Color of a mirror on the Earth

[accdd]

If I place a mirror on the ground facing upward, what color will those who observe it from space see?

 

[berkeman]

What do you think and why? What do you see when you place a mirror on the ground and look down at it from above while you stand over it?

 

[accdd]

During the night, black.
During the day, I don't know whether blue or red.

 

[berkeman]

Why red?

 

[accdd]

Because if I stand outside the atmosphere, the reflected light has to travel a lot farther, but I don't know if that is enough to make the reflection in the mirror turn red.
I expect that depending on the angle it should change the reflected color.

 

[berkeman]

I don't see any red...

https://www.popsci.com/uploads/2019...EZXDEYW6JU.jpg?auto=webp&width=785&height=785

 

[anorlunda]

This photo is from the ISS. A narrow band of sunrise/singer is visible. But that is not a reflection of the ISS's image, so a mirror on the ground would not show red.

So I think a man in orbit should see mostly black in a reflected image from a mirror on the ground.

 

[DaveC426913]

They'll see blue.
The mirror reflects what it sees.
60 miles of atmosphere is not tintless, whether you're looking up from below or down from above.

 

[jbriggs444]

reflected light has to travel a lot farther, but I don't know if that is enough to make the reflection in the mirror turn red.

Distance does not make light turn red.

At least until you get to seriously large distances where cosmological redshift is a factor. Even then, it is not the distance, exactly, that is causing the redshift. Rather, heuristically at least, it is that the distance is increasing over time. The more distance there is, the more effect a percentage increase per unit time will have.

 

[Vanadium 50]

If you took a photograph of the sky and placed it on the ground would you expect it to look like the mirror? If not, why not?

 

[anorlunda]

They'll see blue.
The mirror reflects what it sees.
60 miles of atmosphere is not tintless, whether you're looking up from below or down from above.

Then continents would look blue seen from the ISS. Or blue tinted. That is not seen from space.

 

[accdd]

If you took a photograph of the sky and placed it on the ground would you expect it to look like the mirror? If not, why not?

No.
If I look at the mirror from a given angle I see: the light scattered by the air between me and the mirror and the light scattered by the air placed symmetrically on the opposite side. The light scattered by the latter has to go a long way before it reaches the observer who is outside the atmosphere.
For the same reason that the sky at sunset is red, I expect that the color seen by the mirror may be red, or a color between blue and red. Am I wrong?

 

[accdd]

I don't see any red...

View attachment 316108https://www.popsci.com/uploads/2019...EZXDEYW6JU.jpg?auto=webp&width=785&height=785

Aren't the colors in this photo mainly due to the color of things? In a mirror, the color observed should be almost only that of the light scattered by the atmosphere.

 

[jbriggs444]

For the same reason that the sky at sunset is red, I expect that the color seen by the mirror may be red, or a color between blue and red. Am I wrong?

You are talking about the effect of Rayleigh scattering and atmospheric dust. For a long optical path, blue will be scattered away preferentially while the reds and oranges will suffer less effect. Yes, this occurs at sunrise and sunset when the optical path through the atmosphere is perhaps hundreds of kilometers long.

The optical path vertically through the atmosphere is much shorter than that.

 

[jbriggs444]

Aren't the colors in this photo mainly due to the color of things? In a mirror, the color observed should be almost only that of the light scattered by the atmosphere.

In a mirror, the color observed should be a combination of three things:

1. The light from the original object (somewhere in space) that is not scattered by the atmosphere on the way down, is reflected by the mirror and is not scattered on the way back up.
2. The light from the sun that is randomly scattered toward the observer.
3. The light from the sun that is randomly scattered away from the observer and then reflected.

The one of these that is due to the "color of things" is number 1. That one does not involve scattered light except to the extent that the original object in space was illuminated by Earthlight rather than sunlight.

In a picture of the Earth from space, the color observed should be a combination of:

1. Reflection from direct incident sunlight that was not scattered on the way to the target and not scattered on the way back up to the receiver.
2. Reflection from indirect incident sunlight that was scattered on the way down and not scattered on the way back up to the receiver.
3. Incident sunlight scattered to the receiver.

One can see the effect of scattering on 1 and 2 above by looking at the greenish patches of shallow ocean near the shoreline. Assuming that the picture is undoctored, this would plausibly be due to scattering of everything but blue/green in the shallow water.

 

[DaveC426913]

Then continents would look blue seen from the ISS. Or blue tinted. That is not seen from space.

I am pretty sure that uncorrected images of Earth from space are blue tinted.

Regardless,

During the day, what does the mirror "see" and thus reflect?
The same thing we see when we look up.
Blue sky.
Why would a mirror not reflect the blue sky it sees?

 

[Baluncore]

Blue-tinted light is scattered by the atmosphere into the mirror, arriving only down the very narrow incident path defined by the observer and the mirror orientation.

A similar amount of back-scattered upward-travelling blue-tinted light will be added to the exit ray. But then some of the originally incident blue-tinted light will be lost during passage up the exit ray.

The result will be a mirror with an apparent blue tint.

 

[sophiecentaur]

Thanks @accdd for an interesting question. When I first read this I though "what a daft question" but it has evoked plenty of different ideas and replies so not so daft after all! It got us all going.

I'm assuming that there is no specular reflection of the Sun so, with no atmosphere, the (image in the) mirror would be black (space).

With some atmosphere, Rayleigh scattering produces a fairly uniform indirect field of blue-enhanced sunlight (the blue sky). This link shows the variation with angle of the scattered light and the 'high frequency tilt' (1λ⁴ ) which gives the blue tinge. Note there is still plenty of red wavelength light in it; it is desaturated, despite our subjective and over cooked appreciation of it. Something I had never considered before is that the component of light, scattered in the higher atmosphere could also be subjected to a 'low frequency tilt' because of re-scattering (??).

If the mirror were replaced by a black disc, the observer would see sunlight, scattered from the atmosphere and the 'colour' be much the same as our blue sky. The specific angle / height of the Sun would affect the details but the link shows it as not very directional - definitely a lot of backscatter.

In addition, light scattered from a particular area in the sky will hit the mirror and be reflected towards the observer in space through the atmosphere. Passing along this path (the only path that's relevant ) it will lose some of the blue component.

Conclusion: The space observer will see light from both causes; direct scattering and a modified image of the sky, as seen from Earth. The result will be almost twice the brightness of the sky, looking up but possibly a bit less blue saturated because of the red enhancement of the light reflected by the mirror.

 

[hutchphd]

There is a useful magic number to keep in your quill of usefull numbers. For midspectrum (green ~500nm) light the Rayleigh scattering rate for air at STP is ~1% per km .

 

[sophiecentaur]

There is a useful magic number to keep in your quill of usefull numbers. For midspectrum (green ~500nm) light the Rayleigh scattering rate for air at STP is ~1% per km .

If the atmospheric effects (blue sky) only extend to 20km altitude, the absorption would not be more than say 20%. So the contribution of what you see via the mirror could be perhaps 80% to what the space observer would see from direct scattering of theSun. Using that crude estimate, if there were a nearby reference black on the ground, the result from the mirror would be around 180% relative to the black area. That would probably make the mirror appear a bit but not much brighter than clean snow - noticeable under the right conditions.

 

[Baluncore]

Using that crude estimate, if there were a nearby reference black on the ground, the result from the mirror would be around 180% relative to the black area.

Shadows on satellite images are very black. Shadows are really blue, since they are illuminated by the blue of all the sky, but not the Sun. I suspect the blue is removed from the entire image by a digital haze filter.

There is a ground-truth black reference on the ground. It is the deep well that is hypothesised in the question; "Can you see stars (or a satellite) during the day, from the bottom of a well".

 

[sophiecentaur]

Shadows are really blue,

I think this must be a subjective thing. If you look at the sky and not of the Sun, depending on which precise direction you are looking you will see a 'blue' that is the result of the high frequency tilt. Look at the RGB values on any of your most impressive blue sky photos and there is still a lot of R and G. There's about four times the amount of far blue light scattered as far red but, between those extremes, the levels of other wavelengths follow a just slightly curved line so the integral effect will be to desaturate the 4:1 effect. Then, we never look directly at the Sun and the sky doesn't change colour when we stand in the shadow of a wall.
If you look in an object in half shadow (on Earth) does the darker bit look particularly blue, compared with the bit with Sun on it? I think the answer is no. But our brain is hard at work eliminating any colour difference due to the change of illuminant so that evidence is inconclusive.
Also, colours seen from satellites are affected by the context of the surrounding cloud and the intense 'black' of space (at least half the field of view. Our vision wasn't evolved to deal with that and it's likely that the cameras will have colour balance which could be biassed. Why would the cosmetic / entertainment value of those stunning shots be toned down at all?

 

[Baluncore]

A "haze filter" was used to remove UV that may affect a photographic image.
That filter is built into digital cameras.
Some blue can also be subtracted to sharpen a digital image, in effect by removing blue haze.

 

[sophiecentaur]

Shadows on satellite images are very black.

What shadows? Shadows on objects like the craft are not subject to scatter so they can be very contrasty. The shadows on the Earth would have to be the day / night terminator but the 'shadow of the Earth on the Moon (total eclipse) makes the Moon a dim red colour but very visible.

And most space images - seriously astro or camera snaps - are treated for colour, levels and curves so they are mostly 'unreliable witnesses', I suspect. (Same goes for human eyes)

 

[Baluncore]

What shadows?

The shadows you see in images such as Google Earth, of the Earth, taken from satellites during the day.
The sky looks blue, and that light illuminates whatever is in the shadow of objects on the Earth. If an object in the shadow is white, then it will look blue like the sky.

 

[sophiecentaur]

The shadows you see in images such as Google Earth, of the Earth, taken from satellites during the day.
The sky looks blue, and that light illuminates whatever is in the shadow of objects on the Earth. If an object in the shadow is white, then it will look blue like the sky.

Google Earth is tiffled from beginning to end to make people watch it and to bring out the features they want. Also, the images are compressed so you can watch them with low bandwidth. Entertaining but I have often found my Times World Atlas is better ( loads of hand drawing).
You are right about the spectrum of the illuminant. Our brains always battle with this. Such an unusual scenario is very hard to predict.

 

[DaveE]

"Color of a mirror" sounds like a Zen koan... Just saying...

 

[Drakkith]

I'd argue that a mirror has no color. It is clear, just like the cornea and lens of your eye is or the lens of my refractor telescope. The difference between them being that if I stand in front of my refractor's lens I can see what's behind it but not much of what's in front of it, while a mirror let's me see what's in front of it but not much of what's behind it.

Now if you're asking what the image seen in the mirror is, that's a different story.

 

[sophiecentaur]

Now if you're asking what the image seen in the mirror is, that's a different story.

So, someone sees something but doesn't realize it's in a mirror. What colour is it? The thread title may not be too precise but the Summary explains the question that's really been asked. Sounds like the OP could be a successful AD Man - he grabbed our attention.

 

[sophiecentaur]

It could be worth while looking at the CIE chromaticity chart.

The resulting chromaticity of a mix of colours (points on the chart) is perceived as the centroid (weighted by intensity) of all the input colours. Assuming the spectrum of sunlight lies along the edge of the curve, the centroid lies in the white patch. Scattering will produce weighted values in favour of the blues and greens so the centroid will still be near the white patch - not near the spectral curve - and be perceived as bluish. That light will then be reflected upwards to the observer and this 'bluish colour' will be pulled back towards the white, making it a less saturated blue. The relative path lengths down and up will affect the result.

 

[Curiosity_0]

I think black from the space will make bluish tint very hard to see

 

[Baluncore]

I think black from the space will make bluish tint very hard to see

Then why does the sky look blue against the black of space, when viewed from Earth.

 

[sophiecentaur]

I think black from the space will make bluish tint very hard to see

The Earth's surface will be next to the mirror - not the black of space. The 'background' the observer will be comparing the mirror image with will be surrounding land which could be darker or lighter than the patch of blue sky image that's reflected. That would / could depend on the inclination of the Sun and the proportion of the land that is shadow.
The "black from space" would be at the edge of the observer's vision and I don't think it would actually count in the colour perception of that mirror image.

 

[accdd]

There are many solar power plants using large mirrors on earth. Their color as seen from the satellite is blue.

(source google maps: planta solar 10 Spain)
I have another question: if I have a photo of the sky taken with a smartphone, can I get physically meaningful data from the colors in the picture? Or can the RGB matrix I get cannot be used for anything?

 

[hutchphd]

I have another question: if I have a photo of the sky taken with a smartphone, can I get physically meaningful data from the colors in the picture? Or can the RGB matrix I get cannot be used for anything?

There is useful information. In particular you get three numbers (R,G,B). The exact calibration of these numbers will be far more useful if you have a white (or known uniform gray) image and perhaps a "black" image. Of course if you have a JPEG or other compression scheme it is more complicated.