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

[nausetmarsh]

Tangent Alert: Perhaps this is the phenomenon to which the protagonist in Schubert's "Die Winterreise" (The Winter Journey)refers in what is, I believe, the last song, "The Three Suns". According to the storyline, the narrator goes crazy which is why he sees three suns at once. Maybe someone more enterprising than I would be interested in researching whether the *actual* (as opposed to the narrator) poet might have observed such a thing. If the phenomenon has anything to do with heat, it would be less likely to occur in the winter, I suppose . . .

[socaldad]

The idea that the cloud patterns are different in front of each of the suns is very hard to prove since you are seeing both the reflection (from where ever) and the primary image superimposed. No real way to separate them. The only problem that I have with the window theory is the difference between pictures 9805 and 9810. If the observer did not move then the distance between the second and third images in each picture would remain constant. If the observer moved closer or further from the window then changes in spacing would make sense.
Since we have no reason to suppose that the observer is intentionally striving to deceive then it is most likely that they did not realize that a window could cause this phenomina. If there was a window in the optical path then problem solved. If there was no window then WOW this is really interesting.

[danofold]

It is refreshing to hear an actual first-hand report on this phenomena. Having said that, I must join those who have demonstrated that this time we are observing a sunrise/sunset though a double pane window/door.
I will also state that the photographer was not deliberately misleading us or APOD. I have seen a few of these 'atmospheric phenomena' discussions resulting in the through-the-window effect.

[apodman]

(obnoxious boldface blah blah blah)

Yeah, bold guy, your opinion is much more important than anyone else's, so make sure we see it. Next time why not large print, full caps, and color? Sorry we can't make it flash for you, but you could always convert it to a flashing .gif file.

---

BTW, phenomenon is singular, phenomena is plural.

[TheSciencePundit]

@DaveB

As the two left most suns clearly have different clouds in front of them I cannot see that it is possible for the effect to have been caused by glass close to the camera.

Clearly? Not so clearly to those on this board who've looked closely at those pictures with tools like Photoshop.

@dockwatcher

The reflection is 10% smaller. Does that rule out plate glass?

Not at all. In fact, it's what one would expect if the glass was at an angle with respect to plane of vision of the camera. In other words, even if the reflected images of the sun are the same size as the projected image on the glass itself, the reflected images are farther from the camera and hence appear smaller.

the reflections seam fuzzier as they go out. If it was glass wouldn't they stay sharp?

The reflections aren't as bright as the original and so don't drown out the background as much, hence apearing fuzzier.

[Elias Chasiotis]

From the glossary of the book "Color and light in nature" by D. Lynch and W. Livingston (Cambridge University Press, second edition, 2001), p.266:
"multiple suns (moons) unexplained occurrences of multiple images of the sun".

[drollere]

this is an internal reflection in multilayer architectural glass, either shatterproof or thermal layered, or both. the images are far too stable and invariant to be atmospheric, and the fact that the person was inside a room, rather than outdoors, pretty much clinches it.

for this to be an atmospheric phenomenon, there would have to be multiple perfectly vertical and perfectly flat (not turbulent) thermal discontinuities in the atmosphere. that just doesn't happen.

the reflection also exactly reproduces the very small "second sun" just below the virtual image, which is caused by an atmospheric inversion at a great distance. an atmospheric reflection would "see" this inversion from a different angle, and therefore either not reproduce it, or reproduce it in a different form. the fact that it is identical indicates the reflections are local to the viewer. the sun's reflection in the water is also reflected: you can verify this by putting the image in photoshop and boosting the saturation or contrast. an atmospheric reflection would not reproduce this reflection, since its locus would be physically much higher in the sky (as the image shows, many times higher in the sky than the height of an ocean drilling rig).

the geometry of linear perspective demonstrates that the sun's image is always at the eye level of the viewer, for all positions of the viewer at low altitudes (sitting, standing, different floors of a building, etc.). a glass perpendicular to the direction of view toward the sun would also be perpendicular to the ground, and produce no vertical reflections. the only way the reflection could occur is if the glass were slightly oblique sideways to the sun's rays.

note that the angle of the sun to the glass changes with the solar position on the ecliptic, so this illusion would occur within a limited time only twice a year (or only once a year, if the sun's rays are perpendicular at a solstice). the horizontal displacement indicates the plane of the window is slightly oblique; given the latitude of the viewer, the location of the sun in the ecliptic, and the angular displacement of the reflections at a given distance from the window, and the thickness of the internal layers of the glass, you could calculate the angle of the wall to true north. the smaller vertical displacement indicates the window is not plumb, or the casing is tilted (e.g., in the supplementary photos, the railway car is on a banked curve). due to the seasonal path of the sun, vagaries of weather, lifestyle and attention might mean it is years before an occupant notices the effect.

next time naomi sees this phenomenon, she should first step onto her balcony to photograph it.

[drollere]

The reflection is 10% smaller. Does that rule out plate glass?

no. i'd guess the glass is shatterproof, which means there is a layer of soft or flexible plastic between two sheets of glass. we don't know the optical properties of this layer, but the light imaging the reflection had to pass through the layer 3 times (front to back, back to front, and then front to back before exiting out the back), so it is optically really 3 times thicker.

refraction occurs at the boundary between two transmission media, and there are altogether six boundaries in this image (two to pass through the front and back glass layers, and two for the internal reflection). that, and the rubbery layer, are more than enough to degrade the image.

[silentbanshe]

In this case, might it not be due to the curvature of light and the earth. In non-euclidian geometry light is said to be curved rather than straight, and so is the earth. Might this not be caused by the two curvatures creating an at their "edges" creating an odd angle, and the light is thereby reflected onto water particles in the air. They have been known to be reflective. Maybe an odd angle caused these particles to reflect the image of the sun for a brief instant.

[kamoses]

In the attached image, I took a flashlight and a scanner with two clear windows, much like a 2-pane window, and shined a flashlight through it and took a picture in the same orientation as the sun, window, and camera in the picture. What you will notice is the the reflected images go to the RIGHT, not the left. Can one of the window people explain why the sun would reflect differently in the picture?

[silentbanshe]

"this is an internal reflection in multilayer architectural glass"
But for it to be a reflection from glass, wouldn't the picture need to be taken from an angle not perpendicular to the glass? If it was through glass, there wouldn't and at a non-perpendicular angle, it would not cause a reflection of the image but rather a glare. In this case it must be atmospheric.

[orin stepanek]

If caused by glass; The photographer could repeat it again morning after morning. Wouldn't that have been noticed the next night? I still wish Barry or Noemi would reply as to whether or not the photo was in or out doors.
corrected phrasing!
Orin

[backstagediver]

@TheSciencePundit:
I believe you are wrong on all of your points. I agree with DaveB that the clouds are different in front of each of the reflections. It may be hard to see from the first three, but 9810 has distinctly different clouds. This cannot be accounted for by a reflection by windows.

As for a reflection from a window pane reducing the apparent size of the sun by 10%, your theory is correct in a sense, but if you apply math to your theory, it is wrong. In order for an image to be reduced in size by 10%, the image of the object needs to be 10% farther away. The sun is ~93 million miles from earth. 10% farther is ~102 million miles, an increase of about 9 million miles. For a double-paned window to make the sun appear 10% smaller, the two panes would have to be separated by 4.5 million miles. I don't think they make windows that big.
Now this itself does not rule out the idea that a window created the reflection. Convex surfaces on the inside of the window could account for apparent size changes, but I believe this is incorrect and the effect is created by the atomsphere.

Your comment about the reflections being fuzzier does not really address dockwatcher's observation. He was referring the the blurring of the edges more as the images progress, not the transparency of them.

Edit: kamoses beat me to this part.
Also, I can easily debunk the window reflection, as the reflections are on the wrong side of the image. The pictures show that the proposed window would be angled, with the far side on the left of the photo and the near side on the right. This picture illustrates.
(/Edit)

I believe this effect is caused by thermoclines or some other boundary of air masses with significant differences in density and therefore optical properties. I see similar effects when scuba diving. Thermoclines or haloclines can cause mirages and occasionally reflections between layers of water, and is not just horizontal planes, but can appear vertically too. This effect can be extrapolated to appear in the atmosphere on much larger scales since air transmits light more easily than water. Also the fact that the reflections are shifted vertically upward requires the reflecting surface to be angled slightly towards the viewer at the top. This would be unlikely for a window, but I see no reason for atmosphere to do so. We see it happen all the time over road surfaces, so why not in atmospheric boundaries?

[geckzilla]

In the attached image, I took a flashlight and a scanner with two clear windows, much like a 2-pane window, and shined a flashlight through it and took a picture in the same orientation as the sun, window, and camera in the picture. What you will notice is the the reflected images go to the RIGHT, not the left. Can one of the window people explain why the sun would reflect differently in the picture?

I took two panes of glass out of some picture frames and tried to make some reflections. I'm not totally sure about it but when I did this, I could move the reflection up, down, left, and right by slightly altering one pane so that it wasn't quite parallel to the other piece of glass. It seems to have more to do with what angle the two glasses have to each other rather than the viewer's angle to the glass. In fact, as long as I maintained the same angle between the two panes, regardless of which way I moved them together, the reflection always maintained the same position. If this is true, then it would have been even easier to fool the photographer since the viewer's position doesn't matter.

[Sarba Guha]

My guess is that it is a reflection of the sun "from the ocean" back into the atmosphere. However, it is the part of the ocean that has "curved around" beyond the horizon that we are looking at. A certain part of it will be more "normal" to the sun rays than the ocean that the camera captured (or the part we can see) and therefore, has a greater chance of reflecting the rays back to the atmosphere in its own direction.

Now, that might happen every day but it is only when the "clouds" in the sky are perfect for our line of sight, that it will be reflected "back to us" again. That is what "could be" happening. Therefore, we get "multiple suns". The third sun to the left is less bright because the sun rays strike the waters at a much greater angle to create it and therefore it is less intense.

There could by many such reflections back but an individual sees only one reflection from his vantage point (just as everyone sees a "different rainbow").

The problem with my own explanation is that if I were right, there should be other suns on the "right hand side as well". I am assuming that in this case the clouds only on the left were "just right" for it to happen "in this particular case".

Unfortunately, the clouds do seem to be evenly distributed on both sides and that makes my argument rather shaky. I have been seeing APOD regularly for the last six years at least but have to admit that I am no physicist.

Further Question from me: Is the brightest spot really the sun or is it also a reflection from the sun which is yet to rise (or is "just emerging")? It looks so bright that it is probably real.

[Hoyos]

It took me a while to go through all posts.
Here goes my 2 cents.

Against atmospheric-reflection theory:
It seems that the "Etruscan vase" effect is the same in all mirages. Since this is a nearby effect caused by refraction of the light in the nearby ocean, it is therefore very unplausible that reflections are happening in the outer atmosphere.
Birefringence seem unlikely, too. We are seeing 3 images instead of 2. Is there a trirefringence? Also, it would require fine tuning to make the images so different in brightness.

As well put by geckzilla, cloud patterns cannot be used here. Unless one goes through a Fourier transform of the entire image and analyze it.

Comments on the glass-reflection theory:
No doubt there is a glass in between as pointed out with no doubts by MadCat-75.
This theory also explain the identical Etruscan-vase patterns.
It is only strange that the reflections are on the LEFT side of the sun while it was expected to be on the RIGHT side as pointed out by blastoff.
So, to solve this, it would be nice if Noemi and Barry could tell us if the glass is tilted clockwise (as one would naively guess from the pictures point of view) or counterclockwise (which would explain the images being in the left side).

It has been a nice discussion and hope we can solve it
Cheers,

[Eamon Shute]

@TheSciencePundit:
I believe you are wrong on all of your points. I agree with DaveB that the clouds are different in front of each of the reflections. It may be hard to see from the first three, but 9810 has distinctly different clouds.

It is not possible to say that, as the main image is severely overexposed, but both images have a band of cloud running just below the middle of the image.

As for a reflection from a window pane reducing the apparent size of the sun by 10%, your theory is correct in a sense, but if you apply math to your theory, it is wrong. In order for an image to be reduced in size by 10%, the image of the object needs to be 10% farther away. The sun is ~93 million miles from earth. 10% farther is ~102 million miles, an increase of about 9 million miles. For a double-paned window to make the sun appear 10% smaller, the two panes would have to be separated by 4.5 million miles. I don't think they make windows that big.
Now this itself does not rule out the idea that a window created the reflection. Convex surfaces on the inside of the window could account for apparent size changes, but I believe this is incorrect and the effect is created by the atomsphere.

The increased size is caused by flare - the sun is so bright that light is scattered around the image. It is nothing to do with the atmosphere - it also happens with photos taken in space.

Your comment about the reflections being fuzzier does not really address dockwatcher's observation. He was referring the the blurring of the edges more as the images progress, not the transparency of them.

The blurring can be explained by the glass not being perfectly flat.

Also, I can easily debunk the window reflection, as the reflections are on the wrong side of the image. The pictures show that the proposed window would be angled, with the far side on the left of the photo and the near side on the right. This picture illustrates.
(/Edit)

The reflections could be on either side of the image, depending on how the surfaces are inclined to each other. Also, the sideways deflection of the light ray is only a few millimetres and therefore does not cause the multiple images. They are caused by the ray being reflected through slightly different angles.

[green1]

IMG reflect.JPG (114.1 KiB) Viewed 3725 times

I just decided to attempt to see what a reflected image would look like by first erasing the two reflected images, then copy and pasting them and increasing the transparency of the copies. You will see that the copies are still identical with the original. Even the reflected glow on the water is still evident on the first copy. The second copy still shows the cloud glow, although the transparency is set at 90%. In addition, where the reflections overlap they brighten, rather than appearing stacked. This does not come anywhere close to the original photos in question.
I don't believe for a minute that this is a result of reflections on glass. It is a natural result of the sun shining through the atmosphere and reflecting from the sea or polar ice, or refracting through cold polar air.

[parkerjh]

This is likely the case of nature imitating the unnatural. The images must be the result of one of three scenarios. Internal refelections in the camera optics, multi-surface reflections through a window in front of the camera, or multiple plane reflections off of boundary layers of the atmopshere.

Camera inter-lens reflections would appear farther spaced apart than they are, and generally inherit other aberations changing the shapes of the reflections

Multi-surface reflections would create dimmer, duplicate images of the sun, superimposed on the scene brightness levels (additive). In this case the regions where the second images exist would not appear behind the cloud structures. In these images the cloud structure areas are too dark for the scene and reflected images to be additive. The light of the second image is attenuated by the cloud areas in front of it.

Boundary layers in weather systems are often stratified. More often we see them as horizontal layers, but the more rare vertical layers do occur. This image is similar to an image that would be produced if multiple large shear curtains could be hung, with large distances in between shears (the arrangement hung at a slight angle from the oberserving normal). Each shear (or strata) allows part of the light to trasmit in the forward direction and some to reflected backward at an angle. The three images could actually be produced by as few as two boundary layers. Behind the main solar disk one can see the scattering due the light coming through the boundary layer reflecting the second image.

[TheSciencePundit]

@backstagediver

My first objection was to DaveB's use of the word clearly. It's nice that you believe that the clouds are different, but looking at the images up close, it's simply not clear at all that the clouds are different.

You completely missed my point regarding relative size. The distance from the earth to the sun is irrelevant. In my explanation, I granted that the sun images could be the exact same size on the glass itself. But if the glass is tilted, then one of those images will be farther from the camera and hence appear smaller. If you're still having trouble with this, here's an experiment you can do.

1. Draw two identical circles on a piece of paper.
2. Tilt the paper so that one of the circles is closer to you than the other.
3. Take a photograph.
4. Measure the relative sizes of the two circles in the photograph.

Bear in mind, I'm not claiming that this is the actual explanation for the different sizes, I'm merely offering it as one plausible explanation for how a double-paned window reflection can appear smaller (or larger) than the original. It's also quite possible that variations in the separation between the panes could create the effect.

The apparent blurring of the edges can indeed be adequately explained by the diminishing brightness of the reflected images. You should play around with Photoshop or GIMP some time; you'll see what I mean.

[KateMScott]

At first I was convinced that this was a natural occurrence, but when I looked at the photo larger, I know that it is an effect created by a window or camera lens.

The bottom of sun flattens where it hits the horizon, but the 2nd sun (shadow) matches the actual sun. I would like to argue that a sun higher in the sky would not have a flat bottom (or at least have a smaller one). The optical illusion of a flattening sun is created by light refraction in our atmosphere. And if that same atmosphere also theoretically creates the reflected image of the sun, it would not change the way the image is distorted (flattened) as it travels to our eyes - the image would still be refracted based on its angle in the sky (would not be distorted as much when it is further from the horizon). Therefore, because we see the shadow sun is flattened exactly the same way as the real sun, it proves that the reflection is within close proximity to the viewer (from a window pane or camera).

Note: This is just a guess. Please let me know if this really does make sense.

Kate
Los Angeles, CA

[luchio]

Here You can see that reflections from window glass can be placed on every side, all You need is to change angle betwen camera and glass:

ksiazyc.jpg (8.42 KiB) Viewed 3629 times

Here are visible reflections of balcony, which prove that pictures were made by closed window.

balkon.jpg (33.41 KiB) Viewed 3634 times

About the clouds in fron of sun: we canot see which clouds are in front of Sun, because they are overburned by sunlight, so there is no chance to say that in front of reflections are clouds, cause this can be only "shadows" of clouds in front of Sun.

But this is only my opinion.
PS. Sorry for my poor english

Luchio, Poland

[photojhh]

Regardless of the debate about angles of incidence, atmospheric effects, etc., one thing is absolutely certain: There was something with reflections between the photographer and the sun. MadCat-75 made that point by outlining reflections in the wide angle shot. I brightened that part of the photo and was going to post it, but luchio just beat me to it.

Local reflections are quite visible, complete with mirror-image symmetry. The presence of optically active planes (minimum of two) would certainly cause some effect when viewing a bright object like the sun. Several other posts in this discussion show photos that duplicate the effect to the degree that a consensus seems reachable: window glass.

[Howard Cornelsen]

IMHO, all objections to "window glass reflections" have been put to rest.

At a glance, the wide-angle shot shows reflections, probably of curtains as well as balcony.

The sun reflections show clouds. The original sun doesn't because its brightness overcomes the clouds, visually and photographically.

As a photographer, my immediate reaction to the original image was that the cause was clearly reflections. A number of people have brought up other interesting (less mundane) possibilities, but it seems to me that all questions are tending toward resolution in the window-glass reflection theory.

Nonetheless, it's a great set of shots, Noemi! I wish my horizons were that near level.

And the thread has been an enlightening read as well, with a lot of input from various fields of knowledge.

But for me, the "multiple reflections from window glass" supporters have the answer.

--Howard

[owlice]

There is no reflection of the "second sun" on the water. I think this lends support for the window theory.