The code assumes the projection is north-up and it is a descending scene.
For scenes over Antarctica, which uses the polar stereographic projection
where north is not always up, and ascending daylight scenes at the poles,
this assumption is not valid. The result is that the algorithm does not
correctly identify shadows since it is often looking in the wrong direction
for the shadows. For the L8 version, the correct approach is likely to use
the new angle coefficients to calculate the solar angle and correct
satellite viewing angle.
Original issue reported on code.google.com by rd.alt.5...@gmail.com on 19 Mar 2015 at 6:35
Original issue reported on code.google.com by
rd.alt.5...@gmail.com
on 19 Mar 2015 at 6:35