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

[Fil]

In my view, the multiplication of the number of Suns occurs between the camera focal plane and the atmosphere where the Sun gets deformed or where the clouds are. The reason for this assumption is that the extra images of the Sun have this same shape/image. So, unless there is something with a discontinuous distribution of multiple refraction indexes between the lens and the place where the sun appears to touch the sea, I think this is a reflection issue in the lens system used in front of the sensor.
Now obviously I am not saying the photographer lied when he said he saw multiple sunsets! What the photographer says is subject to his interpretation of what he saw, and with such a bright object, too much light can play tricks to the eye.. For instance, some people can have a nervous eye movement that stops for very few hundreds of a second and moves to a different position. If an eye looks at a bright sunset and a residual image is formed, then moves slightly to the side and another residual images forms, this can guive the sensation of multiple "overexposed images" in the eye, giving the impression of multiple objects. Its like watching the new car with LED brake lights at night, if the eye moves rapidly we get the impression that there light trail made by the braking car is not continuous, as it is in fact blinking at about 50Hz or something similar.. At night, you can only observe this high frequency by moving your eye and not noticing a continuous "light trail".
But at least the effect seen in the pictures, I attribute it to reflections internal to the system formed by camera objective and digital sensor. (digital sensors tend to be more reflective than emulsion film, and some cameras may not have an anti-reflective coating on the rear lens).

If one were able to analyze an image and separate the "extra reflexes" from the original image, Internal reflexes actually could be used in a productive way to made an High Dynamic Range (HDR) image, by combining each reflected image with the "0-order" (original) image.

[apodman]

The distant stratocumulus cloud decks in the picture act as beam splitters and double refractors.

Mystery science.

[polymath]

Confirmation bias and ad hominem ridicule.

[danamongden]

Well, I don't know for sure about this triple sunrise, but I did once see a double sunrise with my naked eye and not through a window. I was outside, ran back in to get my camera and took a photograph, which I am attempting to upload/include here:



This is a cropped version at full pixel resolution, but I do have a larger image showing the context (http://www.danamongden.net/archives/photos/IMG_0195.jpg). It is over land, so this is not an ocean reflection. There is quite a bit of cloud cover, so the humidity above and below the clouds are likely quite different. My take on it was that the humid air in and below the clouds caused a refraction to give me the second (lower) sun image.

Again, I stress that this was visible to the naked eye and was not viewed through glass. I was outside.

dan@zurg.net

[geckzilla]

Polymath, one of the main reasons you support the atmospheric cause of the sun ghosts seems to be the initial statements by RJN about the photographer not trying to fool us. You seem to be basing a lot of your argument on the integrity of the individuals who created and then chose the image to be an APOD. Does it seem to you that the mundane explanation of a double glazed window would diminish their reputation? Did you read my post about my realization that the Diacons are not actually residents of that apartment, but that it was most likely taken while they were in a hotel room, on vacation, in an unfamiliar place? People make mistakes. Especially when something can be romanticized into something fantastic, our perception can be occluded by emotions.

[Chris Peterson]

To be clear, I don't think anybody is arguing that false sunrises, semi-specular solar reflections from clouds, and other such phenomena aren't real. These things are well understood, and for the most part well documented. They also have nothing at all in common with the APOD in question.

[apodman]

Note that in danamongden's photo (1) unlike the images in the APOD, the mirage caused by atmospheric reflection is inverted, and (2) unlike the images in the APOD, the displacement of the mirage from the original is vertical from a reflection off a horizontal layer.

---

Confirmation bias and ad hominem ridicule.

Bias? I entered this discussion with no opinion. I have been presented with evidence and logic on one hand and word salad descriptions of nonexistent phenomena on the other. Which side would any rational person choose?

Ridicule? I simply called one of your statements what it was. Much of the rest of what you say also deserves to be refuted and dismissed, but nobody has the time to debunk every baseless idea in this topic.

The second sun in Danamongden's photo is caused by internal reflection inside the camera, and is a common phenomenon. The reflection is exactly 180 degrees around the mid point of the frame.

Wrong. Danamongden says it was visible to the naked eye, such phenomena are well known and real, and there's no reason not to believe it. (See, polymath, this guy is arguing against an atmospheric explanation and I say he's wrong, and you argue in favor of a different supposed atmospheric phenomenon and I say you're wrong; where do you think my bias is now?)

[Eamon Shute]

The second sun in Danamongden's photo is caused by internal reflection inside the camera, and is a common phenomenon. The reflection is exactly 180 degrees around the mid point of the frame.

[polymath]

My fifth grade teacher sent me to sit on a stool in the classroom corner and wear a conical cap for asking in geography class if the continents moved. I spent a class-week there ignoring the ridicule of the teacher and the classroom bullies before the teacher relented. In eighth grade geography, the teacher, a different one at a different school in a different district, taught continental drift theory.

I've seen a double sunrise, at sea plowing east'ard through a gentle swell, with my naked eyes. The fishing boat captain saw it too. When a group of fishing boats met up for a midday gam while the nets soaked, other fishermen said they'd seen it too and had seen the phenomena before. The atmospheric conditions that day were similar. I assumed the stratocumulus clouds were above the Gulf Stream some fifty or more miles distant where the surface air temperature was about 80 F. Where I was it was in the 40s.

I won't be shifted off my position, not because I'm hardheaded, but because I conclude the available evidence favors a majestic atmospheric effect. But also because I accept that the cosmos is a fascinating and often currently unexplainable place. I celebrate the majestic beauty depicted in the picture.

[kamoses]

The second sun in Danamongden's photo is caused by internal reflection inside the camera, and is a common phenomenon. The reflection is exactly 180 degrees around the mid point of the frame.

Here is the center of the image.

Clouds seem to be very similar to the APOD pic. Though it is a more vertical version, there is a bit of horizontal displacement, showing that it is possible to have a refraction translate in the horizontal direction. Flipped...well, could it not be possible the APOd suns are flipped? I honestly can't really tell whats goin on with those clouds. They're all pretty flat and lack the detail to fully tell.

[apodman]

My fifth grade teacher sent me to sit on a stool in the classroom corner and wear a conical cap for asking in geography class if the continents moved. I spent a class-week there ignoring the ridicule of the teacher and the classroom bullies before the teacher relented. In eighth grade geography, the teacher, a different one at a different school in a different district, taught continental drift theory.

So you use the fact that you were right about one thing to argue you will always be right about everything. I don't get it.

I won't be shifted off my position

The very essence of bias.

[polymath]

Attacking me doesn't disprove my viewpoint. The logical fallacy of an ad hominem attack is it shifts discussion off topic and ignores the assertion under debate by attempting to discredit the messenger, and by faint implication discredit the messenger's assertion.

[Eamon Shute]

Here is the center of the image.

The original image has been cropped at the bottom. You can tell because the height is less than two-thirds of the width. If you measure one-third of the width of the frame (=one half of the original height) from the middle of the top edge you will get a point mid-way between the two suns (as far as it is possible to tell with such fuzzy images).

[dockwatcher]

Sorry for diverting from the glass in front argument. Since it gets down to if this picture is doctored or not all that is needed is to provide a viable explination (or two) that would put the burden of proof back on the photographer (Who should have taken dozens of pictures from difrent locations).

I couldn't find the footer that discribed the camera used but if it is the canon s51s then it may be important to note it has a thru the lens view finder in aditionto the lcd. since it takes stills and movies simultanious I would think that would add a prism in the light path.
Does anyone have the manual. It is also suposed to have a chapter on silly effects.

[ricardelico]

What do atmosphereists have to say about the reflections evidencing a window pane in image 9802? Does anyone dare seriously state those are not windowpane reflections? Do I hear Hey, those are camera-lens reflections! or extremly rare but nonetheless possible atmospheric Gdansk mirages occurring right at the balcony!

-- Windowpaneist

[Hoyos]

Gentleman (and ladies, too),
lets not leave our ego here.
Poor APOD reader that has to go through all posts.

There is no point in disputing that a double or triple sunrise may exist. As we know, the atmosphere has lots to teach us still.
So, lets focus on this particular phenomena and try to explain that particular photo. Wasn't that the reason why RJN opened this forum?

So, the point I made well back in page 5 was that the atmosphere-made reflection or refraction is very unlikely because the "Etruscan vase" patter of the primary sun is reproduced in the others.
The Etruscan vase is caused by the refraction of the light in the warmer layers over the nearby sea.
If the other suns are caused by the atmosphere, they shouldn't see the same pattern, but a different one (see the various patterns in http://www.atoptics.co.uk/atoptics/sunmir2.htm).

In order to capture the same pattern, the reflection of the sun has to occur between the body of water and the viewer.
Having said that, I challenge the atmosphere-theory supporters to refute this argument.

Cordially,
Hoyos

[Chris Peterson]

Sorry for diverting from the glass in front argument. Since it gets down to if this picture is doctored or not all that is needed is to provide a viable explination (or two) that would put the burden of proof back on the photographer (Who should have taken dozens of pictures from difrent locations).

There is not the slightest reason to think the image is doctored. There is not the slightest reason to think that the photographer is trying to fool anybody. There is nothing to suggest that these few images are anything other than quickly snapped vacation images of an odd thing that somebody saw out their window. Most people would not take it farther than that- taking dozens of shots from different angles. And there is no burden of proof on the photographer at all. This is just an odd image that somebody submitted and asked for an explanation.

I couldn't find the footer that discribed the camera used but if it is the canon s51s then it may be important to note it has a thru the lens view finder in aditionto the lcd. since it takes stills and movies simultanious I would think that would add a prism in the light path.

The information is in the intact EXIF file headers for each image. Most any modern image viewer or editor should give you that information when you open an image. You can read about the camera at dpreview.com. I don't think we need to be concerned with the camera optics, since there is no reason to think the effect in the images is related to them. The photographer took the pictures in the first place because she saw the effect with her naked eyes. Also, you can see the typical internal reflections of the lens system in the images, and they look about as expected.

[Dave Shaffer]

Folks,

Plain and simple: It’s the window glass. I hope/think the images discussed below will set to rest all the objections to this explanation.

Anybody who thinks otherwise does not understand how the atmosphere works.

The atmosphere is stratified horizontally, with small vertical systematic variations in the index of refraction. Hence, displacement due to the atmosphere will be VERTICAL. The exceedingly long path traversed by the setting/rising sun results in a maximum displacement on the order of a degree (twice the size of the sun). As noted by some other commentators, this refraction also elongates and smears our view of the sun (as can be seen in the image in question at the bottom of the sun).

Any horizontal variations in the index of refraction of the atmosphere will be way too small to cause more than very tiny lateral shifts in the apparent position of the sun - certainly nothing like degrees or more as seen here (bigger than the common, well-known vertical effect!). Additionally, as noted by others (and as seen in the solar images here and elsewhere), such shifts would tend to smear out the solar image. The crisp lateral displacements noted here would not occur without both a large change in the index of refraction and a discrete boundary in the atmosphere - larger than could possibly occur.

Optical astronomers worry all the time about the vertical dispersion of the atmosphere - it affects the spectra of stars taken at low elevation (where “low” here means tens of degrees of elevation). They NEVER worry about lateral (horizontal) dispersion - IT JUST DOESN’T HAPPEN!

Contrary to some comments, there will be NO effects due to the latitude of the observation. As noted above, the atmosphere is ->locally<- horizontally stratified.

Many of the objections to the dual-pane window theory seem to involve the sided-ness of the images, the curvature (mis-alignment) of the images, and the spacing of the images. The argument being that the spacing between the glass would have to be highly variable in order for these effects to be due to the glass. (The other primary objection seems to be the appearance of the clouds in front of the sun. This has been settled by the Photoshop overlays which show identical clouds, with the realization that the “primary” image is overexposed/saturated.)

Well, it turns out, even fine (?!) American glass is pretty crappy in this regard.

As the full moon was rising Wednesday evening in Flagstaff, Arizona, I took some pictures through one of our double-pane windows. All the effects that “can’t possibly be due to the glass” are, in fact, quite obviously seen here.

These four pictures were taken within a few seconds of each other, with my Pentax K20D digital SLR camera. The exposures are 1/13 or 1/10 of a second, at f/2.8. Autofocus, and image stabilization were turned on. I moved the camera 2 or 3 inches laterally between exposures, going from left to right. (Unfortunately, there is a window screen between the window and my camera. The screen causes the up-down / left-right spreading of the lunar image - rather like a very low dispersion diffraction grating. The screen, however, IS NOT the cause of the multiple lunar images.) The exposures have been highly stretched in Photoshop to bring out the third image of the moon. The overexposed lunar image is the brightest thing in the frames. That the overexposed moon is always in the same sky position can be checked by comparing it to the faint house lights seen below the moon in each image. Note, also, the frame of the tilted window and how it moves across the field of view as I move my camera. (The vertical striations in some images are an artifact of the camera image processing (I’ve seen this before for this camera for longish exposures)). The window was angled about 45 degrees (i.e. swung open) with respect to the line of sight to the moon.

In frame 1 (http://astrodave.us/frame1.jpg) with my camera about as far to the left as I could hold it and still see the lunar images, the reflected images of the moon are to the LEFT of the primary image. The faint third image is almost obscured by the window frame. The images are not equally spaced, and they are not in a line.

In frame 2 (http://astrodave.us/frame2.jpg), I have moved my camera a few inches to the right. The lunar images are not too far from being equally spaced, and the third (left-most) image has moved in relatively closer to the second image.

In frame 3 (http://astrodave.us/frame3.jpg). I have moved the camera another few inches to the right. The ghost images have shifted to the RIGHT of the primary lunar image, indicating a significant change in the (lack of) parallelism between the two panes of glass which comprise this particular double-pane window. (This is a high-quality, modern (about 6 years old) american-made window. I have noted similar effects in other windows in our house, as well as other windows here in Flagstaff. It seems that exceedingly small variations in the relative orientation of the two panes of glass can have VERY large effects on the reflected images - as some of the other commentators here have been trying to say. It is time for the naysayers to give up on this point!) Note that here, the third image is very far away from the second image. Again, very minor glass misalignments can cause very large variations in the relative location and orientation of the reflected images. Note, too, that the second image is above the line between the primary and third image. Lack of linear alignment is another artifact of how the glass panes are oriented.

Finally, in frame 4 (http://astrodave.us/frame4.jpg), with the camera moved yet another couple of inches to the right, it seems that the third image has moved so far from the primary image that it is either “behind” the right-hand window frame or completely out of the picture.

These pictures should put to rest all the objections about multiple reflections due to multi-pane glass. Variations in both the thickness of the glass and the relative alignment of the panes of glass can have large effects on the relative displacement and orientation of the multiple reflected images which occur as light bounces around between the panes.

You can all observe this for yourselves. The sun is generally too bright for such observations - except at sunrise and sunset! However, as shown here, the moon works well. I could easily see a double moon image with my unaided eye and it was obvious that the relative location of the second image moved wildly as I scanned my eye back and forth, and up and down with respect to my double-paned windows.

THERE IS NO DOUBT THAT THIS IS A WINDOW EFFECT. Those who think otherwise need to show us an observational counterexample, and produce a scientifically reasonable theory - no bogus appeals to never-previously-observed horizontal variations in the index of refraction, birefringence, over-the-horizon ice crystals, polar temperature gradients, etc. As shown by me and others, it is trivial to produce double-pane window glass examples which essentially duplicate all the features of the triple sun picture.

David B. Shaffer (Ph.D. Astronomy (Caltech) and semi-pro photographer)

PS This is my first post to this forum. I hope all the formatting and image URLs work OK. I felt compelled to respond because of all the atmospheric refraction nonsense being tossed around.

[Chris Peterson]

The distant stratocumulus cloud decks in the picture act as beam splitters and double refractors.

Mystery science.

Confirmation bias and ad hominem ridicule.

There is no ad hominem attack here. There is a minor error in calling it "mystery science", since it isn't science at all. You made an assertion that has no basis in theory or observation. You have offered no evidence that clouds can act as beam splitters or as "double refractors" (a term that isn't even well defined). Throwing out fancy sounding words without support, and which don't hold up to any kind of examination, has a name: pseudoscience, and it isn't ad hominem to identify pseudoscientific arguments for what they are.

[polymath]

There's a hundred kilometers of waterway between the Gdansk bay shores below the solar images. My examination indicates to me that the ordinary solar image has a vase neck and a vase foot. The first extraordinary mirage has a foot but no neck. No apparent foot or neck on the second extraordinary mirage. I also see that the two vases' foots are as likely to be caused by cloud dispersion as by water surface reflections.

[polymath]

I suppose then that binoculars are based on pseudoscience, that six prisms arranged in tandem with a pinpoint polarized light source directed into them projecting an ordinary image and two extraordinary mirages on a viewing plane are pseudoscientific play. That there's only pseudoscientific evidence in false sunrise, green flash, rainbows and raindogs for prismatic atmospheric lensing effects.

Accusations of pseudoscientific application is itself symptomatic of confirmation bias and an ad hominem attack.

No, optical physics and meteorology allows for the possibility of dramatic atmospheric effects. Optical physics is itself at root behind the window pane and mirror reflection theories, which admirably demonstrate localized birefringence principles. I just don't believe they're meaningfully contributing to the fabulous picture's origins.

[Chris Peterson]

I suppose then that binoculars are based on pseudoscience, that six prisms arranged in tandem with a pinpoint polarized light source directed into them projecting an ordinary image and two extraordinary mirages on a viewing plane are pseudoscientific play. That there's only pseudoscientific evidence in false sunrise, green flash, rainbows and raindogs for prismatic atmospheric lensing effects.

We aren't talking about binoculars. In this case, we have no mechanism capable of acting as a prism. We have no polarized light source. We have no relationship to false sunrises, green flashes, rainbows, or sundogs (none of which are particularly well described by "prismatic atmospheric lensing effects", and none of which except for a false sunrise even produce what could reasonably be called an "image" at all).

[bkinmonth]

Has anyone noticed that by using red/blue 3D glasses and tilting your head to the right, the 3 suns converge and form an image of the Virgin Mary?

Just wondering...

Oh yeah, almost forgot:

[Chris Peterson]

Problems with the multiple glass pane explanation:
1) The differential angle between the two window panes is large (6+ mm over 1 meter) relative to typical glass pane separation distance.

That is one explanation, but not the best one. The angle between a pair of internally reflected images is 1/3 the angle between the reflecting plates. The largest angle seen in the images is about 0.75°, so you could make a case for a 4mm spacing difference over a 1m window. But the images themselves clearly show different spacings, which are reasonably explained by the photographer moving a little between them, resulting in views through different parts of the window. It would be more reasonable to consider something like a 100mm section of window, which only needs a 0.4mm spacing variation. That is entirely within the flatness variation of ordinary float glass. It is far more likely that we have slightly wavy glass than that we have a continuous wedge over the entire width.

2) The differential angle varies and in the latest pictures (over a time duration of about 40 seconds) is reduced by at least 50% as the photos show the suns merging.

That is an argument for reflections within a window, since slightly different viewing positions through typical windows are documented to produce exactly this effect.

3) If the differential angle varies this much over time, what causes this rapid change?

Same comments as previously.

4) If the differential angle variation is a variation in different parts of the window, why didn't the observer notice this variation as the camera was moved?

Perhaps she did. Or perhaps the displacement of well under a degree was simply not observed over this short time. There is no indication that the photographer had made any special study of odd atmospheric effects. My impression is a vacationer looking out the window and seeing something different, and getting off a few shots.

5) The clouds appear to be between the observer and the multiple suns. (I put this last - my personal impression was that the clouds lie between observer and suns, but this is a close call.)

How else could they appear? As evidence for a foreground reflection, we have the simple observation that all three suns show substantially the same cloud pattern. The two ghost suns are partially transparent, so they also allow the clouds behind them to be seen. All this is exactly as expected from foreground reflections.

[apodman]

Chris, I use "mystery science" in the sense used by MST3K to mean non-science dressed up in scientific word salad as in bad old science fiction movies. But you know that.

Polymath, if you know about ad hominem you also know about the straw man that you are guilty of attacking.