Sundogs or 22° halo?

Imaged by Norma Wallace at Christopher Lake, Saskatchewan, Canada.     ©Norma Wallace, shown with permission
It's thought - but not proven - that large plate crystals flutter downwards, relative to local cloud air currents, like autumn leaves. Greatest tilts would be at the ends of the pendulum like swings.

But we do not know for certain.
Atmospheric
Optics
About - Submit Optics Picture of the Day Galleries Previous Next Today Subscribe to Features on RSS Feed

HaloSim ray tracings for plate crystals with increasing tilts and a low sun.

The values are the standard deviation for a Gaussian tilt distribution.

There's no evidence that real crystals have Gaussian tilts!
         

Extreme sundogs are made by almost horizontal plate crystals.    Extreme 22°  halos are from hexagonal crystals of some sort – probably clusters – tumbling in all directions.

Between the extremes are apparently ever taller ‘sundogs’ and 22° halos of various completion.

Both halos have the same ray paths, only crystal orientation distinguishes them and those demarcations are blurred.  

At low sun, plate crystals of increasing wobble give taller sundogs which can also be considered to be ever increasing segments of the 22° halo.   The latter does not need random orientations – plates and columns with large tilts also generate it.
 
At higher sun the sundog moves further from the sun and is no longer concident with the 22 degree halo. Increasingly tilted plate crystals then produce a sundog and a segement of the 22° halo!