The softly glowing night sky

By Phil Plait | June 11, 2012 6:57 am

There’s a lot more going on over your head than you know.

For example, the atmosphere of the Earth thins out gradually the higher you go, and when you get to about 100 kilometers (60 miles) up, different physical processes become important. One of them is called chemiluminescence — light produced by chemical processes. This can make the upper atmosphere glow in different colors. It’s faint, and best seen from space… where we conveniently keep several astronauts. Neuroscientist and amateur video maker Alex Rivest has collected pictures of this airglow taken by astronauts and made this eerie and beautiful time lapse video:

Alex took the original astronaut pictures and enhanced them somewhat to bring out the faint airglow. You can see it in lots of pictures taken from the space station, and I’ve commented on it many times. One thing I’ve been meaning to do, though, is find out what the physical process is that’s causing the air to glow, and why it creates different colors — you can clearly see green, yellow, and red glow in many of the pictures!

Alex comes to the rescue on that as well. On his blog, he discusses how he made the video and why the air glows (based on a somewhat terser explanation at the Atmospheric Optics website).

The way this works is simple in general, though complicated in detail — much like everything else in the Universe! Basically, during the day, in the upper atmosphere ultraviolet light from the Sun pumps energy into oxygen molecules (called O2; two oxygen atoms bound together — this is the stuff we breathe). This energy splits the molecules apart into individual atoms, and these atoms have a little bit of extra energy — we say these atoms are in an excited state. Like a jittery person who’s had too much coffee, they want to give off this energy. They can do this in a couple of ways: they can emit light, or they can bump into other atoms and molecules and react chemically with them.

If you have an excited oxygen atom sitting in space all by its lonesome, it can either dump that energy by emitting green light or red light. Usually, it’ll emit green light in less than a second after becoming excited, and it’ll emit red light on much longer timescales, like minutes. This is important, so bear with me.

At a height below about 95 km, the atmosphere is thick enough that collisions between atoms happens all the time. In fact, an excited oxygen atom doesn’t have to wait very long (usually microseconds) before another atom or molecule bumps it. If collisions happen faster, on average, than about once every 0.1 seconds, then an oxygen atom doesn’t have enough time to emit green light before getting smacked by another atom or molecule. When that happens, the other atom can steal its energy, and no green light is emitted. So below that height we don’t see any green emission.

At heights of 95 – 100 km or so, collisions happen less frequently, giving the oxygen atom time to blow out a green photon (a particle of light). So at that height we do see the green glow. This layer is thin, like the shell of a bubble, and we see it as an arc due to limb brightening (which you can read about here if you want details). In the picture above, you can see it as a very thin green arc above the diffuse yellow glow (which I’ll get to; hang tight). Normally it wouldn’t be very bright, but looking along the edge of the shell is like looking through a very long slab that stretches for hundreds of kilometers. The light builds up, making it bright enough to see.

Higher up, above 100 km, the oxygen atoms are much farther apart because the density is lower. The odds of two of them colliding are a lot lower, so the time between collisions can be pretty long, long enough to give the oxygen atoms time to emit red photons. That’s why we see that red glow higher up, where the air is ethereally thin.

As I said above, the oxygen atoms can also smack into other molecules and react chemically. When there’s hydrogen and nitrogen around, one of those chemical end products is what’s called a hydroxyl radical — an oxygen and hydrogen atom bound together (designated OH). These radicals can vibrate, like weights attached to either end of spring, and emit red light in the process as well. That also contributes to the red sky glow at great heights.

I’ve pointed out this red smear in various astronaut photos before (like here and here), and wondered what it was. Now I know!

There’s more going on, too. Below that green line (at roughly 50 – 65 km high) is a somewhat fuzzier yellow glow. It turns out that’s from sodium, which emits yellow light when it’s excited. It was thought for a long time that this sodium might be coming from sea salt blown into the air, but it turns out to have a more heavenly source: meteors! As these tiny rocks from space burn up in our upper atmosphere, they leave sodium behind. It’s not much, but sodium is a very enthusiastic atom, and glows brilliantly. So even though there’s much less of it than oxygen, it’s still pretty bright.

There are other processes, too, which contribute different colors at fainter amounts. For example, when two oxygen atoms combine to form an O2 molecule, it has a bit of residual energy left over. It can get rid of that by emitting a blue photon. This is usually pretty faint, and occurs at 95 km, right at the bottom of the green layer. That’s not a coincidence! Remember, that’s the height where collisions become frequent, so that same process that quenches the green glow — oxygen atoms smacking into each other — is what causes the blue glow.

And if all this sounds familiar, it may because these same processes are what causes aurorae to glow at different colors, too! In that case, though, the source of energy isn’t light from the Sun, but fast subatomic particles from the solar wind or solar storms. These come zipping in like little bullets, slam into the air, and blast apart oxygen molecules like shrapnel. After that, the process of the atoms giving off energy is pretty much the same as what I’ve outlined above.

The levels of complexity of all this get serious pretty rapidly past what I’ve described, with electrons jumping from one energy level to another, Einstein coefficients, forbidden transitions, and collision probability cross-sections. You can find out all about those online if you wish, and more power to you if you do.

But in fact, I’d say this whole topic seems to go from relatively simple to fiendishly complex in an almost — ahem — quantum leap.

Still, it’s fascinating, and I had a lot of fun poking around websites and quantum mechanics descriptions trying to figure this all out. The pictures taken of the Earth from space are always lovely and engaging and awe-inspiring, but they become even more so when there is understanding — when there’s science — behind them.

Knowing is always better. Always.

Image credit: NASA

Related Posts:

JAW DROPPING Space Station time lapse! (Seriously, this is incredible)
The fiery descent of Atlantis… seen from space!
Turns out, it *is* a river in Egypt
Psychedelic space station stars and cities
A celestial visitor, seen from space


Comments (26)

  1. Wzrd1

    Next up, an experimentally proven cause for terrestrial gamma ray bursts, a fair number of which show evidence of pair production.
    Phil, do you remember the old days when the atmosphere was considered boring by physicists? ūüėČ

  2. Nigel Depledge

    So . . . it’s not quite clear to me if an O atom (these are radicals, too, BTW) can emit a green phton and a red photon or just one or the other.

    My chemical intuition tells me it should be one or the other (because of, among other things, E=h[nu]). Is it, or not really?

  3. The sodium layer is the same one used with adaptive optics lasers, right? I didn’t know that emitted it’s own light.
    Science rocks!

  4. ceramicfundamentalist

    quick question: if green light is emitted seconds after uv dissociation of O2 molecules and red light is emitted minutes after dissociation then why do we see these airglow patterns in the night sky? shouldn’t all the atoms have recombined very quickly after the sun has set?

  5. David Taylor

    Great question from ceramicfundamentalist. In addition, does the luminosity show a declining light curve with increasing time of the source out of direct illumination by solar UV?

    Dr. Rivest’s work is wonderful!

  6. Other Paul

    I don’t know what it is or why it happens, but as soon as I see that it’s a vimeo I know it’s pointless clicking the play button. Jerky, stuttery, audio and video just make it unwatchable. Doesn’t happen with youtube, which always plays well. Very strange.

  7. Jay

    Nigel (#2): The same atom can emit both photons. The green photon (higher energy) is from a higher excited state to a lower excited state, and the red photon is from the lower excited state to the ground state. The details are in, particularly the footnotes.

  8. ceramicfundamentalist (4): I’ll be honest, that’s complicated and I’m not 100% sure. I spent quite a bit of time thinking about that very question yesterday as I wrote this post, in fact.

    Part of it is as I describe: you’re looking through a LOT of air – 100s of km worth – so even a faint glow builds up. Also, there are other processes creating the green glow besides the prompt oxygen emission. Also, the geometry may be that we’re looking toward the twilight line, where the green glow would be strongest.

    Or it could be something I’m missing. Looking this up online was extremely difficult; I found contradictory or inconsistent explanations, and some that were extremely technical and (haha) unilluminating. I read papers dating back decades, as well as many from just a few years ago. It’s maddening, but I wasn’t able to find a definitive answer. I’m hoping that someone reading this may know, and I may try to contact some atmospheric scientists to get more info too.

  9. Christine Bayne

    That video is the most stunningly beautiful thing I have ever seen. Cripes, it should have come with a warning. Pass the tissues, please.

  10. Wzrd1

    Phil, from what I’ve been reading online, it appears to be a bit in the knowledge gap and there is some ongoing research (typically suffering from sparse funding).
    Some potential additional energy sources are resonance between layers of the atmosphere, which pump the atoms to higher levels, cosmic ray induced energy to the upper atmosphere, even thunderstorms generate RF energy that resonates between layers and even between the ionosphere layers.
    But, from all that I’ve been reading, we simply don’t really know, as it’s not really easy to get long term measurements of the upper atmosphere.
    Hence, the planned satellite missions to observe the upper atmosphere and magnetosphere, HAARP and many other efforts, each to their own layer to study.
    And also hence, my hint in my previous comment about terrestrial gamma ray bursts, which previously were thought to be impossible, until they were observed by satellite and are still an ongoing area of research as to their ultimate cause (OK, lightning is known as a cause, but not the precise physical process).

    The more we look, the more we find.
    And in science, it isn’t typically the “eureka moment” that moves our understanding forward, it’s more often “Hmmm, THAT’S interesting!”

  11. JES

    Just out of curiosity… This “air glow” (unlike the auroras) doesn’t seem to be at all localized, but, well, everywhere. How come we can’t observe it from ground level? Shouldn’t the night (or at least the evening) sky be glowing similarly?

  12. TimO

    If you were CGI’ing a movie and used that clip no one would buy it.

    Just goes to show that nature can be wilder than you can imagine….

  13. Matt B.

    @6 Other Paul: Is Vimeo set to HD? If so, try turning it off.

  14. sam

    bunch of UFOs in there lol

  15. Pete

    Wow! Lots 0f satellites!

  16. Other Paul

    @12 Matt B – Thanks for that. Though the result is contrary to expectation. Flipping the HD switch results in a big notice saying HD IS ON and the stuttering stops, switching again yields HD IS OFF and the stuttering returns. Either their feedback is misleading or switching to HD actually ‘improves’ things. I leave improves in quotes since although audio and video is now smooth and continuous, the lipsync is about a second out and the video plays slightly slomo. But at least it’s watchable, which is what I need for these astrovidialities. (Still odd that, whatever I do, youtube has none of these issues, HD-on-off-regardless).

  17. Wzrd1

    Other Paul, perhaps one of two things is going on.
    Either your flash version is out of date.
    Or, your ISP is filtering that less used source, to conserve bandwidth.
    I’ve personally ran into BOTH over the years, either the ISP is lagging in response to demand of clients in against of limiting less lawful use OR older, less compatible versions of software causing the very problems you describe.
    On a few occasions, a bad install, which responded far better with a system reimage, fixed the issue.
    Without a LOT of interesting testing, one cannot diagnose the issue further to help you.

  18. Jonnie

    I wonder what colours other planets might glow in. You never really see those kind of effects in sci-fi artwork/SFX.

  19. Anthony

    JES, the airglow can be observed from the ground but it appears much fainter because we’re looking upwards through a thin layer rather than horizontally through the layer as the astronauts do. It can be detected by astronomical telescopes, especially those doing spectroscopy as the airglow occurs at specific characteristic wavelengths (or frequencies if you prefer). In fact for infrared astronomy the airglow is a serious nuisance, that diffuse red emission from hydroxyl radicals is much brighter in the infrared than it is in visible light and seriously hampers infrared astronomy done from the ground. It’s like trying to look at the stars during the day!

    The airglow does get fainter over the course of the night as you would expect but it doesn’t completely disappear. The O2 emission fades away in the first couple of hours but the hydroxyl emission will only drop to half its initial brightness by the end of the night. On top of the gradual decline the airglow brightness from any particular patch of the sky varies randomly on timescales of minutes due to density waves in the upper atmosphere.

    These atmospheric phenomena are of course interesting in their own right (and make for some great photos!) but they have also motivated significant efforts on the behalf of infrared astronomers to remove the airglow from the much fainter signals from distant galaxies or cool stars. Various technologies with suitably exotic names have been used, OH suppression masking, Rugate filters, volume phase holographic filters and fibre Bragg grating OH suppression.

  20. Anthony

    To see what the airglow looks like from the ground in the infrared check out the timelapse movies on this page:

    Monochrome and low resolution so not exactly pretty but a great illustration of just how different the night sky is in the infrared. Notice how bright the airglow is in comparison to the stars, and how uneven and variable it is.

  21. Other Paul

    #21 @Wzrd1 – Thanks for your observations. I guess this isn’t really the place for a long discussion of this but this is the only place I’m getting anything useful! I have googled the topic to no avail. I can’t think of a useful search term which doesn’t flood the hits with general jitter-stuttery issues. Nobody else (unless it’s buried in those hits) is discussing what I’m experiencing, which is vimeo=bad and youtube=good. I find it hard to believe that flash ‘knows’ the source and behaves accordingly, but you never know. Play me a video without telling me its provenance and, if it plays OK I’ll say it’s youtube and if it stutters I’ll say it’s vimeo. The stutteralitude really is that reliable an indicator of source! ISP throttling according to source is, I suppose, more believable.

  22. Brian Too

    @ 9. Christine Bayne,

    +1 on your comment.

    Nice to see something good from Moby too.

  23. Nigel Depledge

    @ Jay (7) –

  24. JES

    Thank you so much for the information, Anthony (#19-20)!

  25. NoseyNick

    Like Christine, I’m in tears. WOW. One of those “pale blue dot” or “eagle has landed” or “crucibles… enriched guts… universe is in us” moments that makes me cry like a baby.

  26. Mmmmm doubly ionised goodness!


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar