Gas Giant Colors

[Caelus]

I was wondering what colors different gas giants can be, and whether their temperature has anything to do with it. From our own solar system we can tell that Jupiter-like planets at their distance from the sun run from brown to yellow and white in color, and the ice giants are blue. But if they where closer to the sun, would they be a different color? I was also wondering if gas giants formed from other gases would be different colors, and what the likelyhood of different-chemistry planets existing is. Thanks.

[chornedsnorkack]

There are three basic colours relevant here. White, blue and brown/yellow.

White can come from either dense clouds, or else Rayleigh scattering.

Seriously, Rayleigh scattering.

True, Rayleigh scattering preferentially scatters blue. However, if there is no absorption then red would penetrate deeper into atmosphere, but eventually be scattered back equally completely.

However, warm gas giants do absorb red.

Red is absorbed by water - whether in a massive ocean or in vapour. Thus a massive and otherwise clear atmosphere of hydrogen which is warm enough to keep water as vapour would look blue. Whereas if the water forms dense clouds in upper atmosphere, the clouds would reflect red along with blue. And snow would do likewise, while massive ice just like massive water is blue.

Methane looks much like water. Both have absorption lines in IR at about 3 micrometres, and a series of overtones providing visible absorption into red light.

So planets as cold as Uranus are blue.

What about gas giants colder than Uranus, though? Most methane would rain out... if hydrogen is cooled to 33 K at about 13 atmosphere level, it gets critical. And critical opalescence of hydrogen, or dense hydrogen clouds, might look pretty pure white.

Hydrogen does have collision induced dipole absorption lines, though. Do those collision induced dipole absorption lines have visible overtones strong enough to make liquid hydrogen perceptibly blue?

[eburacum45]

This is interesting;
http://en.wikipedia.org/wiki/Sudarsk...classification The Sudarsky classification includes some of the hotter gas giants, hotter than those with water clouds. See the colours of the gas giants arranged down the right hand side of the page for examples.

[grant hutchison]

Atmospheric colours are quite strongly affected by temperature if you extend your range of interest towards hot jupiters.
In Celestia, we used the calculated reflection spectra from Sudarsky, Burrows & Pinto to produce default colours and texture maps (by Andrew Tribick) for five temperature classes of extrasolar gas giant, as described in the referenced paper.

ETA: eburacum45 got there before me.

Grant Hutchison

[Spaceman Spiff]

First off, by "wondering what colors different gas giants can be", you are presumably asking about their light spectra through the visible range of wavelengths (380-720 nm), right?**

Light interacts primarily with electrons. It is scattered or absorbed in the presence of electrons, which come in a variety of "phases". Here are the most relevant:

1) free within an ionized gas (can absorb in the presence of the electric field of an ion)
2) attached to atoms and ions
3) attached to molecules
4) attached to molecules that have condensed into solid state aerosols and grains, or liquid droplets.

The most important thing to take away from this is that every type of material, in terms of composition and "phase", absorbs and scatters light uniquely.

The prevalence and importance of each of the above four "phases" depend on (a) the elemental composition of the giant planet's atmosphere (defined as that layer responsible for light reflected/emitted by planet), and (b) its equation of state (how pressure changes as a function of density and temperature). The first of these provides the raw materials, and the second arranges them in "phase". Very roughly speaking, one may assign decreasing temperatures (T) to the above 4 "phases" moving down the list 1-->4. Pressure (P) also plays a role, and in general one may place the above phases on a P-T diagram. Physics and Chemistry are at work to determine what kind of "stuff" is present as a function of depth through the giant planet atmosphere. One rule of thumb is that chemistry is much more effective at higher temperatures (to a point) and/or in the presence of moderately energetic light.

Next, before proceeding, go back and read the bold statement, above.

To finish off this overly long post:
Two giant planets of equal bulk compositions will almost certainly appear differently if their P vs. T profiles differ, or if their atmospheric compositions differ (e.g., due to convective mixing from the interior, mixing due to wind currents, heterogeneous settling of heavier matter towards the center over time). Two giant planets of equal bulk compositions, but differing ages will appear differently, since a planet's interior cools over time, affecting P-T relation within the planet as well as its thermally emitted spectrum. The intensity and spectral shape of the light incident from the parent star will affect the P-T diagram, the chemistry and phase of the matter, the thermally emitted spectrum, as well as the distribution of photons available for scattering.

I hope you found some of this helpful. The link mentioned by eburacum45 is a really nice article, and several examples of the effects of composition/state of matter (especially temperature) on the scattered visible light spectrum are provided.



**We will not worry about the neuro-physiological/psychological aspects of color formation in human vision, or the differences between that and multi-broad-band filter images from man-made camera-detector systems.

[JCoyote]

OK it might help if I redefine the original question... are there any colors or color combinations that could NOT happen on a gas giant?

Or are all combos possible?

[Spaceman Spiff]

There are three basic colours relevant here. White, blue and brown/yellow.

White can come from either dense clouds, or else Rayleigh scattering.

Seriously, Rayleigh scattering.

True, Rayleigh scattering preferentially scatters blue. However, if there is no absorption then red would penetrate deeper into atmosphere, but eventually be scattered back equally completely.

This is similar to the color of milk -- which like our atmosphere is weakly absorbing at visible wavelengths, but scatters preferentially light of shorter wavelengths. When the thickness of the medium becomes sufficiently large, the medium scatters everything -- white light incident --> white light scattered.

And even Earth's blue sky is a series of accidents -- despite the 1/(wavelength)⁴ scattering efficiency. Like the bluish, translucent edges of milk where the milk is thin, the Earth's molecular sky is blue because it's just thick enough to scatter light but not too thick. If thick enough, the sky as seen from the ground would brighten and whiten considerably, and even thicker molecular atmospheres would then begin to dim to ruddy red.

Aerosols and droplets add more fun.

[Spaceman Spiff]

OK it might help if I redefine the original question... are there any colors or color combinations that could NOT happen on a gas giant?

Or are all combos possible?

I can't think of any colors/hues in particular that "could NOT happen", although the possibilities do narrow as one constrains the parameters of the planet.

If you're asking whether you could...

"Picture yourself in a boat on a river with tangerine trees and marmalade skies..."; Cellophane flowers of yellow and green towering over our heads..." (Lennon & McCartney)

or something similar, then I guess that sort of thing would be ruled out unless you're taking something extra in your coffee in the morning.

[chornedsnorkack]

I observe that Sudarsky completely ignores gaseous bodies colder than Jupiter, even to mention their omission.

[George]

And even Earth's blue sky is a series of accidents -- despite the 1/(wavelength)⁴ scattering efficiency. Like the bluish, translucent edges of milk where the milk is thin, the Earth's molecular sky is blue because it's just thick enough to scatter light but not too thick. If thick enough, the sky as seen from the ground would brighten and whiten considerably, and even thicker molecular atmospheres would then begin to dim to ruddy red.

Yes. I've read that the blue photons we see have only scattered on average only a little over 1x.

What is odd, however, is that the overhead sky is still blue even during sunset and sunrise. The much greater optical path at this time of day should make the overhead sky more yellow or orange since much of the blues would have scattered away. But our thin upper ozone layer, apparently, is the cause of absorption of the longer wavelengths, which gives us the darkened blue appearance. [Chappuis Effect]

[IsaacKuo]

Okay, so most of the bulk components of atmospheres appear either white or blue. Phosphorus, sulfur, and hydrocarbon compounds may appear red/orange/yellow/brown.

What would provide green?

With blue, red, and green under the same temperature conditions, you could have just about any color.

[George]

What would provide green?

The absorption of red by methane contributes much to the green color of Uranus.

With blue, red, and green under the same temperature conditions, you could have just about any color.

Then add selective scattering to the mix and you get colors you wouldn't expect. The atmosphere of Mars is great example of the color conundrum. It was assumed Mars would have a blue sky, but that quickly changed. Then there is that blue halo around the Sun caused apparently by selective scattering.

What I'd like to know is why Saturn's northern sky sometimes appears blue? I assumed Rayleigh scattering would explain it if the Saturn sky were clear, but pure scattering (eg isotropic atmosphere) should do as Spaceman Spiff states, namely that all the colors would scatter back given enough atmosphere causing a white, sunlight color result.

[Spaceman Spiff]

The absorption of red by methane contributes much to the green color of Uranus.

Then add selective scattering to the mix and you get colors you wouldn't expect. The atmosphere of Mars is great example of the color conundrum. It was assumed Mars would have a blue sky, but that quickly changed. Then there is that blue halo around the Sun caused apparently by selective scattering.

What I'd like to know is why Saturn's northern sky sometimes appears blue? I assumed Rayleigh scattering would explain it if the Saturn sky were clear, but pure scattering (eg isotropic atmosphere) should do as Spaceman Spiff states, namely that all the colors would scatter back given enough atmosphere causing a white, sunlight color result.

You will see the "bluening" effects of molecular scattering along the edges of the opaque atmosphere (just as along the edges of low-fat milk -- and it you can't see it, try mixing some water to the milk until the edges begin to appear bluish).

But you've also got to be careful now with the images taken by Cassini (or Voyager) or by the Hubble Space Telescope. Most of the time, 3 or more broad band images are combined to roughly approximate human color perception, but this is not done exclusively -- depending on the purpose of the observation.

[Spaceman Spiff]

Okay, so most of the bulk components of atmospheres appear either white or blue. Phosphorus, sulfur, and hydrocarbon compounds may appear red/orange/yellow/brown.

I don't think it is quite that simple.

[George]

You will see the "bluening" effects of molecular scattering along the edges of the opaque atmosphere (just as along the edges of low-fat milk -- and it you can't see it, try mixing some water to the milk until the edges begin to appear bluish).

Yes, and there are images of this on the web where the far end of an aquarium that contains milk droplets, opposite the end from the white light source, will appear redish, yet bluish when looking at the sides of the aquarium. Yet I assume looking at the end where the source enters the aquarium it would look more whitish, no reds or blues. [This is one logical point you seem to be making and I am trying to get a grip on it for various circumstatnces, but not for short optical paths, of course.]

But you've also got to be careful now with the images taken by Cassini (or Voyager) or by the Hubble Space Telescope. Most of the time, 3 or more broad band images are combined to roughly approximate human color perception, but this is not done exclusively -- depending on the purpose of the observation.

Yes, but the blue seems to be a near true blue color representation, which Cassini is capable of producing. [Added: They claim Rayleigh Scattering explains it but now I'm not so sure.]

[eburacum45]

I observe that Sudarsky completely ignores gaseous bodies colder than Jupiter, even to mention their omission.

Yes, that puzzled me when I first read the classification. I think Sudarsky et al were just concerned with the characteristics of observed exoplanets, and since exoplanets as cold as Neptune and Uranus are rarely discovered these days, they were not considered in detail.

Your idea of a very cold, whitish gas giant is intrigung. Presumably there would be some Rayleigh scattering as well, even far from the local star.

[Spaceman Spiff]

Yes, and there are images of this on the web where the far end of an aquarium that contains milk droplets, opposite the end from the white light source, will appear redish, yet bluish when looking at the sides of the aquarium. Yet I assume looking at the end where the source enters the aquarium it would look more whitish, no reds or blues. [This is one logical point you seem to be making and I am trying to get a grip on it for various circumstatnces, but not for short optical paths, of course.]

In the end, it all depends on how optically thick the milky water is. If the milky mixture is pretty dilute, I suspect you'd still see a slightly bluish back scattering from the (white spectrum) incident light side of the aquarium. But one can experiment!

Yes, but the blue seems to be a near true blue color representation, which Cassini is capable of producing. [Added: They claim Rayleigh Scattering explains it but now I'm not so sure.]

It's hard to gauge the precise geometry from that image (view is somewhere near the pole, with incident sunlight coming from lower right, Cassini's view -- glancing toward the pole?).

If it is true that we're seeing the effects of mostly molecules via molecular (Rayleigh) scattering, due to the clouds of various liquid droplets having sunk much deeper, then I would have to surmise that there must also be selective molecular extinction present as well. Besides molecular H, Saturn's atmosphere contains methane, ammonia, and probably hydrogen sulfide (H2S). Of these, I know that methane selectively absorbs longer wavelengths.

[Spaceman Spiff]

To those who'd like to dig deeper regarding the physics of giant planets, this paper appeared in Friday's arXiv/astro-ph.

[chornedsnorkack]

Yes, that puzzled me when I first read the classification. I think Sudarsky et al were just concerned with the characteristics of observed exoplanets, and since exoplanets as cold as Neptune and Uranus are rarely discovered these days, they were not considered in detail.

Your idea of a very cold, whitish gas giant is intrigung. Presumably there would be some Rayleigh scattering as well, even far from the local star.

Note that we only just discovered Sedna and Eris which are mostly ice, just like Uranus and Neptune.

They are probably not massive enough to hold hydrogen and helium even at that distance, and their surface tends to get covered with reddish tholins. But note the colour contrast between Pluto, which can hold nitrogen atmosphere at perihelion, and Charon which cannot. What is the atmosphere of Eris like?

Suppose that Solar System were to contain icy bodies the size of Titan, or Mars, or Earth, sufficiently remote not to have been discovered till now and cold enough to hold appreciable amounts of hydrogen.

Could they have, say, white clouds of liquid hydrogen?

[eburacum45]

Could they have, say, white clouds of liquid hydrogen?

A cloud has to be suspended in an atmosphere of some sort. What would a cloud of liquid hydrogen be suspended in?

[IsaacKuo]

If we're talking about the "color" of really cold planets, wouldn't the "color" be dark grey or black? Human eyesight is what perceives color, but eyes are poor at perceiving color in the dark.

Now, if we were talking about an object the size of a basketball, then a human could just shine a flashlight on the object to see its color better. But we're talking about something the size of a planet, so a flashlight isn't going to provide enough light to see the whole thing.

So I think we should consider what the planet would look like under natural light, either by a human in orbit or with human eyes through an optical telescope (which will make it look bigger from a distance, but can't increase its inherent brightness).

If we use a long exposure camera or digital sensors to amplify the planet's brightness, then we're not really talking about the "natural" color of the planet.

[grant hutchison]

A cloud has to be suspended in an atmosphere of some sort. What would a cloud of liquid hydrogen be suspended in?

Helium, I suppose. One could even have hydrogen snow through a helium atmosphere.

Grant Hutchison

[chornedsnorkack]

If we're talking about the "color" of really cold planets, wouldn't the "color" be dark grey or black? Human eyesight is what perceives color, but eyes are poor at perceiving color in the dark.

Now, if we were talking about an object the size of a basketball, then a human could just shine a flashlight on the object to see its color better. But we're talking about something the size of a planet, so a flashlight isn't going to provide enough light to see the whole thing.

So I think we should consider what the planet would look like under natural light, either by a human in orbit or with human eyes through an optical telescope (which will make it look bigger from a distance, but can't increase its inherent brightness).

If we use a long exposure camera or digital sensors to amplify the planet's brightness, then we're not really talking about the "natural" color of the planet.

What are the surface brightnesses where human eye can detect the colours of nebulae and galaxies?

[grant hutchison]

If we're talking about the "color" of really cold planets, wouldn't the "color" be dark grey or black? Human eyesight is what perceives color, but eyes are poor at perceiving color in the dark.

It's not really that dark at the fringes of the solar system. At Pluto's mean distance from the sun, the light level is about equivalent to standard indoor lighting, for instance.
You need to drop the level of direct sunlight at least a millionfold (compared to Earth) before you run into the photopic threshold and start to lose colour discrimination. So you'd still be seeing colours out to beyond 1000AU from the sun.

Grant Hutchison

[chornedsnorkack]

You need to drop the level of direct sunlight at least a millionfold (compared to Earth) before you run into the photopic threshold and start to lose colour discrimination. So you'd still be seeing colours out to beyond 1000AU from the sun.

Indeed. And hydrogen goes critical at 33 K. Which should be encountered perhaps at 60 AU already.

Solid hydrogen below 14 K should begin from perhaps 400 AU.

[Spaceman Spiff]

Indeed. And hydrogen goes critical at 33 K. Which should be encountered perhaps at 60 AU already.

Solid hydrogen below 14 K should begin from perhaps 400 AU.

The critical T of 33 K has a corresponding critical pressure ~ 1.3Mpa or ~1300 atm. It's triple point pressure is 0.071 atm. So it's a little more complicated than you outlined.

[grant hutchison]

The critical T of 33 K has a corresponding critical pressure ~ 1.3Mpa or ~1300 atm.

There's a bit of decimal slippage in there, I think. One atmosphere is around 100 kPa: so 1.3 MPa is about 13 atmospheres.

Grant Hutchison

[chornedsnorkack]

Uranus, at 20 AU, mainly cools (emits longwave radiation) near tropopause, which has temperature of roughly 59 K and is at pressure of about 0,1 bar.

Underneath the tropopause, visible light penetrates to methane clouds, at the level of 1...1,5 bar.

[eburacum45]

Fomalhaut b, at 115 AU, is the coldest planet I can think of off-hand. Fomalhaut is about 18 times as luminous as Sol, so it is not quite as cold as it would be at that distance in our system. But it might be cold enough for hydrogen clouds.

Here's my image of that planet- perhaps I should lighten the colour a little
http://www.orionsarm.com/im_store/fomalhautb.png

[Spaceman Spiff]

There's a bit of decimal slippage in there, I think. One atmosphere is around 100 kPa: so 1.3 MPa is about 13 atmospheres.

Grant Hutchison

Yep, thanks!