*Where* does sunlight scattering occur?

[DaveC426913]

I observed a strange phenomenon the other day which led me to ask this question.

I was in Bahamas on a cruise, where it is very bright. This might have enhanced an otherwise more subtle effect.

(OK, get the drinking jokes out of your system now... )

I was outside, right next to the hull of the ship on a tender. It was about 11AM, so sun was very high in a virtually cloudless sky. Yet I was so close to a 12 story structure that I was in shadow. It was clear and there was no haze.

I looked up at the sky directly above me, so I was looking through sky that was in ship-shadow into cloudless sky. The sky directly above me was so dark that I thought for a good minute that the sky was completely overcast and grey with cloud cover, like it was about to rain very heavily. Anywhere else I looked the sky was bright, saturated blue, but right above the ship it was this medium foreboding grey.

I'm pretty familiar with sky and weather phenomena, so I have a lot to compare to, but this was odd. It took me a few moments of analysis, and some small clouds to drift into the grey before I could convince myself that the sky I was seeing was actually clear as a bell like everywhere else, and not overcast.

Obviously what was happening was that I was looking at the shadow of the ship. But shadows are a lack of light, which means I should have been looking through dark at bright sky beyond. In a clear hazeless sky, I should not be able to see an absence of light.

It leads me to wonder if most of the scattering due to sunlight actually occurs very near the observer, not across the entire expanse of sky. In a sense, the implication of this is that most the brightness and blueness of a sky is actually being generated less than a 50 metres from the observer.

 

[xts]

My ad hoc theory:
Blue light of the sky has an intensity proportional to $\sin^2\phi/\cos\theta$ where $\phi$ is an angle between your sight axis and the line pointing to the Sun and $\theta$ is an angle between sight axis and Zenith.
It is strongest when you look perpendicular to the line to Sun, low above horizon.
As you were in tropics looking vertically and the Sun was almost in zenith, you got the smallest possible amount of blue light of all conditions.

$\sin^2\phi$ - due to scattering dipolar profile

$1/\cos\theta$ - due to amount of air on your sight path

gray (rather than black) may come from scattering on large particles (dust, etc) in lower parts of atmosphere

It leads me to wonder if most of the scattering due to sunlight actually occurs very near the observer, not across the entire expanse of sky. In a sense, the implication of this is that most the brightness and blueness of a sky is actually being generated less than a 50 metres from the observer.

That can't be true. If it would be, you would see blue light from perpendicular beams of white light: e.g. car lamps.ADDED>>>
Or... maybe I was wrong argumenting for $\sin^2\phi$ factor? I must rethink it... Anyway $1/\cos\theta$ seems to be obvious.

 

[DaveC426913]

Just one point: blue light of the sky has an intensity proportional to $sin^2\phi/cos\theta$ where $\phi$ is an angle between your sight axis and the line pointing to the Sun and $\theta$ is an angle between sight axis and Zenith.
It is strongest when you look perpendicular to the line to Sun, low above horizon.
As you were in tropics looking vertically and the Sun was almost in zenith, you got the smallest possible amount of blue light of all conditions.

$sin^2\phi$ - due to scattering dipolar profile

$1/cos\theta$ - due to amount of air on your sight path

Yes, I am well aware of this phenom, as an artist, a photographer, a sailor and a weather enthusiast. You can see in my diagram that I've shown the sky blue near the horizon and whiter higher up.

This was distinctly different. It was darker, not lighter. And grey, not blue or white.

 

[DaveC426913]

That can't be true. If it would be, you would see blue light from perpendicular beams of white light: e.g. car lamps.

I see your point, though I am not sure the two scenarios are equivalent. Many factors between them are different.

 

[xts]

Photographer? So where are the photos? Of course - made with polarized filter...

 

[DaveC426913]

Photographer? So where are the photos? Of course - made with polarized filter...

I considered taking some photos, but there was no way I could. I would have needed a fish-eye lens on my point-n-shoot. The phenom needed the whole sky to be seen, and that's a lot of PhotoShop stitching...

 

[xts]

I must admit - I was totally wrong. Wiki says that blue sky comes from Rayleigh scattering - which intensity is even stronger along the axis than perpendicular. So we have $$\frac{1+\cos^2\phi}{\cos\theta}$$
I am confused - I see no explanation for your phenomenon then... I never saw that, but I've never been sailing on tropical waters...

 

[A.T.]

I don't need to tell you that human visual perception is highly subjective. The eye & brain are correcting the perceived luminosity and tint. Just because a part of the sky looked darker, doesn't mean it was darker. It might have looked darker in contrast to the white(?) ship hull. Just because the sky looked less saturated(blue), doesn't mean it was. The brain normalizes colors.

As for objective effects: A clear sky can be darker right above, because you look trough less atmosphere and thus see less scattered light.

 

[sophiecentaur]

A.T. is probably pretty near the mark. As I was reading this thread the phrase "integrating to grey" kept coming to mind. If the sun is overhead than, as you all agree, there will be a very blue blue on the periphery - bluer than your brain expects. So the not-so-blue bits will look much less saturated (i.e. greyish).

One has to remember that the bluest blue sky and the reddest sunset are not very saturated colours - just stronger than we're used to. They're only bluish and reddish.

 

[TimbreofSilen]

Wouldn't this be the same phenomenon as standing in a well during the day and seeing the stars?

 

[DaveC426913]

Yeah, as I said, I'm pretty perceptive of these kinds of things. While I can't refute the idea that it was a subjective, experiential thing, I plan to keep soliciting for physical explanations.

 

[sophiecentaur]

Wouldn't this be the same phenomenon as standing in a well during the day and seeing the stars?

Is there any evidence that you can actually see stars this way?

 

[A.T.]

Wouldn't this be the same phenomenon as standing in a well during the day and seeing the stars?

Does this work? It doesn't according to this:
http://www.snopes.com/science/well.asp
But this is again about human perception.

To have an objective measure you would need to take two pictures with exactly the same camera settings from the surface and from the well. For the surface picture you could restrict the field of view with a blend right in front of the camera, so that you see the same amount of sky. Or use a strong zoom, such that you don't see the well walls, and the same zoom at the surface.

Then you would compare the brightness of the sky in the two pictures. You could of course use some light sensor directly to measure the intensity.

 

[sophiecentaur]

I might suggest that the reason for some of this effect could be the existence of flare inside your actual eye. They are far from perfect optics - relying on a lot of image processing to get the high standard of imaging that the luckiest of us are blessed with. (Eat yer heart out Legolas) Everyone who uses a camera seriously knows about using a lens hood, even in. circs where you might not think it necessary. Under nearly all conditions, we see the sky, modified by a lot of natural flare and 'bluer' than it really is.

I think Dave's observation was there because of the lens hood, in the shape of the ship and, because this phenomenon is so rarely revealed, it appeared very extreme. The brain is very funny about some things.
I looked for some evidence. Looking through a lot of pictures in my Aperture Library, I see that the bluest of blue skies (not Caribbean) have typical values of R56, G75, B117, which shows how unsaturated a 'really blue sky' blue is. I also looked at 'really red' sunsets, finding typical values of R193, G90, B38. Also, by hunting around in areas of sunset pictures, it was very easy to find pretty much equal values of RGB (grey!).
So, it is possible to find areas of sky, in many pictures, which are actually grey - when there is an appropriate scattering away of some wavelengths and scattering towards of some others.

The only way, Dave, to be sure is to go and have another holiday and take the camera with a fisheye lens and other lenses - making sure to put the exposure on manual.

Boy, what an excuse!

 

[agentredlum]

Have you ever looked up at the sky from inside an airplane window at high altitude? I have and I think it looks more white than blue lending credence to your feeling that scattering occurs closer to the surface. There is another fact indicating your feeling is correct. The atmosphere is thicker near the surface and has more dust particles increasing the likelyhood of scattering.

Here's 2 questions, where you wearing appropriately shaped optics? (Prescription eyeglasses) and what color are your eyes?

 

[mikeph]

Polarised sunglasses + polarisation of diffuse sky radiation could lead to the perception of dark regions.

 

[sophiecentaur]

Aeroplane windows could be responsible for some of the effect. They are not good, optically (they don't need to be) and are made of plastic which will be abraded and become slightly 'frosted' over time.
If there were appreciably less scattering at this height then why would you expect to see "white"? If there were no scattering then you would just see into space without any scattered sunlight . The blue sky actually gets darker and darker (not more saturated blue) as you go up. The proportion of blue vs red light would be much the same, at any given angle.
It's quite hard to look up out of a plane window so I have no idea what it looks like. I must remember to look next time. I think you are mostly looking out horizontally.
As you get higher, the horizon gets further away. Some of the extra scattered blue light that you see has originated at a greater distance. This light then has to travel through a greater distance and, just as with a sunset, there will be a 'reddening' of what you see because the blue is scattered away again. Although not a lot, this will go towards reducing the saturation of the blue that you see - making the horizon appear whiter. Light coming in vertically only passes through, perhaps 100km of air (most of it pretty rarified) Light which is skimming the atmosphere from the horizon can be passing through many hundreds of km of relatively dense air.

 

[xts]

It's quite hard to look up out of a plane window so I have no idea what it looks like. I must remember to look next time.

I am 185cm tall too ;) but airplanes sometimes turn - even at high altitudes they slant by 30deg or more. At 35,000 feet you may spot brightest stars when the Sun is low above horizon.

 

[sophiecentaur]

Turning . . . . good idea.

 

[agentredlum]

Aeroplane windows could be responsible for some of the effect. They are not good, optically (they don't need to be) and are made of plastic which will be abraded and become slightly 'frosted' over time.
If there were appreciably less scattering at this height then why would you expect to see "white"? If there were no scattering then you would just see into space without any scattered sunlight . The blue sky actually gets darker and darker (not more saturated blue) as you go up. The proportion of blue vs red light would be much the same, at any given angle.
It's quite hard to look up out of a plane window so I have no idea what it looks like. I must remember to look next time. I think you are mostly looking out horizontally.
As you get higher, the horizon gets further away. Some of the extra scattered blue light that you see has originated at a greater distance. This light then has to travel through a greater distance and, just as with a sunset, there will be a 'reddening' of what you see because the blue is scattered away again. Although not a lot, this will go towards reducing the saturation of the blue that you see - making the horizon appear whiter. Light coming in vertically only passes through, perhaps 100km of air (most of it pretty rarified) Light which is skimming the atmosphere from the horizon can be passing through many hundreds of km of relatively dense air.

Well, it wasn't white like printer paper. As i remember it, it was more like a dull white, sort of grey, but definitely not blue. If all scattering is equal at high altitude (20000 ft) even if the scattering is diminished, then the overall impression may be of a whitish, greyish hue. Depends on how close to the sun you look.

Of course the higher you go, the less scattering so the sky should get darker and darker.

Notice the color of the sun in this skydiving video, not yellow, is it? Notice the color of the sky in the near vicinity of the sun, not blue is it?



Heres a jump from 30 km. The sun is so white, i have never seen white like this on earth.

http://www.youtube.com/watch?v=IWugcnyqLMk&feature=related

 

[DaveC426913]

Notice the color of the sun in this skydiving video, not yellow, is it? Notice the color of the sky in the near vicinity of the sun, not blue is it?

Heres a jump from 30 km. The sun is so white, i have never seen white like this on earth.

There is no such thing as true colour in video or photography. The only way to assess the colour would be to employ a colour calibrating card in the shot.

 

[agentredlum]

There is no such thing as true colour in video or photography. The only way to assess the colour would be to employ a colour calibrating card in the shot.

If I were to ask an astronaut (Dave Bowman) to describe the color of the sun from space, what would he say?

Are you saying all colors in photographs and videos are not to be trusted? Dave...?

 

[sophiecentaur]

Well, it wasn't white like printer paper. As i remember it, it was more like a dull white, sort of grey, but definitely not blue. If all scattering is equal at high altitude (20000 ft) even if the scattering is diminished, then the overall impression may be of a whitish, greyish hue. Depends on how close to the sun you look.

Of course the higher you go, the less scattering so the sky should get darker and darker.

Notice the color of the sun in this skydiving video, not yellow, is it? Notice the color of the sky in the near vicinity of the sun, not blue is it?



Heres a jump from 30 km. The sun is so white, i have never seen white like this on earth.

http://www.youtube.com/watch?v=IWugcnyqLMk&feature=related

I guess that, amongst all the other things the sky divers were having to think about, they would not have remembered to set the colour balance to anything in particular nor to put a strong neutral density filter in front of the lens for accurate colourimetric measurement of the Sun's image. It would have been Point-and-Shoot and grossly overexposed at times, I'm sure. This, for electronic sensors, would drive them into limiting so all information would be lost. Not real evidence of anything really.

 

[agentredlum]

I guess that, amongst all the other things the sky divers were having to think about, they would not have remembered to set the colour balance to anything in particular nor to put a strong neutral density filter in front of the lens for accurate colourimetric measurement of the Sun's image. It would have been Point-and-Shoot and grossly overexposed at times, I'm sure. This, for electronic sensors, would drive them into limiting so all information would be lost. Not real evidence of anything really.

Allright, how about this evidence?

http://www.universetoday.com/18689/color-of-the-sun/

 

[sophiecentaur]

Evidence of what? It just shows the same effect that is visible in the setting sun - just much less evident because of the few km of atmosphere in between, compared with what you get when the sun is at the horizon.
It says that the sun's temperature is about 6kK - no surprises there.
Those pictures were taken using the precautions I specified earlier! They may have even been 'tiffled' to make them look right on your monitor screen!

 

[agentredlum]

Evidence of what? It just shows the same effect that is visible in the setting sun - just much less evident because of the few km of atmosphere in between, compared with what you get when the sun is at the horizon.
It says that the sun's temperature is about 6kK - no surprises there.

No surprises there? THATS THE REASON IT LOOKS WHITE FROM SPACE! BECAUSE IT'S 6000 Kelvin.

Look at image 7) I this was taken by astronomers working at kitt peak.

http://www.science20.com/solar_fun_of_the_heliochromologist/the_color_of_the_sun_revelation

 

[sophiecentaur]

No surprises there? THATS THE REASON IT LOOKS WHITE FROM SPACE! BECAUSE IT'S 6000 Kelvin.

Look at image 7) I suppose you know better than astronomers working at kitt peak.

http://www.science20.com/solar_fun_of_the_heliochromologist/the_color_of_the_sun_revelation

You seem to be angry about something and I don't know what. Which is "image 7" and what are you getting at?
What has all this to do with the OP and the subtle effects of atmospheric colouring?
I don't see how I am arguing with the "Astronomers at Kitt Peak" but I know that there will be things about which I know more than some of them - the ad hominem argument is pointless.

Your average photo of the Sun just doesn't tell the truth. Fact. That's all I have been saying. What have you introduced into the thread that takes us anywhere?

 

[agentredlum]

You seem to be angry about something and I don't know what. Which is "image 7" and what are you getting at?
What has all this to do with the OP and the subtle effects of atmospheric colouring?
I don't see how I am arguing with the "Astronomers at Kitt Peak" but I know that there will be things about which I know more than some of them - the ad hominem argument is pointless.

Your average photo of the Sun just doesn't tell the truth. Fact. That's all I have been saying. What have you introduced into the thread that takes us anywhere?

I'm not angry, from your previous post you seemed angry to me. What is the point of rejecting evidence? Of course you can say there is something else going on but all you have offered is speculation about filters being used.

What does this have to do with the original post? Everything! I'm sorry you don't see that. Color perception is the key here and color temperature scales are important to the perception of color. Holding on to the colloquial idea that the sun is yellow is not a good idea when contrary evidence is presented.

Just scroll down the page and you will find 7)

I edited post # 26 to eliminate personal attack perception. That was not my intent, so i fixed it.

 

[DaveC426913]

Are you saying all colors in photographs and videos are not to be trusted? Dave...?

That is exactly what I am saying, yes.

They auto correct colour to make for a good picture, based on what camera manufacturers deem "a good picture" under average circumstances for average users. (Having studied it in college and 10+ years in the photo industry, I could go on at length about white balances and neutral greys. De-correcting for auto-colour correction was a large part of my work.)

My little point-n-shoot has at least six settings to correct for colour temp. of lighting. It's default state is auto-correct. More sophisticated cameras have more sophisticated algorithms for correcting.

Unfortunately, what you want is exactly the opposite. You want a system that does no correction at all.

Without calibration, cameras cannot be used to compare colours like you are trying to do.

 

[agentredlum]

That is exactly what I am saying, yes.

They auto correct colour to make for a good picture, based on what camera manufacturers deem "a good picture" under average circumstances for average users.

My little point-n-shoot has at least six settings to correct for colour temp. of lighting. It's default state is auto-correct. More sophisticated cameras have have sophisticated algorithms for correcting.

Unfortunately, what you want is exactly the opposite. You want a system that does no correction at all.

Without calibration, cameras cannot be used to compare colours like you are trying to do.

Have you ever made a pinhole camera?
I have, and i have looked through it.
Do your own eyes have correcting algorithms?

I am not a professional photographer. However, i have taken many pictures, i never mistrusted colors in photographs before. I have also used video recording machines and never noticed discrepansies between colors so i must 'chew' on your statement that colors are misrepresented, Dave.

By the way, i was on your side on this one...

 

[agentredlum]

I guess the only way to settle this is to ask an astronaut. Are there any astronauts out there? What color is the sun from space?

 

[sophiecentaur]

That is exactly what I am saying, yes.

They auto correct colour to make for a good picture, based on what camera manufacturers deem "a good picture" under average circumstances for average users. (Having studied it in college and 10+ years in the photo industry, I could go on at length about white balances and neutral greys. De-correcting for auto-colour correction was a large part of my work.)

My little point-n-shoot has at least six settings to correct for colour temp. of lighting. It's default state is auto-correct. More sophisticated cameras have more sophisticated algorithms for correcting.

Unfortunately, what you want is exactly the opposite. You want a system that does no correction at all.

Without calibration, cameras cannot be used to compare colours like you are trying to do.

Your next range of cameras can shoot in RAW, which produces larger files but allows you to get the colour balance better by picking on a portion of the picture, or in the batch, with a reliable grey.

But we are verging on the subject of what colour 'actually is'. I have to insist it's totally in the mind of the viewer but that we can measure spectrum and equivalent 'black body' temperature very accurately. Colourimetry is based solely on a consensus of subjective opinions about colour matching of different combinations of differently produced primaries.

I was thinking that the thread was simply about the relative amounts of scattering of light from different directions and with the sun in different positions. Even though the measurements are a bit flawed, the results from my camera do show what I am getting at- and that is that there are places in the sky where you might expect R, G and B signals to be much closer to equal than in other places and that nowhere will the colours you see be very pure.
Get those tickets booked and tell us about it.

 

[sophiecentaur]

Have you ever made a pinhole camera?
I have, and i have looked through it.
Do your own eyes have correcting algorithms?

Yes they do. They are constantly correcting for the effects of the lighting on the colours of objects. If they didn't, you would see the same object at midday and at sunset and think it was a different object, because it would appear to be to different colours. The agenda they are following is not one of scientific measurement but of making the best sense of what they see of the world around.
Why should a pinhole camera make any difference to what colours you see, compared with just looking directly at a scene?

btw I saw that picture of the sun ("7)") - not sure what it was supposed to prove, though. The three 'representative' coloured objects were obviously there for some sort of reference. However, because the chromaticity values for those references weren't quoted (or even the values for the sun's surface, then the picture is no accurate evidence of anything.

The fact that the atmosphere makes a difference to the spectrum of the sun's light is very obvious, no?

 

[DaveC426913]

Have you ever made a pinhole camera?
I have, and i have looked through it.
Do your own eyes have correcting algorithms?

As sophie pointed out, absolutely. (Your brain that is.) In fact, our personal perception is far more heavy-handed at auto-correction than cameras.

Look around you right now. What lighting condition are you in? Tungsten? Fluorescent? Daylight? Did you actually have to think about it? Regardless of what it is, you will see it as white.

A camera (even with its auto-correcting feature) sees the difference between daylight and fluorescent so powerfully that you'll think your pix are ruined.

 

[sophiecentaur]

As sophie pointed out, absolutely. (Your brain that is.) In fact, our personal perception is far more heavy-handed at auto-correction than cameras.

Look around you right now. What lighting condition are you in? Tungsten? Fluorescent? Daylight? Did you actually have to think about it? Regardless of what it is, you will see it as white.

A camera (even with its auto-correcting feature) sees the difference between daylight and fluorescent so powerfully that you'll think your pix are ruined.

One man's "heavy handed" is another man's survival fitness. Homo sapiens, way back, was far more interested in recognising, consistently, the reflected colours of meat, mates and foliage than in assessing the colour of the Sun. We still have to take the jumper outside into the street to see just how near it matches the socks, though, when we're in Marks'.