1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Why does the atmosphere get colder with elevation?

  1. Nov 15, 2012 #1
    ~50 years ago, Richard P. Feynman asserted that the reason the atmosphere is colder at higher elevation is because "The ground is heated by the sun, and the re-radiation of heat to the sky comes from water vapor high in the atmosphere; so at high altitudes the air is cold--very cold--whereas lower down it is warm" ~ FLP, Vol II, 9-4, paragraph 2.

    I'm not even sure what that means, but I believe it is wrong.

    Mind you, atmospherics was not Feynman's expertise, and the field has advanced considerably since he gave his lectures. He's quite clear that his lecture on electricity in the atmosphere was speculative. So, I'm not trying to take anything away from the worth of the FLP. I know of very little else in the FLP which has been superseded with the ensuing half a century. Besides, that chapter is what led me to hunger for an understanding of the dynamic of the atmosphere.

    My understanding of what makes the atmosphere colder at higher elevations is that convection currents act as a "heat pump" drawing heat from higher elevations and "squeezing it out" as the pressure increases. I can't recall where I read that, but it was probably somewhere close to James F. Price of the Woods Hole Oceanography Institute.
     
  2. jcsd
  3. Nov 15, 2012 #2
    Hetware,
    Greetings.

    Are you asking specifically about the troposphere? Above the troposphere the temperature zigzags back and forth between increasing and decreasing.
    (increases in stratosphere, decreases in mesosphere, increases in thermosphere).
     
  4. Nov 15, 2012 #3

    Drakkith

    User Avatar

    Staff: Mentor

    While convection does happen, it actually brings hot air from the surface to higher altitudes. The main reason the lower atmosphere is warmer is because most of the radiation from the Sun is absorbed at the surface and then transferred to the air from there through physical contact.
     
  5. Nov 15, 2012 #4

    Yes.
     
  6. Nov 15, 2012 #5
    That's what Feynman specifically says doesn't happen. The temperature drop through adiabatic expansion is greater, according to his, un-demonstrated assertions, than the pressure differential. That is, warm air rising will cool faster than the ambient temperature differential, due to adiabatic expansion.
     
  7. Nov 16, 2012 #6

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    The atmosphere is very complicated. Two over-simplified models can be relied upon to yield conflicting predicted results for temperature at a given height. I don't think there's a lot of point in arguing, using anything but the most advanced and detailed model.
     
  8. Nov 16, 2012 #7

    Drakkith

    User Avatar

    Staff: Mentor

    Is it not possible that both effects happen?
     
  9. Nov 16, 2012 #8
    No, convection has the opposite effect. To my knowledge convection never increases a thermal gradient, only decreases it. The main reason why is that energy from the sun is transferred into the ground, but not the air, because the air is transparent and the ground isn't. So an object further from the ground will be further from a heat source, so will be colder.
     
  10. Nov 16, 2012 #9

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    On the scale of our atmosphere, convection is more important than heat conduction - the equilibrium is indeed a temperature gradient. But if we would have that equilibrium, we would have nearly no convection. The sun heats the ground, increases the gradient and induces convection. So I think the answer is "both".
    There are exceptions, where the gradient is too small to give convection.
     
  11. Nov 16, 2012 #10

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    It's correct. Feynman was essentially writing about the greenhouse effect.

    Imagine things would look like if the atmosphere was completely transparent. This would result in a very different world than the one on which we live. Surface temperatures would fluctuate much more than does ours because of the surface would receive the full brunt of the Sun at day and radiate to the darkness of space at night. You can see this to some extent in the desert where humidity is very low. (However, even in the desert, other greenhouse gases such as CO2 still mediate temperatures.)

    In this world with a transparent atmosphere, there would be a rather thin layer just above the surface where atmospheric temperatures fluctuate over the course of a day. Beyond this layer, the atmosphere would be very close to isothermal up to the point where the atmosphere stops acting like a dense gas (that's the mesopause in our atmosphere).

    We don't have a transparent atmosphere thanks to those greenhouse gases that absorb in the infrared, and also to ozone that reacts chemically with ultraviolet radiation. It's those greenhouse gases coupled with daytime heating / nighttime cooling at the surface that steer conditions away from isothermal. In the lower atmosphere, water is by far the most potent greenhouse gas. What Feynman wrote was essentially correct, at least with regard to the troposphere.
     
  12. Nov 16, 2012 #11

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    If you look at the maximum height that clouds can be found is round about the tropopause. If I am right that the max height of clouds indicates the maximum height that convection can work the it would seem to me that the temperature lapse rate must be affected by convection. If energy were not being transferred to this height by convection, wouldn't that increase the lapse rate?
    I know it's a dodgy subject so I could be falling into a pitfall here. Am I?
     
  13. Nov 16, 2012 #12

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    That's because there's a temperature inversion at the tropopause. The ozone layer is above the tropopause. The ozone-oxygen cycle absorbs energy from incoming ultraviolet sunlight, heating up the stratosphere. Temperature increases with increasing height in the stratosphere. Temperature inversions create an absolutely stable atmosphere. There is little if any convection in an absolutely stable atmosphere.
     
  14. Nov 16, 2012 #13
    I should qualify what I said about Feynman regarding the consequence of adiabatic expansion not sufficing to cause a continual updraft. He states that such an updraft can be sustained if there is sufficient water vapor because as the rising air cools, the water vapor condenses releasing latent heat, thus warming the air, which in turn causes it to expand and rise still more.

    Kerry Emanuel, in Divine Wind: The Science and History of Hurricanes, gives a decent account of atmospheric heat. He doesn't make the claim Feynman makes regarding the behavior of dry air, but he does agree that water plays an important role in the process of cooling Earth's surface through the same mechanism Feynman described.

    Emanuel asserts "A detailed calculation of the radiative balance of the earth's surface and atmosphere yields an average surface temperature of 135°F, much warmer than observed." That is ignoring convective effects.

    He also observes that:

    "the atmosphere is nearly opaque to infrared radiation. The radiation absorbed by gases in the atmosphere is re-emitted, both upward and downward. In consequence, the earth's surface receives both infrared radiation emitted downward from the atmosphere and solar radiation. In is a remarkable fact that, on average, the surface receives more radiation from the atmosphere than directly from the sun."​
     
    Last edited by a moderator: Nov 17, 2012
  15. Nov 17, 2012 #14

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    Maybe, but the surface gets both the re-radiated IR and the direct visible light, so gets more in total than does the atmosphere. The ground becomes hotter than the atmosphere, and the proof is the existence of large scale convection. Convection (up to the tropopause) acts to flatten the gradient, but adiabatic expansion limits the ability of convection to do so.
    I liken it to sand dunes on a beach. The wind and waves pile up the sand; gravity acts to level it out; sand particle interactions limit the flattening out.
     
  16. Nov 17, 2012 #15

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    I now appreciate the Ozone effect but I'm not sure whether that answers all of the 'why' question. It may just be stating a situation. I think that I am, in effect, asking how much the convection effects the position of the tropopause. I would imagine it's a matter of where one effect runs out and the other takes over.
    The height that turbulence reaches is governed by the a mount of heat energy at ground level, I think. I seem to remember calculating and equating the change in mgh for a given amount of thermal energy and that gives an indication of the height that a bubble of hot air can reach. Afair, that corresponded to the max height that you find clouds.

    I wonder what temperature distribution you would get if you could suppress convection (by a series of vertical baffles, for instance). I suggest that the tropopause would end up quite a bit lower. Someone is bound to have studied a model like that.
     
  17. Nov 17, 2012 #16

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    Read again. Feynman specifically says exactly the opposite. If an isolated packet of rising warm air cools faster than the environmental lapse rate that isolated packet will fall back to earth. This happens for example, when wind forces air upward over a mountain on a clear day. For a rising packet of air to continue rising, the air in that packet must cool at less than the environmental lapse rate. When this happens, the rising packet will be warmer, and thus less dense, than the surrounding air. It will continue rising. This is exactly what Feynman says, sans the term "lapse rate".

    In fact, he goes on to say that the observed lapse rate inside clouds is less than would be predicted by assuming isolated packets of rising air. This is because of entrainment (a term that Feynman does use). The rising packets are not quite isolated; they eventually do mix with the surrounding air. That surrounding air cools the rising air, or alternatively, the rising air warms the surrounding air.

    Another way to look at it: Convection, when it occurs, transfers heat from the surface to the atmosphere. Convection can only occur when the lapse rates are conducive to convection.


    And evaporation. Evaporation has a much greater cooling effect on the surface than does simple convection (e.g., thermals).


    That convection would not exist were it not for greenhouse gases. The ground would be a whole lot warmer than the atmosphere if the atmosphere was transparent in both the visible and infrared portions of the spectrum -- and there would be no large scale convection.
     
  18. Nov 17, 2012 #17

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    This is interesting. Are you saying that it is not actual 'contact with the ground' by the air that causes the heat transfer and that it is absorption of IR by the air above the ground (and most of the transfer would need to be quite close to the ground, too)? Is this totally reasonable? I say this because, unless it works that way, it doesn't seem likely that you could get significant, thermals over small areas of ground (ploughed fields etc) as used by gliders and birds - the heat transfer being much more general and diffuse into the higher regions of the atmosphere. If it is true, then it would be difficult to explain the blanketing effect of quite high layers of cloud at night.
    Maybe there are two mechanisms, one which explains the heating effect close to the ground and another that explains the effect over thousands of metres.
    This is all fiendishly complicated and it's hard to get a measure of the significance of all the mechanisms involved.
     
  19. Nov 17, 2012 #18
    Air is not a very good conductor of heat. But according to what I posted above from Kerry Emanuel, air does absorb IR radiation very effectively. It would make sense to me that most of the surface to air transfer is radiative as opposed to conductive. Radiative absorption certainly allows for a larger volume in which the transfer will take place.

    As for thermals, I don't see the disconnect. If air readily absorbs IR, then it would make sense that air over a warm radiator near the surface would be heated significantly more than nearby air over a colder surface.
     
  20. Nov 17, 2012 #19

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Yes but, if it absorbed it so easily, there would be no IR left for high cloud cover to make any difference to night temperatures. Do you see the conflict here? it probably needs figures in the explanation.
     
  21. Nov 17, 2012 #20

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook