On 12. November routine plate stuff prevailed over large areas around Rovaniemi. As I drove around, in one location my eye caught an uppervex Parry arc so I stopped to photograph. Visible was also circumzenith arc and even though it was far from dazzling, I aimed the lens at zenith with Kern in mind. After all, you have to account for the sun elevation upon judging how good the cza really is. Those bright at higher sun may lose much of their lustre at low sun. And the lower the sun, the better for Kern. In this case the sun was low: elevation at 1315 local time was 4.6 degrees.
As shown above, the attempt was not in vain: there really seems to be a full circle around zenith. True, the Kern is extremely faint and there are banded artefacts in the image, so one must be careful here. But the regular circle seems different from these artefacts.
The image above shows three versions of a stack of 21 photos taken during 1m 59s (sun has not been tracked in any part of the process). One is blue-minus-red image and two others are made with background removal technique with two different gaussian blur and median noise values. Nicolas Lefaudeux had instructed me on this method. It is also described on his site. The fourth image is a simulation made with HaloPoint.
Because of the Kern’s visibility display rather close to the noise threshold, it would probably be best to refrain from further analysis. But because I made the simulation anyway, let’s say that the Kern’s relatively uniform intensity seems to point at crystals towards triangular shape and I had to use quite extreme triangularity in simulation as depicted by the crystal figure below.
The display has also exceptionally long Schulthess arcs – they extend all the way to 46° halo. The Lowitz oriented population used to simulate them gives also 46° contact arcs. In the actual display 46° contact arcs may be visible too, but it is hard to say because the artefact bands are oriented in similar direction.
For reference, below is also a single image straight from camera. Enhancing it (not shown) seemed also give out Kern arc, but it was even closer to noise threshold than the stack. In any case, the diamond dust was so uniform stacking did not really bring that much benefit.
As shown above, the attempt was not in vain: there really seems to be a full circle around zenith. True, the Kern is extremely faint and there are banded artefacts in the image, so one must be careful here. But the regular circle seems different from these artefacts.
The image above shows three versions of a stack of 21 photos taken during 1m 59s (sun has not been tracked in any part of the process). One is blue-minus-red image and two others are made with background removal technique with two different gaussian blur and median noise values. Nicolas Lefaudeux had instructed me on this method. It is also described on his site. The fourth image is a simulation made with HaloPoint.
Because of the Kern’s visibility display rather close to the noise threshold, it would probably be best to refrain from further analysis. But because I made the simulation anyway, let’s say that the Kern’s relatively uniform intensity seems to point at crystals towards triangular shape and I had to use quite extreme triangularity in simulation as depicted by the crystal figure below.
The display has also exceptionally long Schulthess arcs – they extend all the way to 46° halo. The Lowitz oriented population used to simulate them gives also 46° contact arcs. In the actual display 46° contact arcs may be visible too, but it is hard to say because the artefact bands are oriented in similar direction.
For reference, below is also a single image straight from camera. Enhancing it (not shown) seemed also give out Kern arc, but it was even closer to noise threshold than the stack. In any case, the diamond dust was so uniform stacking did not really bring that much benefit.