The subhorizon photos of 7/8 December show what seems to be part of a diffuse circular halo surrounding the subanthelic point. Similar effect can be simulated with random oriented crystals (arrow in the simulation on the left). It arises from raypaths 3-5-6-7-3 and 3-5-6-7-2-3, which in plate crystals make respectively the Liljequist and sub-Liljequist parhelia.
The simulation on the right provides a 180 degree view of halos from random orientation centered around the subanthelic point. The bright dot is the subanthelion. The narrow small radius halo is from raypath that I have not checked. It seems to disappear if the crystals deviate even a little from regular hexagonal shape. Then there is the diffuse 3-5-6-7-(2)-3 subanthelic halo and blue circle.
Finally, a note on the Parry crystals used in the simulation on the left: no triangles this time, regular h/d 1 hexagons were used. Simulations with Jukka Ruoskanen’s software.