Sun-satellite angle ranges are not consistent

[adybbroe]

from pyorbital.orbital import get_observer_look
from pyorbital.astronomy import get_alt_az, sun_zenith_angle
from datetime import datetime
import numpy as np

dtobj = datetime(2018, 1, 16, 14, 0)
lons, lats = np.array([-30., -27.]), np.array([0., 0.1])

sunz = sun_zenith_angle(dtobj, lons, lats)
sata, satel = get_observer_look(-75., 0., 35786023.0, dtobj, lons, lats, 0)
sunalt, suna = get_alt_az(dtobj, lons, lats)
suna = np.rad2deg(suna)

print(suna)
print(sata)

The above code produce the following output:

[ 173.60123028 -178.56645841]
[  90.          269.90995988]

So, the sun azimuth angles seem to be in the range of -180 to 180, but the satellite azimuth angles are in the range of 0 and 360.

I suggest that PyOrbital always return angles in degrees between -180 and 180!

[adybbroe]

The thing is that the above azimuth angles should result in nearly the same azimuth differences. When you pass under the sun as above, the two angles jump, but the relative difference should not.

print(suna - sata)
# [  83.60123028 -448.47641828]

ssadiff = np.mod(suna - sata, 360)
ssadiff = np.where(ssadiff > 180, 360 - ssadiff, ssadiff)
print(ssadiff)
# [ 83.60123028  88.47641828]

But, if you had done something like this (as it was in Satpy before January 16, 2018) the azimuth difference angles jump!:

ssadiff = np.abs(suna - sata)
ssadiff
# array([  83.60123028,  448.47641828])

ssadiff = np.where(ssadiff > 180, 360 - ssadiff, ssadiff)
ssadiff
# array([ 83.60123028, -88.47641828])