Mystery crystals

A reader writes:

I was walking down the street at three in the morning after a night out, in the middle of winter [here in Australia], and there was twinkling frost all over the top of a parked car. And the next parked car. But not the one after that.

I kept looking, and the difference was that cars that were parked under a tree had no frost, but cars that were in the open were frosty.

The air temperature was pretty low, but it wasn't below freezing - I checked later and the local weather station said it got down to about 4 degrees C.

Did frost fall down out of the sky and somehow... stay?


It was a clear night with no breeze, right?

On a clear night, the sky above you is a window to deep space. There's no sun keeping things warm, no diffuse sky radiation making the sky blue and at least a bit warm wherever you look; just a blanket of air, and then space.

Heat can pass by convection, conduction and radiation. Radiation, for most items humans encounter, is the least important of these three paths. But if an object has a wide view of something which, like deep space, is close to absolute zero, then it can radiate enough heat to drop below zero Celsius, even if the ambient air temperature is a little above freezing.

If there's even a light breeze, the passing above-freezing air will keep surfaces too warm for frost to form, by allowing heat to move by convection - in this case forced convection (as in the case of a computer CPU's heat sink cooled by a fan). Likewise if a surface is directly connected to something with a large heat capacity, allowing that surface to stay warm by conduction (as in the case of the CPU itself, in physical contact with its heat sink). The thin steel roof of a car will form frost in these conditions; a solid block of steel would not, because radiation wouldn't be able to cool all of it enough before the sun came back up.

The less of a direct view a surface has of the sky, the smaller this already-small effect will be. So cars - or rubbish bins, or other thermally-isolated surfaces - that're in the "shade" of a tree or building probably won't frost up. (There could be some interesting odd cases, if for example a car is parked next to a skysraper covered with IR-reflective glass.)

This same phenomenon can be used to make ice in a desert, if that desert has clear, still nights. Wide shallow trays of water held up off the sand by narrow supports can freeze surprisingly quickly.

Psycho Science, as I have brilliantly decided to call it, is a new regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

10 Responses to “Mystery crystals”

  1. Mohonri Says:

    Very, very interesting. I ran into the same phenomenon when we still lived in Texas. We would park our two cars in the driveway. The one on the left was mostly under a tree, while the one on the right was under a mostly open area. The car on the right would regularly require defrosting in the morning, but the one on the left would be clean as a whistle.

    I always wondered how that happened, and now I know.

  2. kai Says:

    I'm liking the new site. Nice clean WordPress Theme.

  3. Bern Says:

    That's the greenhouse effect in action. Or rather, not so much... 4ºC in winter means the air is likely to be very dry, so there's much less water vapour than usual to backscatter the IR coming up from the ground.

    Which is why cold-but-cloudy nights are generally quite a bit warmer than cold-and-clear nights.

  4. RichVR Says:

    I was always under the impression that space itself has no temperature. While objects in space would eventually, over a long period of time, radiate heat. So that an object in space is eventually 2.7 degrees Kelvin.

    • Anne Says:

      You're right that space does not have a temperature. In fact, many things don't have temperatures. If you let something sit for long enough, all its components (atoms, molecules, radiation, loose electrons, whatever) will shift energy randomly back and forth until it's distributed "evenly". In statistical mechanics, you can actually predict the distributions that occur. A system where that's happened is "thermal", and it has a well-defined temperature. Most systems under terrestrial conditions are thermal, or very nearly; in fact the reason building lasers is hard is because they need a non-thermal distribution in the lasing medium. But when you have very low densities or very loosely-coupled components, it can take a really long time for things to "thermalize", and in the meantime they don't really have a well-defined temperature. You can sometimes talk about temperatures of different components: for example, in much of the galaxy the interstellar medium is at millions of degrees (and between galaxies it's even hotter), but the radiation field, the photons, are at 2.7 K (except for the relatively few photons from stars and whatnot, which not only have a hotter energy distribution but a non-uniform distribution of directions).

      All this is to say that, no, space doesn't have a temperature. So you need to think about how you're exchanging energy with it. If you're exchanging radiation with dark sky, then it's effectively 2.7 K (this is how they're planning to keep the James Webb space telescope cool). If you're exchanging radiation with the Sun, it's 5700 K — except that the Sun occupies only a small fraction of the sky. If you're exchanging energy with the interplanetary medium, it's much hotter, but there's also the solar wind to consider. In practice, radiation dominates, so you'd think about the Sun and the cosmic microwave background, and maybe planets if you're close to one.

  5. matt t Says:

    Years ago, I asked myself if cold could radiate - it certainly sometimes feels like a cold object is radiating its cold at you. I figured that no, cold can't radiate, any more than dark can. It's just an absence of heat radiation that you feel on your skin.

    This would be the same thing, although it had never occurred to me. I've noticed that the road underneath trees is often dry on cold mornings, while exposed ground is wet - I'd always figured that the trees caught falling dew during the night, keeping the ground dry. But maybe it's just the ground being warmer under the tree that prevents dew condensing onto it there? I'll check this out next time it happens!

    Looking forward to more psycho science, Dan. :)

  6. Stark Says:

    OK, So I've understood this was the reason for frost formation on clear calm nights where you are within, generally speaking, 1-4c of freezing... but here's a frost phenomenon I don't get :

    I come out to my car on a cold morning - approximately 3c - no frost on car but it is covered in condensation. I get in, start the car, hit the windshield wipers and BAM the condensation turns to ice as the wipers make their first pass. My first thought was super cooled water ( but the ambient temp is over freezing so I think that's out. So... Mister Dan the Science Man.... whats up here? ;)


  7. Mohonri Says:

    I think the same effect might also apply when the temperature is at or near the dewpoint.

  8. hagmanti Says:

    I don't think supercooling is out. If it's cold enough to freeze water (and make frost), it's cold enough to supercool it. Same exact mechanism (radiative cooling) but if there's no impurities (or the temperature of the water doesn't get too low given the amount of impurity present) and no vibration (no breeze, or this whole discussion is moot), then it'll supercool instead of freeze.

    I'd guess the difference is:

    A) How clean the car in question is (clean cars would be more likely to have supercool dew)
    B) How clean the air is (lots of dust / particulate matter would make it more likely that small droplets would freeze as they condensed-- that would be my guess, but the physics of water freezing out of air get really complicated really fast, and what'd I'd really love is to do some experiments on it)

  9. Darrell T Says:

    Very interesting. I'm not sure I agree completely with the explanation, though. As written, it sounds almost as though the vicious cold of space is sucking heat from the exposed surface, which is, of course, nonsense. The amount of energy a body loses via radiation is primarily a function of its temperature (Google "black-body radiation" for details) and has nothing to do with its exposure to nearby or distant energy sinks.

    I think the missing part of the explanation, inherent in the original discussion but never made explicit, is that the part of an object's "view" that doesn't look out on a good energy sink must be looking at something that is also radiating energy in the electromagnetic spectrum, most likely in the infrared in a terrestrial environment. For the car parked under a tree, most of the car's view of the sky is obscured by a body that is emitting infrared light at essentially the same rate as the car, so its energy exchange with its environment is balanced.

    I suppose it's also possible that the tree remains slightly warmer than the air over the course of the night, which would cause a slight convection current as it warms the air around it. I suspect, though, that such a current would be too slight to have much effect on the car parked beneath the tree.

Leave a Reply