54
$\begingroup$

The light from the moon is light being reflected from the sun. The sun, in space, is white. But on Earth, when the light is filtered through an atmosphere, the light appears yellow. So then, why is moonlight white through the atmosphere?

$\endgroup$
8
  • 22
    $\begingroup$ Also, I disagree with the premise that the sun appears yellow through the atmosphere. Sometimes it appears yellow, sometimes white, sometimes red; as does the moon. This site has a good explanation. $\endgroup$
    – Jack
    Jun 27, 2018 at 15:22
  • 2
    $\begingroup$ Somewhat related: Why doesn't a full/gibbous moon high in the sky ever seem to look orange? Shouldn't it? $\endgroup$
    – uhoh
    Jun 27, 2018 at 15:28
  • 4
    $\begingroup$ The Sun looks yellow/orange/red as it gets closer to the horizon. The Moon does the same. The Moon looks white when the Moon is high in the sky. The Sun is so intense that you can't look at it when the Sun is high in the sky. $\endgroup$ Jun 27, 2018 at 18:54
  • 3
    $\begingroup$ Full Moon looks quite "yellow" tonight. $\endgroup$
    – ProfRob
    Jun 27, 2018 at 22:19
  • 2
    $\begingroup$ Moon looks more yellow or even orange when it's close to the horizon. This is similar to sun. In the afternoon sunlight is almost white. $\endgroup$
    – rus9384
    Jun 27, 2018 at 22:29

5 Answers 5

64
$\begingroup$

The light from the moon is light being reflected from the sun.

This is at least one reason you should not expect the Moon to have the same color. Sunlight hitting an e.g. blue object would appear predominantly blue and similar. So the color we might expect to see from the Moon is going to be adjusted by the color properties of the Moon's surface.

As it happens lunar regolith is essentially gray in color (i.e. neutral), so it more or less reflects the spectrum it receives.

The sun, in space, is white

No it is not. There are a couple of reasons.

The Sun is an exceptionally bright object (for us) and will saturate our color vision so that it seems like it's white when we look at it directly.

The Sun is actually a yellow star. It's spectrum peaks in the (visual) blue-green range. There are stars that have a red or blue color.

White is a tricky one. White is a perceived color. There is no "white" wavelength in the EM spectrum. White is a balance between the three different color sensors your human vision uses. The Sun itself is not white because the color combination "white" we're balanced for is based not on the light from the Sun in the vacuum of space, but on the light that reaches the surface of the Earth and also what is reflected from the Earth (which is part of the ambient lighting we normally experience).

why is moonlight white through the atmosphere?

It seemed a nice clear sky last night around my neck of the woods and the Moon looked yellow to me. This will vary with your location and atmospheric conditions, but also with your personal perception of color.

$\endgroup$
15
  • 7
    $\begingroup$ -1. Sunlight is about as white as white can be. It is the definition of whiteness, as perceived by the human eye. We don't perceive the Sun itself as white because we can't look at the Sun except when it's very close to the horizon and the color is changed due to excessive Rayleigh scattering (and even then it's unadvisable to do anything but take a glancing look at the Sun). $\endgroup$ Jun 28, 2018 at 4:26
  • 23
    $\begingroup$ @DavidHammen White in terms of human perception is not at all as simple as "sunlight". Color science is rather more complex than you might think. $\endgroup$ Jun 28, 2018 at 6:10
  • 5
    $\begingroup$ Even little kids see that the sun looks yellow and always paint it that way. $\endgroup$
    – RedSonja
    Jun 28, 2018 at 6:27
  • 4
    $\begingroup$ @RedSonja - That's because the only time you can bear to look at the Sun is when it is very close to the horizon. That is when the excessive Rayleigh scattering due to the very low elevation angle makes the Sun appear yellow, or even red. $\endgroup$ Jun 28, 2018 at 10:25
  • 2
    $\begingroup$ You are all correct. Perception will change with what is perceived as the neutral balance. Thus a twitter feud over whether a dress is white and gold or blue and black. Everyone's brain will interpret an image according to previous knowledge about the physical world and try to represent it as a mental image. The sun IS yellow, by definition. It sometimes LOOKS white. But then again, it can look any color to anyone, according to whatever their brain is telling them at the moment. Stop bickering, please. $\endgroup$
    – Stian
    Jun 29, 2018 at 11:57
43
$\begingroup$

Reflected moonlight is actually slightly reddened compared with the incident solar spectrum (Ciocca & Wang 2013). That same light is then transmitted through out atmosphere in exactly the same way as sunlight.

Any phenomenology (which appears disputed) as per the claim in the question is purely down to the nuances of our colour perception.

The reflected moonlight spectrum

enter image description here

Figure 8: Averaged geometrical moon albedos measured by GOME from July 1995, November 1995, and September 1996.

From ESA's GOME moon measurements, including instrument characterisation and moon albedo.

$\endgroup$
4
  • 6
    $\begingroup$ If I read the legend correctly, the two spectra are taken at different angles and thus different amounts of atmosphere to pass through. Seeing as the question includes that as a possible mechanism, perhaps some further explanation in in order? $\endgroup$
    – AaronD
    Jun 27, 2018 at 19:14
  • 6
    $\begingroup$ @AaronD Indeed, the Moon was observed at higher altitude so is less reddened by the atmosphere than for the solar spectrum. Thus the intrinsic difference is (a bit) greater. $\endgroup$
    – ProfRob
    Jun 27, 2018 at 19:42
  • 1
    $\begingroup$ I hope you don't mind, I stumbled across your answer and noticed that the link is paywalled, so I added a 2nd source from answer to Why is this moon red? $\endgroup$
    – uhoh
    Jan 25, 2019 at 8:05
  • $\begingroup$ I feel the need to point out that calling it "purely down to the nuances of our colour perception" is basically dismissing the entire question. Colo(u)r is the nuances of our colour perception. $\endgroup$ Mar 7, 2021 at 8:14
26
$\begingroup$

Surprisingly, the moonlight is actually slightly warmer color than sunlight, as the moon reflectance is higher for longer wavelengths.

Yet, on clear nights, with the full moon high in the sky (as little atmospheric influence as possible) the landscape around us appears blueish, because of the Purkinje effect: at low illumination levels our red color sensitivity decreases (as our vision system gradually switches from daylight (cones) to night time vision (using rod cells)).

Yes, the moon surface is bright enough and its color can be seen correctly (not distorted by the Purkinje effect). So how how does it relate to the question, where we ask about the moonlight color? In my opinion, we tend to perceive the general impression of the color by looking at the surroundings, not just the light source.

That's why the sunlight looks "warm" (more yellow) and the moonlight looks "cold" (more blue - because our color perception is changed by the insufficient light level), even though the real colors are pretty much the same.

In fact, this is can even become a real problem when taking very long exposure photos at night! They look almost like daylight photos, destroying the intended atmosphere of mystery.

Objectively, the photograph is correct, but it is not what we see with our own eyes. Add some blue tint and the night feeling is back.

Example: (note the stars, this is certainly not a daylight photo!) full moon Abrahim Asad - Mathimago fanni Athiri, Fuvahmulah Beach on Full Moon Night LongExposure (CC BY-ND 2.0)

$\endgroup$
5
  • $\begingroup$ The full moon is too bright I think for the Purkinje effect to be significant. $\endgroup$
    – ProfRob
    Jun 27, 2018 at 21:24
  • 3
    $\begingroup$ @RobJeffries exactly, that's why I think the "cold moonlight" impression (distorting our moon color judgement) is caused by the "blue" moonlit surroundings, even though the moon surface itself is bright enough to show its true white color. $\endgroup$
    – szulat
    Jun 27, 2018 at 21:45
  • 2
    $\begingroup$ @RobJeffries From the linked Wikipedia page, "This is why humans become virtually color-blind under low levels of illumination, for instance moonlight.". Note that it's the color of things lit by moonlight, not the surface of the lit moon itself. $\endgroup$
    – JollyJoker
    Jun 29, 2018 at 8:43
  • 2
    $\begingroup$ @JollyJoker Indeed that is what I had misunderstood about this answer. However, the question is clearly about the appearance of the moon, not about the appearance of objects illuminated by the moon. The full moon is bright enough that scotopic vision is not an issue. $\endgroup$
    – ProfRob
    Jun 29, 2018 at 10:19
  • 1
    $\begingroup$ It's still a valuable contribution to this discussion, especially given the abundant conversation about how the sun's color is discerned in its reflection by earthly objects. Thanks for posting this. $\endgroup$
    – lly
    Jul 1, 2018 at 12:32
7
$\begingroup$

I agree with Jack's comment, it's partly misconception and party due to the way rods and cones work.

In "By the light of the silvery Moon: fact and fiction" (from where the illustration and quotes were obtained), by Marco Ciocca and Jing Wang, they explain:

Page 365: "The rods are effectively colour blind. Scotopic vision has a higher sensitivity to light levels than photopic, but is effectively colour blind. We convoluted the sunlight spectrum shown in figure 5 with the photopic response above and the result is shown in figure 8.

Figure 8 Figure 8: Perception of sunlight and moonlight due to our eyes.

During the daytime photopic vision clearly dominates, and the resulting spectral distribution of perceived light, even without considering the lack of blue components of the solar spectrum due to atmospheric scattering, is clearly shifted towards yellow-green (560 nm) because cone sensitivity is biased towards that part of the visible spectrum $^{[20]}. There is little sensitivity below 450 nm, with the peak at 560 nm returning to zero at 700 nm. Therefore sunlight appears strongest around 560 nm (yellow-green, the colour of tennis balls, unsurprisingly), and that is why we think of sunlight as golden.

On the other hand, rods show a much more sensitive response towards shorter wavelengths (rod vision shows a peak around 510 nm), but are unable to detect colour. To show the peak response of the rods we convoluted the Moonspectrum of figure 5 with the rods’ response. The result is also shown in figure 8. If the light was so low that only the rods were detecting it, the Moon would appear white because the rods are colour blind and our convolution would not lead to a colour bias. In reality, however, at the light level of moonlit landscapes human eyes are working in a condition that has been dubbed as mesopic vision, a combination between photopic and scotopic vision in dim lighting $^{[16, 17]}$. Cones are still furnishing information but at a reduced rate, while the rods, with their superior light-level detections, allow low light-level sensitivity. Under these conditions, the rods couple to blue receptors in the cones to show blue light $^{[3]}$. There seems to be be, therefore, physiological reasons why moonlight appears bluish."

Page 366: "Our perception, then, of moonlight as a colder, silvery light with a blue tinge is acombination of two effects: a physical one due to diminishing Rayleigh scattering as the Moon gets higher in the sky and a physiological one (knownas the Purkinje shift $^{[6]}$). The physiological shift is because cones are more sensitive to yellow-green light, whereas the rods respond best to green-blue light $^{[20]}$ and couple back to blue cones $^{[3]}$. This bias of our eyes towards shorter wavelengths in dim light creates the illusion of a colder bluer moonlight, even though moonlight is actually overall redder than sunlight."

Note: Figure 5 on it's own is misleading due to the above explanation and this text in the paper which explains figure 5:

Page 363: "We also compared the Moon at altitude 57° spectrum with sunlight in figure 5, which is normalized the same way as in figure 4. Since direct sunlight will damage the spectrometer, the data were collected with the telescope–spectrometer pointed in a direction at as small an angular distance away from the Sun as prudent. We did so to protect the instrument while at the same time trying to minimize atmospheric effects.

The data presented in figure 5 needs to be compensated and convolved with the spectrum of eyesight under bright and dark conditions, as shown in figure 8.

References:

[3] Khann S M and Pattanaik S N 2004 Modeling blue shift in moonlit scenes by rod cone interaction J. Vision 4 316

[6] Naylor J 2002 Out of the Blue, a 24-hour Skywatcher’s Guide (New York: Cambridge University Press) pp 7–11, 86–87, 195–197

[16] Stockman A and Sharpe L T 2006 Into the twilight zone: the complexities of mesopic vision and luminous efficiency Ophthalmic Physiol. Opt. 26 225–39

[17] Pokorny J and Cao D 2010 Rod and cone contributions to mescopic vision Proc. CIE (Int. Commission on Illumination) Conf. Lighting Quality and Energy Efficiency, (Vienna)

[20] Cornsweet T N 1970 Visual Perception (New York: Academic) pp 145-8

$\endgroup$
3
$\begingroup$

There are lots of good an sophisticated answers here.

One important one I've not seen yet is this: Colour is relative. I don't mean this in a scientific or general way as others have listed (correctly). But simply: the immediate background colour is very important.

Try it with any colour sample: place a dark and then a light colour next to the sample. The sample will appear to change, though of course it has not. The Checker Board Illusion is a black/white and grey version of this. But it also works with colours.

The colour the Moon is right next to will change how we see it. At night, the sky is black or a dark cool shade. So even when it's grey it seems bright at white in contrast. When we see the Moon in the day, with a blue sky, it also looks white, but it's fainter. And we're seeing the moon through the blue atmosphere So that's going to make the background contrast different but also make the light more blue.

But seeing the Moon in the daytime is imagined to be unusual. The Moon is supposed to represent night, and cold, and dark. We are social animals. Our perception is not objective but also social. If we only saw the Moon in the Day we would probably think of it as grey blue. But since we consider the Moon mostly at night and draw it next to black - then it becomes white.

The really weird question that I still haven't figured out is: why is the Sun so yellow? Our perception has evolved white light to match the Sun's output, but still we call the sun yellow or Orange.

Colour is weird.

$\endgroup$
1
  • $\begingroup$ Yes indeed it is! $\endgroup$
    – uhoh
    Feb 28, 2021 at 11:41

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .