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 heat in the atmosphere mostly go down?

  1. Nov 6, 2015 #1
    The wiki article on the greenhouse effect (https://en.wikipedia.org/wiki/Greenhouse_effect) has a top illustration which shows the flow of heat and energy (Watts per meter squared) between outer space, the atmosphere, and Earth's surface. The illustrations shows that of the 519 in the atmosphere, 195 (38%) goes up into space and 324 (62%) goes down to earth. The article writes "the atmosphere radiates energy both upwards and downwards; the part radiated downwards is absorbed by the Earth's surface." I believe the infrared radiation absorbed and emitted by the greenhouse gases goes half up into space and half down to earth. Does the heat then mainly go down because up is empty space and down is thicker air. This is weird to me because I am used to heated air rising?
  2. jcsd
  3. Nov 6, 2015 #2
    Hello stfaivus,

    Consider there are different forms of heat flow: Conduction, Convection, and Radiation.

    Conduction occurs from two objects touching.
    Convection is the rising and falling of heated vs cooled air.
    Radiation is related to EM waves produced by a heated body.

    So what you're seeing is that the atmosphere is giving off heat to space. My question to you is which of the three provided would be the type of heat given off into space? Now think about in what ways it could transfer heat to the earth? Are the more ways to transfer heat to space from the the atmosphere or more ways to transfer it to the earth?

    Think about it for a while. Maybe turn on an incandescent light bulb and think about the different types of heat transfer from that bulb.

    And you're right, hot air is rising which allows the air to release some of the energy it's absorbed by doing mechanical work, that when it's big enough it results in lots of air moving tha then blows your hat off and makes you say "Boy it's windy out today." It really is amazing to think of all the ways that energy is getting moved around.
  4. Nov 6, 2015 #3
    Hi DrPapper,
    In the illustration, it shows 350 (Watts per meter squared) absorbed by the atmosphere. If this is all re-emitted, half up, half down, that accounts for 175 of the rising 195 from the atmosphere. That would also account for 175 of the descending energy. There's another 149 that goes down, and only another 20 that goes up. It still seems more goes down, strangely.

    Now to your prompts. Which type of heat can go into space? Radiation only, because there are no atoms to carry the heat (or sound). So I guess that's the solution. Any heat will not pass into space. Imagine heat being created high up, by the absorption of some incoming solar radiation by the atmosphere. At the boundary with space, the heat can only go down. The radiation heated air releases goes equally up and down. So I guess it makes sense more energy flows down from the atmosphere than up in this diagram.

    Thinking of an incandescent bulb or a candle, the heat goes out by conduction to the surrounding cooler air, by radiation of its visible light and infrared radiation, and I can imagine eddies of convection in the room, perhaps when the rising heated air hit the ceiling, but I haven't seen evidence for convection currents around a light bulb or candle. I do see convection currents in the candle's liquid puddle of wax around the burning wick. Here specks of black burnt wick swirl.

    Does this reasoning make sense?
  5. Nov 8, 2015 #4

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    You are correct that at any one point in the atmosphere, thermal radiation is emitted uniformly in all directions. That does not mean that the atmosphere as a whole must emit thermal radiation in the manner you describe.

    Consider a parcel of air near the surface of the Earth. That parcel absorbs thermal radiation emitted by the surface of the Earth. It also absorbs thermal radiation emitted downward by higher portions of the atmosphere. To be in thermal equilibrium, that parcel must emit the same amount energy it absorbs. Some of this is directed upward, some downward.

    That upward-directed thermal radiation will be absorbed by some higher parcel of air. This higher parcel of air will also absorb downward-directed radiation from even higher parcels of air. To be in thermal equilibrium, this higher parcel must radiate all of that absorbed energy as emitted energy, half directed downwards, half upwards. This continues all the way to the top of the atmosphere. The natural consequence is that more energy is directed downwards than upwards.
  6. Nov 8, 2015 #5

    Yes, you are on the right track. I'd say to not be too focused on the specific numbers they're giving and consider that since the atmosphere will radiate it's energy equally in all directions and that most of that radiation will go into other parts of the atmosphere than into space. Since those molecules can also transfer heat by conduction they can transfer it again to other air molecules but also the Earth's crust. So then the planet get radiated energy and conduction energy. The universe can only get the radiated energy. :D The convection really just keeps the energy moving about in the atmosphere until it's transferred as radiation or conduction.

    If you look carefully around a light bulb or candle you'll see that near the tip of the flame or surface that object appear warped and dancing about. That's because the convection as well as a change to the index of refraction of that air being warmer (bends light differently). But to press the conduction and radiation, notice that grabbing that light bulb will transfer a lot more energy a lot quicker than the radiation will. Please don't actually grab it, but you get the idea. LOL So when you have physical contact you can get not just the radiation but also the conduction.
  7. Nov 9, 2015 #6

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    You are steering stfaivus in the wrong direction, DrPapper. Conduction and convection play a rather minor role in the Earth's energy budget, accounting for less than 4% of the energy that leaves the Earth's surface. A form of energy transfer that you missed, latent heat, accounts for a bit over 17% of the total. The vast majority (a bit over 79%) of the energy leaving the Earth's surface is in the form of thermal radiation.
  8. Nov 9, 2015 #7

    I strongly support your two posts on this thread. I might quibble a bit on your numbers, however. My readings and calculations on the thermal energy that leaves the surface of the planet breaks down into 71% being absorbed by the atmosphere, 23% being radiated into space, 5% being conducted into the atmosphere from the surface, and 1% being added to the atmosphere by enthalpic cycling.

    I use this latter term to refer to the energy transfer that takes place when water is evaporated from the Earth's surface at a higher temperature and condenses out within the atmosphere at a lower temperature. This percentage is based on the estimate of the Earth's mean annual precipitation at one meter.

    I do not include the "latent heat" transfer in my numbers because my readings suggest that the enthalpy of condensation goes into the condensate, not the surrounding atmosphere. There, the condensate either evaporates once more or falls to the Earth as precipitation and is removed from the atmospheric system. The apparent temperature increase seen in laboratory experiments on the enthalpy of condensation is a statistical artifact. When condensation occurs, it is the "cooler" molecules that are removed from the vapor system first. Consequently, the mean thermal energy (temperature) of the remainder increases. However, the vapor system has lost thermal energy to its environment (the condensate), not gained any. This apparent paradox occurs because temperature is an average value, while thermal energy (heat, in engineering terms) is a total value.

    By the by, almost all of the thermal energy that leaves the Earth's surface is absorbed within the first one hundred meters of the atmosphere. Most atmospheric radiation is absorbed by the atmosphere. Only very small proportions of atmospheric radiations reach either the Earth's surface or outer space.
  9. Nov 9, 2015 #8
    I believe you are misreading the diagram.

    The atmosphere is between the Sun and the Earth surface and plays a role in the transmission of energy from the Sun onto the Earth. The atmosphere is also between the Earth and space, and plays a role in the transmission of energy from the Earth surface to space.

    You and Wikipedia are conflating the two energy flows. For the sunlight that flows thru the atmosphere to the Earth, more gets thru than is blocked. However the infrared emitted from the earth surface that is absorbed by the atmosphere, half gets re-emitted up, and half down, just as you state. I think both of those energy flow diagrams try to show the sun incident energy and the earth emitted in one diagram, to show that the earth at a steady state temperature must have 343 Watts per meter hitting it, and 343 watts per meter leaving, or the temperature would change. (I believe the 343 watts per meter number is incorrect).

    The earth surface will have a black body temperature, and radiate with the spectrum for that temperature. Just as an electric stove heating element does not glow visibly at a low temperature, the earth also only "glows" in the infrared. And just as the higher stove top settings increase the energy emitted, higher temperatures increase the emitted energy. The average temperature is a steady state temperature at which the emitted energy equals the absorbed energy, taking into account the atmosphere.
  10. Nov 10, 2015 #9

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    No, he's not. Here's a more detailed and more recent diagram of the Earth's energy budget:


    The atmosphere as a whole emits more thermal radiation to the Earth's surface (340 W/m^2) than it does to space (200 W/m^2). This is the correct physics. Thinking that it should emit the same amount of energy in both directions is incorrect.
  11. Nov 10, 2015 #10
    Perhaps I am misreading his confusion:

    "I believe the infrared radiation absorbed and emitted by the greenhouse gases goes half up into space and half down to earth."

    That belief is accurate. The energy flows shown are not confined to the energy absorbed and emitted by greenhouse gases. If you look at energy that is absorbed by greenhouse gases, roughly half will be emitted with an upward result and half with a downward result. The majority of the upward energy flow in the diagram looks to be convective heat up.

    My comment was "The average temperature is a steady state temperature at which the emitted energy equals the absorbed energy, taking into account the atmosphere." This is true. The EARTH will absorb energy and emit energy in equal amounts, when it is at a steady state. The earth average temperature reflects the steady state, the point at which entire earth (including the atmosphere) emitted energy (from a body with a given temperature) equals the absorbed energy of the entire earth (including the atmosphere). There are other factors, but the essential model is a steady state. The Wikipedia drawing has the earth receiving a net energy of 340.3 absorbing a net of 0.6, and 339.8 as the net outgoing energy. That is not a exact steady state, but one with slight global warming.

    The energy flow from the earth in the drawing is not exclusively re-emitted energy. In the legend it says: "conduction, convection, and evaporation". I don't really have an intuitive sense of how the energy of a hurricane goes .... but it might all end up thrown onto the earths surface. But infrared photons that are absorbed by the atmosphere and re-emitted a small time interval later, are emitted in a random direction. If we approximate the earth surface as a plane, half will emit and intersect with that plane, and half will not.

    I think this is a case of different people saying slightly different true things. I'm probably not being as clear as I think. And the drawing is a bit confusing.

    Certainly, if I am wrong in the more narrowly limited statement I tried to pinpoint, let me know. I believe we are just talking about different energy ... infrared radiation, vs net heat.
  12. Nov 10, 2015 #11
    Heated air can rise, but the main sorce of heat to atmosphere is the reflectión of solar radiation on earth surface, so the nearer to the surface the hotter the athmosphere at first sight. So that is the cause to be hotter near the ground, without need to say that heat _goes_ down. It comes from below, really
  13. Nov 11, 2015 #12
    Part of the confusion in this thread comes from trying to show all five of the relevant atmospheric heat budgets in a single diagram. These budgets include atmospheric inflow (solar radiation, terrestrial radiation, conduction, and enthalpic cycling); atmospheric outflow (longwave radiation to the surface, and to outer space); surface inflow (insolation plus atmospheric longwave radiation); surface outflow (longwave to the atmosphere, longwave to outer space, conduction to the atmosphere, and enthalpic cycling [see post #7]); and finally earthshine (emitted longwave terrestrial radiation, reflected solar radiation from the surface, and upscattered and reflected solar radiation from the atmosphere--mostly clouds).

    Just taking the first of these we have the atmosphere receiving most of its thermal energy (heat, in engineering terms) from terrestrial longwave radiation (69%), shortwave solar radiation (26%), conduction from the Earth's surface (4%), and enthalpic cycling (1%). All of these percentages are approximations, and subject to amendment as the scientific evidence accumulates.

    To address one persistent source of confusion, net (incoming minus outgoing) longwave atmospheric radiation varies considerably from one elevation to another at a single location, and even more from one location to another over the Earth's varying climatic regions. This radiation rarely travels more than a hundred meters or so before being absorbed by the so-called "greenhouse gases"--primarily water vapor. More thermal radiation goes downward from the denser, more humid lower hundred meters than goes upward from the more rarefied and drier upper atmosphere. You simply cannot treat the extremely complex atmosphere as a single homogeneous system.
  14. Nov 12, 2015 #13
    Thank you everyone on this thread! This seems an important and timely issue to get very clear on.

    My two initial questions are more clear now, especially from the slightly more detailed NASA energy flow diagram provided by DH on #9 and the interesting discussion throughout. The diagram (https://en.wikipedia.org/wiki/Greenhouse_effect#/media/File:Greenhouse_Effect.svg) I was initially trying to interpret shows 519 for total heat and energy absorbed by the atmosphere, which moves 195 (38%) up into space and 324 (62%) down to earth. All units are Watts per meter squared. It also shows 452 absorbed by the atmosphere from the surface of the earth, 350 of which is by greenhouse gases. My two questions were: one, why does 62% of the heat and energy go down? and two, if 350 of the 452 absorbed by the atmosphere is by greenhouse gases, how are the other 102 absorbed?

    The second question is easily answered by the NASA diagram. It shows 86 from latent heat and 18 from thermals, or conduction. The first question, though more clear, is still a little confusing to me. For now I can interpret the first question like this. All the energy absorbed by the atmosphere-infrared radiation IR rising from earth (358) + latent heat (86) + thermals (18) + incoming solar radiation (77)=539-is all converted into IR. 200 (37%) goes up into space and 340 (63%) goes down to earth.Thinking about heat conduction from atmosphere to surface does not appear to be an issue, which makes sense, as most of the solar radiation is converted into heat at the surface, thus the surface is warmer than the atmosphere, and the thermal flow goes upward.

    I'm still not clear on why more IR goes down. I like DH's explanation in comment #4 of how the absorbing and re-emitting of IR equally up and down by parallel parcels of atmosphere sends more IR down than up. This reminds me of how I think about sound and waves and calculus in general. Except when I thought about it more the amplifying effect with the parcels of air seems symmetrical if one assumes uniform density and greenhouse gas distribution throughout the atmosphere, so just as much IR would get bounced up as down. But of course, thanks to gravity, the density is far from uniform.

    I'm guessing that because a lot more density and greenhouse gases are lower than higher, this allows us to think of the center of the atmosphere releasing its IR as lower to the ground. Then, if the IR released from the atmosphere is mainly going up, I could see how DH's parcels of air would amplify more IR back down than up.

    How is this sounding? I think part of the confusion was started by me first asking about heat rising or falling, when we are talking about infrared radiation, which is different than heat. I'd call it heat's electromagnetic messenger.
  15. Nov 12, 2015 #14
    I must have misunderstood the question, I'm just saying there's more ways for air to transfer heat to the Earth (and what happens to it there) than there is ways for the atmosphere to transfer it to space. My list of transfers wasn't meant to be exhaustive just to illustrate there's a lot more going on than just radiation.

    By the way that diagram you shared is pretty awesome! :D
  16. Nov 13, 2015 #15

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    That belief is incorrect. Look at the diagram I posted. The Earth's surface receives about the same amount of energy in the form of backscattered thermal radiation as the Earth as whole receives from the Sun at the top of the atmosphere. The atmosphere as a whole radiates considerably less energy than that into space.

    The only way the atmosphere as a whole could radiate the same amount of energy upward and downward would be for the atmosphere to be isothermal, which it is not. There's a significant temperature gradient in the atmosphere, a drop of about 6.5 degrees kelvin per in the troposphere.

    Don't think of it in terms of "the center of the atmosphere." The thermal radiation received by the surface of the Earth comes from the bottom of the atmosphere. The thermal radiation emitted by the Earth as a whole comes mostly from the upper reaches of the troposphere, plus a small contribution from the Earth's surface via the "atmospheric window", a band in the thermal infrared in which none of the greenhouse gases are active. It's the opacity of the atmosphere in the thermal infrared and the temperature gradient in the troposphere that are key to understanding this apparent paradox.
  17. Nov 13, 2015 #16
    The confusion is around the wording of emit. An individual greenhouse gas emitting a photon of IR emits it in a random direction. Half of the time, that would have a net upward direction, half the time a net downward direction.

    On average, and over the long term, there is a balance at the top of the atmosphere.

    In between, there is re-routing of energy. I think the difference in what we are saying is how the net effects end up. I would expect that the molecules at the earth surface that are emitting IR photons do not uniformly emit them upward, but in every direction including down. But there is a net upward. Any particular rock on a hillside might emit IR that warms the nearby ground. Or passes warmth inward. The rock molecules are fixed in orientation, and there may be a preferred direction, but the gas molecules are randomly oriented and moving.

    I think the difference in what we are talking about is just the words:

    "I believe the infrared radiation absorbed and emitted by the greenhouse gases goes half up into space and half down to earth."
    "I believe the infrared radiation absorbed and emitted by the greenhouse gases goes half up into space and half down to earth."

    If it is emitted up, and reflects off a cloud, and ends up "to earth", I want to say that is part of the half "up". If you are saying that CO2 self-orients such that emitted IR photons preferentially aim down, I am skeptical of that. I'm not questioning the specifics of the data from NASA, but why I think there was confusion when the ordinary thinking is that emitted photons are directionally random, from a randomly oriented moving molecule.

    I just realized that I might be reading into the original question, a question about individual photons and their direction. It doesn't exactly say that, now that I have copied it twice. This was (honestly) the moment I realized I might be adding confusion, by misinterpreting the question ... I really read it in a way that implied individual photons.

    I won't go back and correct any mistakes in any prior comments, that arise from my reading. I won't delete the stream of consciousness realization in this one. One can be confused by interchanging the words "radiation" and "photons" ... perhaps my own confusion might help anyone else.
  18. Nov 18, 2015 #17
    Hi everyone,
    I asked to question to NASA and this was Dr. Lin Chambers wrote back:
    Good questions. Have you found the Energy Budget Story slide set at:
    Though it does not address these specific questions, I think you will find it useful.
    In particular, all objects emit energy according to their temperature (see Slide 29). Since certain gases in the atmosphere are opaque to infrared radiation, IR energy does not transmit very far. So the atmosphere receives IR energy from the lower, warmer part of the atmosphere; while the emission to space is from the higher, colder part of the atmosphere.
    In between, you would see lots of local exchange between layers of the atmosphere, but this diagram is concerned with the net effect at the boundaries (Earth surface and space).
    While hot air does rise, electromagnetic energy goes in all directions – certainly energy from the Sun comes down to the surface. IR energy similarly can go in all directions.
    Have a look at that slide set, and then let us know what questions you still have. We may be able to clarify some things further."
  19. Nov 22, 2015 #18
    Something that puzzled me about this was the 340 watts per square meter. I've seen the solar constant expressed as 1366 watts per square meter. The formula for earth cross sectional area is pi*r^2. The surface area of a sphere is 4*pi*r^2. So the 340 is one fourth of the solar constant. I see that they are expressing the energy budget across the entire surface of the earth.
  20. Nov 22, 2015 #19
    Their accompanying slideshow (http://science-edu.larc.nasa.gov/energy_budget/pdf/EarthsEnergyBudget_Nov14_sm.pdf [Broken]) claims to show "earth's energy budget-A story". I also try to find and present the story of the greenhouse effect on my science arts website on climate change at: www.singingscientific.com.

    Here is a remaining question about this story. Where does each energy flow come from and where is it going? For example, where does the heat from solar radiation absorbed at the surface go? Does more than half go into latent heat (evaporation) and rest go into rising long wave infrared radiation (IR)? I'm guessing this because I am imagining the sunlight hitting the earth's surface, which is about 70% covered in water. Then again, I'm not imaging the greater back IR sent down to earth from the atmosphere. Maybe the different sources of energy flow mix and this question is unanswerable, un-measurable and meaningless?
    Last edited by a moderator: May 7, 2017
  21. Nov 23, 2015 #20

    Buzz Bloom

    User Avatar
    Gold Member

    Hi votingmachine:

    I think that it might be helpful to consider an idea that I learned from personal correspondence with a professor of atmospheric physics at MIT. Unfortunately I do not have the research skills and access to technical literature to find and cite any articles that discuss this idea. I would very much appreciate it if someone could do this for me, or alternatively to cite an article that refutes the idea.

    Here is the idea. IR photons absorbed and emitted by atmospheric gasses do not then all pass through the atmosphere ending up either hitting the earth surface or going into outer space. Many (most?) of them will hit another molecule of gas. To understand the implication of this, it will first be useful to point out that these photons (almost) never affect the temperature of the atmosphere. The evidence for this is seen in the article cited below.
    IR radiation was measured from the atmosphere over a ten year period, and one observation was that the spectrum of this radiation was not thermal. but rather matched the absorption spectrum of CO2 and other GH gasses. (That was not the primary observation, which was the correlation of increases in the measured radiation with the increases of atmospheric CO2.)

    What happens when a GH gas absorbs a IR photon is that a photon of the same energy is quickly re-emitted in a random direction before the molecule interacts with another molecule. If the molecule was near enough to other molecules so that it would interact with one before the re-emitting event occurred, then the excited energy of the molecule could be converted to kinetic energy, and by later further molecule to molecule interactions, the energy of the original absorbed photon would contribute to an increased equilibrium at a higher temperature In the Earth's atmosphere. The only molecules sufficiently close together for an molecule to molecule interaction to occur (with any significant frequency) before re-emission are in the liquid water droplets in clouds.

    In the context presented above, what happens when an IR photon is absorbed by a GH gas molecule in a gaseous environment (not a water droplet) is a random walk: a chain of absorb, re-emit, absorb,re-emit, etc., steps until one of two final events occurs: (1) a re-emitted photon hits the surface of the Earth, or (2) a re-emitted photon escapes to space. The length of each "step" in the random walk towards or away from the Earth's surface is a combination of (a) the distribution of mean free path between GH gas molecules for the interaction radius of the molecule, and (b) the distribution the cosine of the random emission angle with respect to the downward direction. The (a) component depends on altitude since the density of the GH molecules decreases with altitude.

    When I first thought about this scenario, I overlooked the altitude effect, and concluded that the probabilities for the random walk to end at the Earth or in space were equal. I plan to now do a Monte-Carlo calculation for a random walk after I figure out how to calculate the exact way that altitude will effect the step length. I will have to learn how to calculate the mean free path length. Can anyone suggest how to do that?

    Last edited: Nov 23, 2015
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook