute reflectance for viirs_compact ?

[ClementLaplace]

To give you some context I have compared datas between compact file (SVMC files ) given by the reader viirs_compact and the the Mband files (SVM?? files) given by reader viirs_sdr. I'm expecting to have similar results which I get with radiance for all bands and brightness_temperature for the bands M12 to M16.

I have obtained some huge differences for the reflectance for the bands M01 to M11. I have checked into the viirs_compact reader I have seen that the reflectance is computed at those lines.

It seems that there is a term that is missing which is to divide by the cos(solar_zenith_angle) , hence I have applied those corrections

solar_zenith_angle = self.read_geo({"name":"solar_zenith_angle"},info=None)
rads *= (100 * np.pi * a_vis * (dse**2)) / (b_vis * np.cos(np.deg2rad(solar_zenith_angle)))

With the corrections that I have applied I get better results. Was the cos(solar_zenith_angle) missing on purpose ?

[ClementLaplace]

I had a conversation with @mraspaud on slack about this topic. Here is the content

So you have to apply the sunz correction on the SVMC data after you have loaded it, you can check this ticket to know how to apply such correction.

[strandgren]

@ClementLaplace The issue you refer to is about applying the correction without satpy. But since you use satpy, I think this should work:

scn = Scene(filenames=['my_SVMC_files'], reader='viirs_compact')
scn.load(['my_dataset'], modifiers=['sunz_corrected'])

That should apply the regular satpy 1/cos(sunz) correction to the dataset. Please note that this correction starts to be reduced beyond 88 degrees, which is probably not the case in the viirs_sdr reader.