During last year’s meeting of the German halo observers, we decided to drive on top of Mt. Klínovec (Keilberg) after dinner on Nov 27th, 2015. We used the car headlights as light sources for glittering diamond dust displays from ice crystals within the first few meters above the ground, while facing temperatures in the range of –5 °C to –6 °C at wind speeds of 5 – 6 bft. Simultaneously, there appeared a non-glittering, but slowly changing moon halo display in crystals higher up, including a “traditional” Moilanen arc:

(20:03 CET, unsharp masked, for the original image see here)

In the light of our artificial sources, we could make out several of the common halo types, but, moreover, from time to time some rather prominent glittering at about half the radius of the 22° halo. None of us could visually recognize the definitive shape of this arc. On common long-exposure pictures it did not appear very prominently due to averaging out the glitter which attracted the visual attention:

(20:16 CET, exposure time 1 s, another 1 s exposure from 20:11 can be found here)

Luckily I also had recorded a couple of real-time videos (at 25 fps), and this material allowed me to carry out a somewhat unusual analysis method: I extracted the individual frames and re-combined them to maximum stacks in order to preserve the “glitter enhancement” while exploiting the advantage of a longer exposure time. As a result, clear pictures of a circular arc of about 12° radius inside and concentric with the 22° halo were obtained. However, no “lower half” of this halo could be recorded:

(combination of 262 frames from 20:13 CET)

(combination of 141 frames from 20:17 CET)

(combination of 450 frames from 20:31 CET)

For comparison, the source videos can be viewed here (1, 2, 3). I also calculated conventional average stacks (1, 2, 3, gamma value increased to 1.5 for better visibility) from these 6 s – 18 s long sequences. As expected from the conventional long-time exposure photos, the 12° halo segments are completely washed out in the average stacks.

It seems quite likely that the observed segments belong to a “Moilanen circular halo”, i.e. the result of poorly oriented crystals with 34° wedge angles. The moon display proved that such crystals were around, and all halos were caused by a rather thin diamond dust cap wrapping closely the top of Mt. Klínovec, as illustrated by pictures taken from the nearby Mt. Fichtelberg. Near-ground turbulences generated by the rather strong wind will have caused the orientation randomization in the headlight display. If this reasoning turns out to be correct, the correspondence table between the Moilanen and 22° halo families can be completed in the following way (the first two lines were suggested by Nicolas Lefaudeux, as reported in Marko Riikonen’s “Halot”):

Moilanen arc ↔ Parry arc

Mikkilä arc ↔ upper tangent arc

this halo ↔ 22° circular halo