APOD: Unusual Light Pillars Over Latvia (2009 Jan 12)

[neufer]

Sigulda Publicity Stunt= Bright searchlight + Aerodium Vertical wind tunnel + snow crystal input (from bottom)

Correction:
Sigulda Publicity Stunt = Bright floodlights + Aerodium Vertical wind tunnel + water hose input (from top)

It is, in essence, the world's largest snow making machine:

So are you telling us that the publicity stunt you describe is actually responsible for the appearance of the 090112 APOD?

A winter ski resort is more impressive with snow on the ground.
(Even artificial snow. Note that there is not much snow on this side of the tracks!)
http://apod.nasa.gov/apod/ap090112.html

Facts:
1) clear cloudless skies
2) very bright light pillars
3) rapid divergence near top (=divergence of Aerodium Vertical winds)
4) tourist crazy ski resort
5) little to no snow on this side of the tracks!

If so, what is RJN's role in asking for explanations?

He's pulling an APOD/Asterisk Publicity Stunt (which seems to be working.)

[pauln]

Look at the whole picture! How many vertical wind tunnels have they put into this town? If any owls fly over they're going to end up in orbit!!

[ghiggi54]

I am not a scientist; I only play one at home. However, the unusual fanning of the light pillars would seem to have a fairly straightforward explanation. Using the Higginson method of reasoning,it is as follows:
Water crystals in the air are the cause of the pillars, of course. We can reasonably assume that the composition of the crystals would determine the visual size of the light pillar. The density of the crystals would probably be the most likely determining factor. This density can change with elevation. This change would either be a sudden change, or gradual. Because the density of the ice crystals would be one of the determining factors of the visual size of the pillars, a change in density would result in a change of the visual size. The gradual change of density would logically result in a gradual change in the size of the pillar. In this case, it has resulted in a pillar that seems to fan out at the top (the fan being the gradual change in perception). If the atmosphere and ice crystals were layered instead, and the changes in density were sudden, then we would probably see something that looks like a tall lamp with a lamp shade over it (the lamp shade being the suddenly larger light pillar).

[pauln]

I think you're right that the density of ice crystals would have a visual effect, but I think that the effect would be on the brightness of the pillar rather than the width. If you imagine that every crystal is exactly horizontal then you will only get light reflected to your eye from crystals that are exactly over the line between you and the light source. You would get a pillar that is exactly the same width as the light source. Few crystals would mean that little light is reflect to your eye and you would get a dim pillar. Many crystals would reflect a lot of light and you would get a bright pillar.

Now imagine that some of the crystals are a little off horizontal. Those ones will reflect light to your eye if they are a little away from the line between you and the light source. This is what will widen the pillar; mainly flat crystals with a mixture of slightly skewed ones. The more disturbed the crystals are the wider the pillar will be. Clearly the crystals near the top of the pillar are more disturbed than the ones below. This is most likely turbulence caused by turbulence in a shear zone between the still air on the ground and moving air above. In the moving air the crystals are tumbling in completely random orientations so you get no pillar at all. As they drop through the still air they stabilise into a flat orientation. This is why the very top of the pillar is wide and narrows rapidly as the crystals move out of the turbulent zone.

[ralperhus]

I would like to offer as a theory for the cause of the fanning out of light atop the light pillars, the word Backscatter.
To explain: the pillars themselves are refraction of the light between the light source and the viewer, right? Ok- but that which is refracting the light, may also be reflecting backwards, some of the light it is refracting... So, then that light is shining on the crystals which then reflect to the viewer in the shape of the fan based on the same refracting angles...

[nearbyvoicemail]

Upon closer inspection, the "pillars" not only flare out in the upper region of the photo, but also the light appears to travel in the pattern of a vertical wave, resembling the equal-division of a freely-vibrating string (harmonic ratios). Could the fact that heat rises (from the lights) create or affect the apparent vertical angle of the pillars?

[Muscipula]

Keshlam wrote

I'm trying to think through the optics of this image... the question is not just why the column broadens at the top, but why we're getting this flare only in the vertical direction and specifically upward.

It is *what* is happening, and not the reason that is omenes.

Muscipula

[chadair]

This is most likely turbulence caused by turbulence in a shear zone between the still air on the ground and moving air above.

If this were true, then the fans would appear at a constant height above the ground, not an angular height relative to the observer. As you can barely see from the background of the APOD image, and more clearly in the original image posts (http://spaceweather.com/submissions/lar ... 501854.jpg), the fans and the bright cores of the column are all at a fixed angle. Thus this phenomena is related to the light path not the meteorological structure.

I cannot find the photographers contact info, but I'm still looking for the lens used to deduce the angles in the image.

[inalaska2]

i used to live in Anchorage Alaska, 1988-1997. this picture is very familiar to me. during the winter months, when it was dark 1700-0800, you could see this phenomena. in Anchorage the color was always on the pink side, i use to think of it as the Dr Seuss effect. Americans will understand that joke. the cloud layer/ice cloud was always very low over the city when it happened, never on clear nights. and the pink light off the street lights would extend both down toward the ground and up, disappearing into the low hanging fog/cloud. i never noticed the dispersal of the light as shown in this picture, but then the layer above the lights was very low.

[pauln]

This is most likely turbulence caused by turbulence in a shear zone between the still air on the ground and moving air above.

If this were true, then the fans would appear at a constant height above the ground, not an angular height relative to the observer.

Yes, that's a good point. It's hard to get an accurate idea of the relative distance to each of the two most pronounced light sources given the obstructions and the low perspective. They may be similar distances - both of them are roughly the same distance from the bottom of the frame - then again they may not! When you make contact with the photographer, you could perhaps ask if they have other shots. Shots from different places might show the two lights at different relative distances. Alternatively shots taken facing different directions might include dimmer and more distant pillars with fans.

[ralperhus]

If I may hammer my theory a little more: The main vertical column of light is deriving its light from the main point-source. The fan at the top, is derived from scattered light from all directions that the light is scattered and reflected, meaning all directions the light shines. So, light scattered back from a ray at 90 degrees from the viewer does get difraction to a lesser extent than those more aimed directly at the viewer... Maybe I am not saying it so very communicatively... The fan is a conic section of the entire 360 degree output of the lamp.

[chadair]

Using stellarium centered in on Sigulda, Latvia, on Dec 28th 2008 it looks like the pictures were taken at around 12 to 12:30am (the bright stars on the 3rd picture, just left of center that are aligned nearly vertical are Deneb and Sadr, Vega is in the lower right). The APOD picture appears to have been taken first, 2~ hours earlier (I think the only visible star is Vega, which moved from the upper left to the lower right of the white pillar in the other pictures).

I never found the lens focal length, but PiTHON recognized a few stars in the image from Cygnus, and that is enough to figure out image scale. From the APOD image I can then deduce image scale- it is a little distorted as it is a wide angle lens (roughly 17mm), but close enough. The bright streaks in the pillars are 16 angular degrees above the light source, almost uniformly across the image. The fan begins to diverge from the "beam" at about 17 degrees above the light source. The geometry may be complicated a bit since the angle toward the light is NOT parallel to Earth surface and thus NOT parallel to the hexagonal ice crystal alignment (which just be a factor in the cause of the trumpet shape?). Most of the lights are depressed about 1-2 degrees below the horizon.

Looking at GUINEY72's suggestion that two arcs can be inscribed to simulate the beam shape, I come up with an arc of uniform radius equaling roughly 40 degrees.

This does not match the often found 22 degree angle in the Lowitz Arcs, however the arc radius of 40 degrees measured off the picture is getting close to the rainbow angle of 42 degrees relevant to water droplets- though that may be a coincidence.

Chad

[chadair]

It's hard to get an accurate idea of the relative distance to each of the two most pronounced light sources given the obstructions and the low perspective. They may be similar distances - both of them are roughly the same distance from the bottom of the frame - then again they may not! When you make contact with the photographer, you could perhaps ask if they have other shots. Shots from different places might show the two lights at different relative distances. Alternatively shots taken facing different directions might include dimmer and more distant pillars with fans.

Look closely at the original post (linked above) and you will see faint fans far in the background that are all at the same angular elevation.

[lakeside]

There are a number of things going on here. It appears that the light columns are terminated by an overlying cloud deck. The ice crystals presumably are being nucleated at the interface between the lower levels of the cloud deck and the colder air near the ground.
The increase of crystal aspect ratio (the ratio of width to thickness) probably causes improved preferred orientation of the crystals as they grow and descend.. Where less well oriented higher up, near their points of nucleation,they should form a wider reflection pattern. The rays, however are another matter. It is well to remember that these are crystals with sharp points suggesting that electrostatic forces may cause preferred orientation about their vertical axes as their points repel each other. Such preferred orientation within horizontal planes would be expected to cause some form of ray like effect.

J White

[rockwiler]

I think that crystal growth with changing crystal forms is the explanation.

Lets divide the picture into three zones. The top zone is above the spreading arcs of light. The middle zone is where the spreading arcs are. The lower zone is where columnar light dominates.

I would surmise that in the upper zone humid air is crystallizing minute ice crystals that are NOT perfectly flat plates. I would guess they start as shallow dipyramids with a short prism, that is the crystal is like a very short hexagonal cylinder, but rather than having flat end faces it has hexagonal pyramids on each end. This would make them like quartz crystals, except that the pyramids would be crowned by a very obtuse angle rather than an cute angle or right angle. Alternatively, the crystals may be hoppered, which is to say they are recessed on the ends in an inverse pyramid, probably with similar angles to the pyramid but inverted toward the center of the crystal. Either way, the crystals have one or other of these shapes as they fall into the middle zone. I think the explanation actually works better with hoppered faces.

In zone two crystal growth continues, but in this zone the crystal growth no longer supports the pyramid face or hopper faces. Instead, only the common flat end face is able to grow. As the flat faces overtake the pyramid or hopper faces they form with curved faces for a period equivalent to their time falling through zone two. The curved faces cause the arc-like reflections. Once the angled faces are over taken by the new flat growth the crystals lose their ability to disperse light in those arcs.

At this point the crystals have reached the bottom of the middle zone and continue to fall to the ground. Their geometries have probably largely stabilized as the familiar flat hexagonal plates. They produce a typical straight column of light.

I know some will object that crystals don't have curved faces; that the faces are always flat. Mathematically this is true, but curved faces are found in nature, especially in certain minerals like calcite. I have an extensive mineral collection and I could get more than one example off my shelves at any moment. I suspect that it is hoppered crystals filling in with curved faces that is the key to this.

As for the tighter convergence lower on the columns in the bottom zone, perhaps there is some remnant curvature in the end faces that is not completely overcome by later crystal growth, and as the crystals fall the light reflects differently from the faces depending on the viewing angle.

[jaguar]

The "fog" in the picture is probably ice fog. Ice fog forms when the surface gets colder than the dew point of the air/moisture mix adjacent to the surface. This effect is rather common in very cold locations where cold air is caught in a bowl-like area (Fairbanks, AK, with -40C ground temperatures). The normal temperature gradient when ice fog is present is that the surface is coldest with the air temperature gradually increasing as you rise above the surface. There is often a point some distance above the surface where the temperature is 'warm' enough that the ice crystals no longer form. If there is any breeze at the upper surface of the cold air (the inversion layer), there will be some irregular tops to the fog. The temperature above the layer of cold air may be tens of degrees warmer. So, the ice crystals are colder at the surface, with all the strutural and other effects that may be related to that. At the uppermost surface, there may be a mix of ice crystals and water droplets.

[DougStern]

Have lived in Arctic Canada [68 degrees N Lat] 26 years and see these pillars from lights in town many, many times a year. Yes, the result of thick ice crystals with no wind. But have NEVER seen the fanning out at the top. I think you have a very rare ice crystal phenomena here. Sorry but have NO idea why the pillars fan out. Will definitely be on the lookout [with my camera]. Amazing.

[neufer]

Look at the whole picture! How many vertical wind tunnels have they put into this town? If any owls fly over they're going to end up in orbit!!

Two vertical wind tunnels to send the snow up.

The others are actual normal light pillars from the snow coming back down.

[pmanson]

Now that several people on this forum have agreed with my earlier explanation, I'm going to disagree with myself

Looking more at the extra photos linked from the APOD page, the "fan" section looks too crisp and orderly to be just the result of dispersed orientations of the crystals higher up -- it looks more like an optical phenomenon. I know this is horribly unscientific, so I ran some HaloSim simulations. Have a look at this one:



The pillar and fan parts look pretty close to the photos. I'd encourage people to play around with HaloSim and see if they can come up with a simulation that's even closer to the photos. If this simulation is what's happening, the pillar is caused by horizontally-oriented plate crystals, as I described in my original post, and the fan part is an Upper Tangent Arc, caused by horizontally-oriented column crystals. The shape of the arc is dependent on the altitude of the light source relative to the observer - a higher light source (or lower observer) produces a wider arc. In the simulation, I used a light source 0.2 degrees in size at 8 degrees below the horizon. It looks from the photos like the shots were taken from an upper floor of a building, so having the light source below the "horizon" makes sense from the point of view of the orientation of the crystals.

My simulation file is available here.

[keshlam]

pmanson: I agree, that simulation does look a lot like what we're seeing in the photo.

[rockwiler]

It is important to keep these points in mind for any process-related explanation:

The ice crystals are falling.

As they fall they generate a changing optical effect.

Therefore, as they fall, either:

a) the form of the crystals is changing

or

b) they are interacting differently with each other or the atmosphere over the course of their fall.

or

c) both of the above.

If the explanation is a steady-state optical solution the above points can be disregarded.

[Jim R Feliciano]

I know I do not know anything about the crystals that make the pillars. I wonder if the shape of the lamps might have something to do with the upper shape of the pillars. I the lamps are shaped like a bowl or even a trapezoid, then maybe the fanning at the top of the pillars is caused by the direction of the light rays from the original surface. This simple idea might explain why the pillars seem to squeeze at a certain point. When I learned the little about optics that I know, I was told to remember that light travels in straight lines. I think the shape of the lamps and the shape of the crystals work together to make the fanning effect.

[lakeside]

Well, I have heard too much and am now willing to concede (especially after examining the large image) that this is an optical effect and probably is not related to crystal growth. My prediction is that as the observer approaches the light source the fanning out of the light column increases until at the light source it is a full circle overhead. Crystal orientation is of course necessary to all this but the explanation should be left to the mathematicians in the group!

J White

[jteutsch]

looking at the light (red) to the far left of the image posted on http://apod.nasa.gov/apod/ap090112.html one can observe the light acting as though it is passing through a lens. It starts out at the source (ground) then "diffuses out", "re-focuses" and then converges above the tree line. (at a focal point). Looks like the second mode shape of a cantileverd mechanical beam.

[Stephen McDonald]

The pillars aren't actually above the lights, this is an optical illusion. They couldn't be above the lights in reality because the lights have a reflector dish on top sending all of the light downward. The ice crystals which are between the light and the viewer are refracting the light in the same fashion as a water mist produces a rainbow and the same way that upper atomosphere ice clouds create a halo around the moon. The fanout is due to the angle of reflection being sharper at a greater height. As an example, there are several faint pillars in the background which are faint because of the greater distance, but those fainter pillars just appear to fade into the sky much higher than the nearby brighter ones and the faint ones have no fan-out on top: the viewing angle for the top is greater than for the close up ones.