APOD: A Triple Sunrise Over Gdansk Bay (2009 Aug 04)

[green1]

[quote="Naos"] "why do you use a point source at finite distance to represent the sun in your last drawings ? it seems quite strange to me.
I would have used a set of collimated beams coming from a direction"
Naos, the rays are representational only, not intended to be to scale.

"I have tested with Zemax optical calculation software your idea of "NOT perfectly parallel panes of glass". This kind of configuration CAN produce ghosts on the Left OR on the right depending on the sign of the tilt angle between the two panes of glass. "

You don't say how much variation is required. This is obviously a large window or sliding door. Even if it is only 3' square the variance is going to be extremely small. With a 1/4" difference in gap between panes there will not be enough angle to reflect images to the left unless the window is at 90 degrees to the sun, and the camera is perpendicular to the window. Unless you think Poland is still a third world country, to imagine that any commercially available thermal window would have these types of flaws is stretching it a bit.

I would challenge anyone who thinks that this is due to reflections in a thermal pane window to duplicate the effect. If this is window reflection it should be able to be replicated at any latitude at dawn or sunset.

[geckzilla]

"I have tested with Zemax optical calculation software your idea of "NOT perfectly parallel panes of glass". This kind of configuration CAN produce ghosts on the Left OR on the right depending on the sign of the tilt angle between the two panes of glass. "

You don't say how much variation is required. This is obviously a large window or sliding door. Even if it is only 3' square the variance is going to be extremely small. With a 1/4" difference in gap between panes there will not be enough angle to reflect images to the left unless the window is at 90 degrees to the sun, and the camera is perpendicular to the window. Unless you think Poland is still a third world country, to imagine that any commercially available thermal window would have these types of flaws is stretching it a bit.

You still haven't tried the experiment, have you?

Another thing I have noticed that greatly affects the reflections is that the distance between the source light and the panes of glass causes them to diverge even further than when I was only a few feet away. So something like the sun which is quite far away I predict would require very little variance to produce a divergence of reflections. And why do you assume that only third world countries can have flawed windows? A lot of things could have caused the window to be imperfect. It could have been like that since it was constructed or it could have even happened over time. I'm going to send the Diacons an email to ask if they know about the raging debate going on about their windows.

[tlc]

The structure of the lenses of cameras (digital or not) and the iris of the camera lens are such that I would expect to see the sun's secondary and tertiary images shaped like the iris shape. That is seen clearly in many photographs where the sun is being reflected within the lens and the hexagonal or pentagonal (depending on the number of iris leaves) reflections cascade across the image from a corner. The number of images is dependent on the number of surfaces within the lens. Two is an unusually small number for the complex lenses of today.

Also, in order for the reflections or ghost images to occur, the bright point of light needs to be somewhat off center. Given the position of the sun in this image, I would not expect any ghost images or a large overlap between the images.

One thing in favor of an in-camera effect is that the second image clearly shows the boundary between the sun and the water surface. I'm not sure how a lensing effect could do that.

[Chris Peterson]

The structure of the lenses of cameras (digital or not) and the iris of the camera lens are such that I would expect to see the sun's secondary and tertiary images shaped like the iris shape.

And indeed, internal lens reflections are evident in the first image, down and to the right from the Sun.

The triple image of the Sun itself is certainly not a lens effect, because the photographer reported seeing the effect herself, which is what prompted making the images in the first place.

One thing in favor of an in-camera effect is that the second image clearly shows the boundary between the sun and the water surface. I'm not sure how a lensing effect could do that.

This doesn't argue for an in-camera effect, it merely proves that it isn't an atmospheric effect.

[geckzilla]

I tried using Barry's email address on the APOD page to contact him but he is "away from work" from July 31st to August 18th. You may not think this is pertinent to the discussion, but consider this... what if they are on vacation? They could have shot that photo from a hotel room, not an apartment. That would mean they haven't experienced very many sunrises from that location. So perhaps they have only viewed the sunrise from this window once. It would be easy to think it was extraordinary and not a normal, everyday occurrence at that window.

[Naos]

I made some exports from Zemax.

EXPORT2.GIF (10.67 KiB) Viewed 1852 times

EXPORT1.GIF (9.87 KiB) Viewed 1855 times

The sign of the tilt determines the sign of the deviation of the ghosts. Left or right deviation does not depend on the global orientation of the window, and it only depends on the value and sign of the tilt between the two panes of glass.

The deviation is equal to 2 times the tilt angle. Distance between 1st and 2nd suns is about 0.75 deg, so the tilt would be 0.375 deg.
If the window is 1m wide, the gap variation is about 6.5mm which is quite big indeed, compared to typical value of gap (16 mm) inside double glazing.

[Eamon Shute]

The displacement of the second image from the first is less than one degree. The two sheets of glass need to deviate by half this amount, ie roughly one part in 150. This is 1mm over 15cm, which is perfectly plausible for a short distance. The fact that the third image is closer to the second shows that the glass is not perfectly flat, so the angle varies significantly over a short distance, which is not surprising.

The effect is certainly caused by multiple reflections in double glazing, nobody has come up with a valid objection or a sensible alternative. I am surprised that NASA bothered to consult the public when the explanation is so obvious.

[Eamon Shute]

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

[dockwatcher]

Check the boat or oil drick- whatever.

Ignore the double stack; there is an obvious waver in the verticle lines. It looks like a marage senerio would explain at least one image.

[Alnilam]

To geckzilla and others :

Some pictures published here are quite strange in they ray-tracing !!!

Don't forget the reflection rule : the angle of reflection is equal to the angle of incidence !

[green1]

Naos, I'm trying to make sense of your computer simulations.
Where is the light source?
You are showing (in figure one) converging rays which seem to bounce around almost endlessly inside the glass pane then diverge again when exiting the glass? In figure two the rays are diverging into the glass and converging away from the glass. Neither of these is correct. If the rays are diverging before entering (nominally) flat panes of glass they will continue diverging when exiting the glass. The converse is also true.

[orin stepanek]

http://apod.nasa.gov/apod/ap090804.html
Click on Barry; see what happens. I tried and got an email site. I tried to use it to ask about the window thing! It wouldn't send so I guess the debate goes on.

Orin

[Naos]

I only take one ray (representing the direction of the sun). The ray comes from the left of the drawing and goes to right.
This ray hits a a window composed of two surfaces that can reflect and transmit light. The angle of incidence choosen here is 45°
And last : the two surfaces of the windows are NOT PARALLEL. (here 1° tilt angle)

Then I let Zemax compute the ray-tracing.

The results are the set of rays that exit to the right of the drawing.
Their angles from the initial ray direction are the deviations of the "ghosts".

[Chris Peterson]

I made some exports from Zemax...

You are not setting up the system correctly. As presented, it is afocal, and you are (apparently) making the assumption that each exiting ray somehow represents an image of the source. That is not the case.

The fact is that the direction of multiple images seen through a pair of glass plates it completely independent of the orientation of those plates with respect to the source and observer. It depends only on the orientation of the two plates with respect to each other. If they are parallel, there will effectively be no displacement of images (or rather, a single displacement that manifests as a slight fuzziness of the edges). If they are not parallel, the displacement of the images will be in the direction that the plates converge. In practice, with typical double glazed panes, the direction of displacement will vary as you look through different areas of the glass- something that is completely consistent with the images presented. Since the camera was handheld, each shot was through a slightly different region of the window, and as a result each shows a slightly different relationship between the reflected images.

The images are completely consistent with shots made through a double glazed window, regardless of the orientation of that window.

[Naos]

I completely agree with your analysis :

The fact is that the direction of multiple images seen through a pair of glass plates it completely independent of the orientation of those plates with respect to the source and observer. It depends only on the orientation of the two plates with respect to each other. If they are parallel, there will effectively be no displacement of images (or rather, a single displacement that manifests as a slight fuzziness of the edges). If they are not parallel, the displacement of the images will be in the direction that the plates converge. In practice, with typical double glazed panes, the direction of displacement will vary as you look through different areas of the glass- something that is completely consistent with the images presented. Since the camera was handheld, each shot was through a slightly different region of the window, and as a result each shows a slightly different relationship between the reflected images.

The images are completely consistent with shots made through a double glazed window, regardless of the orientation of that window.

However I cannot understand why you reject my simulation. To have a focal system, you just have to put a lens at the exit to convert direction into position on a detector, and thus you have the camera looking at the sun (except its angular diameter) through the window, no ?

[Chris Peterson]

However I cannot understand why you reject my simulation. To have a focal system, you just have to put a lens at the exit to convert direction into position on a detector, and thus you have the camera looking at the sun (except its angular diameter) through the window, no ?

I'm only quibbling with it because it gives a somewhat confusing picture of the actual situation. It's showing the edge smear effect, but not the image displacement effect. The source needs to be shown impinging on the entire surface (as a set of parallel rays), and then the resulting image needs to be determined, which naturally requires including an imaging system in the simulation.

[polymath]

The foreground clouds appear to be stratocumulus lenticularis or undulatus. The clouds on the horizon in front of the sun appear to be the same. The open expanse of sky in between is over Gdansk Bay. At sunrise, large bodies of water are warmer than adjacent land. The slight difference in air temperature over the masses allows cloud formation over land and clear sky over water. The shape of the foreground clouds suggests the wind is relatively calm at stratocumulus heights of 2000 to 2400 m, 7000 to 8000 ft. The distant cloud deck appears and probably is above land across Gdansk Bay, roughly 100 kilometers away and potentially reaching several hundreds of kilometers inland.

Water droplets and/or ice crystals in the distant clouds act like a planetary-scale beam splitter, splitting the solar image into two solar mirages.

I've seen three rainbows in the sky at sunrise. A light drizzle was falling, the horizon was clear. Like a majestic tunnel, the three rainbows formed one after the other.

[a24kgoldphoto]

Sundogs mock suns or parhelia

Sundogs, also known as mock suns or "parhelia", are a pair of brightly colored spots, one on either side of the sun.


Photograph by: Rauber Sundogs form as sunlight is refracted by hexagonal plate-like ice crystals with diameters larger than 30 micrometers and their flat faces horizontally oriented.

Sundogs are visible when the sun is near the horizon and on the same horizontal plane as the observer and the ice crystals. As sunlight passes through the ice crystals, it is bent by 22 degrees before reaching our eyes, much like what happens with 22 degree halos. This bending of light results in the formation of a sundog.

The difference between sundogs and halos is the preferential orientation of the ice crystals through which the light passes before reaching our eyes. If the hexagonal crystals are oriented with their flat faces horizontal, a sundog is observed. If the hexagonal crystals are randomly oriented, a halo is observed.


halos
Terms for using data resources. CD-ROM available.
Credits and Acknowledgments for WW2010.
Department of Atmospheric Sciences (DAS) at
the University of Illinois at Urbana-Champaign.

sun pillars

[tiptuffer]

This couldn't be caused by a pane of glass. The shadow image has to occur before the clouds that you see on the horizon. If they occured at the glass the irregular winged shaped halo caused by the sunlight hitting the clouds in front of it would have been shadowed as well. This is not the case. The second and third image have a better circle shape than the first sun.

[tnuck]

I don't have an oppinion on any atmospheric phenomenon that may cause this, but I'm an ex professional photographer and mechanical test engineer well aquainted with cameras and digital photography. Any image can be concocted in a digital photo-manipulation program and there is no way to tell from this screen shot if that has been done. That said, if the image is truly a one single, un-digitally-manipulated image, it is very unlikely that the effect was the result of anything inside the camera. I say this based on the presense of the cloud bank in front of the two duplicate suns. If the images were a ghosting, the cloud bank would be the same in both and would be very unlikely to line up with the clouds surrounding the ghost images. This also implies that whatever atmospheric effects are responsible, they are taking place behind (above) the cloudbanks.

[wbowman]

Perhaps horizontal atmospheric lensing is caused by the tilt of the earth at northern latitudes. I'll bet that it doesn't occur at the equator. The vertical images suggest a thermal inversion.

[Chris Peterson]

If the images were a ghosting, the cloud bank would be the same in both and would be very unlikely to line up with the clouds surrounding the ghost images.

And that's just what we see: the image of the clouds in front of the primary solar image are repeated in the two ghost reflections. This isn't hard to see, but you have to separate the reflected cloud images from the actual cloud pattern in the sky behind the ghost images. This is additional hard evidence that we are seeing reflections from the window in front of the camera.

[green1]

I will post one more illustration to show how colder polar air could refract the light to give the result of multiple sun images. Remember this is viewed through the cold air "lens" at an oblique angle that will increase the effect. The red dot is the True North Pole, the white disk represents the colder polar air mass (s). (For all the nit-pickers: this is not to scale; only a visual aid to help explain the effect.)
Thanks to Google earth for the image of the globe, APOD & NASA for the images of the sun and moon.

arctic lens small.jpg (89.99 KiB) Viewed 1584 times

[Chris Peterson]

I will post one more illustration to show how colder polar air could refract the light to give the result of multiple sun images.

Although it is pretty obvious that this isn't an atmospheric phenomenon, it is quite certain that it can't be a refractive one at all. Refractive processes in the atmosphere are necessarily dispersive, and there is no sign of dispersion in the ghost images. This must be a reflective phenomenon, regardless of whether it occurs in the glass or by some completely unknown atmospheric process.

[green1]

[quote="Chris Peterson: Although it is pretty obvious that this isn't an atmospheric phenomenon, it is quite certain that it can't be a refractive one at all."]

Chris, you are forgetting that ALL the images in the photo are refracted images. The primary image is refracted (yes, dissipated) through the "lens" of the earth's atmosphere. Otherwise it would not be possible to view or photograph without protection (Safe Sun? ) So it is not a stretch to believe that there could be more than one angle of refraction.