Guest post by Ira Glickstein
The Atmospheric “greenhouse effect” has been analogized to a blanket that insulates the Sun-warmed Earth and slows the rate of heat transmission, thus increasing mean temperatures above what they would be absent “greenhouse gases” (GHGs). Perhaps a better analogy would be an electric blanket that, in addition to its insulating properties, also emits thermal radiation both down and up. A real greenhouse primarily restricts heat escape by preventing convection while the “greenhouse effect” heats the Earth because GHGs absorb outgoing radiative energy and re-emit some of it back towards Earth.
Many thanks to Dave Springer and Jim Folkerts who, in comments to my previous posting Atmospheric Windows, provided links to emission graphs and a textbook “A First Course in Atmospheric Radiation” by Grant Petty, Sundog Publishing Company.
Description of graphic (from bottom to top):
Earth Surface: Warmed by shortwave (~1/2μ) radiation from the Sun, the surface emits upward radiation in the ~7μ, ~10μ, and ~15μ regions of the longwave band. This radiation approximates a smooth “blackbody” curve that peaks at the wavelength corresponding to the surface temperature.
Bottom of the Atmosphere: On its way out to Space, the radiation encounters the Atmosphere, in particular the GHGs, which absorb and re-emit radiation in the ~7μ and ~15μ regions in all directions. Most of the ~10μ radiation is allowed to pass through.
The lower violet/purple curve (adapted from figure 8.1 in Petty and based on measurements from the Tropical Pacific looking UP) indicates how the bottom of the Atmosphere re-emits selected portions back down towards the surface of the Earth. The dashed line represents a “blackbody” curve characteristic of 300ºK (equivalent to 27ºC or 80ºF). Note how the ~7μ and ~15μ regions approximate that curve, while much of the ~10μ region is not re-emitted downward.
“Greenhouse Gases”: The reason for the shape of the downwelling radiation curve is clear when we look at the absorption spectra for the most important GHGs: H2O, H2O, H2O, … H2O, and CO2. (I’ve included multiple H2O’s because water vapor, particularly in the tropical latitudes, is many times more prevalent than carbon dioxide.)
Note that H2O absorbs at up to 100% in the ~7μ region. H2O also absorbs strongly in the ~15μ region, particularly above 20μ, where it reaches 100%. CO2 absorbs at up to 100% in the ~15μ region.
Neither H2O nor CO2 absorb strongly in the ~10μ region.
Since gases tend to re-emit most strongly at the same wavelength region where they absorb, the ~7μ and ~15μ are well-represented, while the ~10μ region is weaker.
Top of the Atmosphere: The upper violet/purple curve (adapted from figure 6.6 in Petty and based on satellite measurements from the Tropical Pacific looking DOWN) indicates how the top of the Atmosphere passes certain portions of radiation from the surface of the Earth out to Space and re-emits selected portions up towards Space. The dashed line represents a “blackbody” curve characteristic of 300ºK. Note that much of the ~10μ region approximates a 295ºK curve while the ~7μ region approximates a cooler 260ºK curve. The ~15μ region is more complicated. Part of it, from about 17μ and up approximates a 260ºK or 270ºK curve, but the region from about 14μ to 17μ has had quite a big bite taken out of it. Note how this bite corresponds roughly with the CO2 absorption spectrum.
What Does This All Mean in Plain Language?
Well, if a piece of blueberry pie has gone missing, and little Johnny has blueberry juice dripping from his mouth and chin, and that is pretty good circumstantial evidence of who took it.
Clearly, the GHGs in the Atmosphere are responsible. H2O has taken its toll in the ~7μ and ~15μ regions, while CO2 has taken its bite in its special part of the ~15μ region. Radiation in the ~10μ region has taken a pretty-much free pass through the Atmosphere.
The top of the Atmosphere curve is mostly due to the lapse rate, where higher levels of the Atmosphere tend to be cooler. The ~10μ region is warmer because it is a view of the surface radiation of the Earth through an almost transparent window. The ~7μ and 15μ regions are cooler because they are radiated from closer to the top of the Atmosphere. The CO2 bite portion of the curve is still cooler because CO2 tends to be better represented at higher altitudes than H2O which is more prevalent towards the bottom.
That is a good explanation, as far as it goes. However, it seems there is something else going on. The ~7μ and ~15μ radiation emitted from the bottom of the Atmosphere is absorbed by the Earth, further warming it, and the Earth, approximating a “blackbody”, re-emits them at a variety of wavelengths, including ~10μ. This additional ~10μ radiation gets a nearly free pass through the Atmosphere and heads out towards Space, which explains why it is better represented in the top of the Atmosphere curve. In addition, some of the radiation due to collisions of energized H2O and CO2 molecules with each other and the N2 (nitrogen), O2 (oxygen) and trace gases, may produce radiation in the ~10μ region which similarly makes its way out to Space without being re-absorbed.
There is less ~15μ radiation emitted from the top of the Atmosphere than entered it from the bottom because some of the ~15μ radiation is transformed into ~10μ radiation during the process of absorption and re-emission by GHGs in the atmosphere and longwave radiation absorbed and re-emitted by the surface of the Earth.
Source Material
My graphic is adapted from two curves from Petty. For clearer presentation, I smoothed them and flipped them horizontally, so wavelength would increase from left to right, as in the diagrams in my previous topics in this series. (Physical Analogy and Atmospheric Windows.)
Here they are in their original form, where the inverse of wavelength (called “wavenumber”) increases from left to right.
Source for the upper section of my graphic.
Top of the Atmosphere from Satellite Over Tropical Pacific.
[Caption from Petty: Fig. 6.6: Example of an actual infrared emission spectrum observed by the Nimbus 4 satellite over a point in the tropical Pacific Ocean. Dashed curves represent blackbody radiances at the indicated temperatures in Kelvin. (IRIS data courtesy of the Goddard EOS Distributed Active Archive Center (DAAC) and instrument team leader Dr. Rudolf A. Hanel.)]
Source for the lower section of my graphic.
Bottom of the Atmosphere from Surface of Tropical Pacific (and, lower curve, from Alaska).
[Caption from Petty: Fig. 8.1 Two examples of measured atmospheric emission spectra as seen from ground level looking up. Planck function curves corresponding to the approximate surface temperature in each case are superimposed (dashed lines). (Data courtesy of Robert Knutson, Space Science and Engineering Center, University of Wisconsin-Madison.)]
The figures originally cited by Dave Springer and Tim Folkerts are based on measurements taken in the Arctic, where there is far less water vapor in the Atmosphere.
[Fig. 8.2 from Petty] (a) Top of the Atmosphere from 20km and (b) Bottom of the Atmosphere from surface in the Arctic. Note that this is similar to the Tropical Pacific, at temperatures that are about 30ºK to 40ºK cooler. The CO2 bite is more well-defined. Also, the bite in the 9.5μ to 10μ area is more apparent. That bite is due to O2 and O3 absorption spectra.
Concluding Comments
This and my previous two postings in this series Physical Analogy and Atmospheric Windows address ONLY the radiative exchange of energy. Other aspects that control the temperature range at the surface of the Earth are at least as important and they include convection (winds, storms, etc.) and precipitation (clouds, rain, snow, etc.) that transfer a great deal of energy from the surface to the higher levels of the Atmosphere.
For those who may have missed my previous posting, here is my Sunlight Energy In = Thermal Energy Out animated graphic that depicts the Atmospheric “greenhouse effect” process in a simlified form.
I plan to do a subsequent posting that looks into the violet and blue boxes in the above graphic and provides insight into the process the photons and molecules go through.
I am sure WUWT readers will find issues with my Emissions Spectra description and graphics. I encourage each of you to make comments, all of which I will read, and some to which I will respond, most likely learning a great deal from you in the process. However, please consider that the main point of this posting, like the previous ones in this series, is to give insight to those WUWT readers, who, like Einstein (and me :^) need a graphic visual before they understand and really accept any mathematical abstraction.
Will says:
March 10, 2011 at 12:52 pm
Phil. says:
March 10, 2011 at 11:16 am
Yes I can read wikipedia also. http://en.wikipedia.org/wiki/Bond_albedo
from which you have clearly quoted verbatim.
The fact is that again you are either missing the point or you are deliberately taking things out of context.
Harry is talking about ALBEDO, you are talking about BOND ALBEDO. You have shifted the goal posts.
ALBEDO in context to Harry’s point is only relevant to visible light. Also from wikipedia, Quote: “When quoted unqualified, it usually refers to some appropriate average across the spectrum of visible light.”
But I referred to Bond albedo and you said in reference to that that albedo only referred to visible which is not relevant to my correct statement.
How can the ratio of the temperatures of Earth and Venus only depend on their distance from the sun when the albedo (bond albedo used correctly as I did before) is so different? Harry’s original statement was wrong.
Thanks, Richard.
Electric blanket I knew the analogy would raise some questions, because an electric blanket works primarily by taking an external power source and radiating heat both upwards and downwards. During the daytime, the Atmosphere does have an external source, as it absorbs some of the shortwave radiation energy from the Sun, due to the Oxygen/Ozone 100% absorption in the UV (0.2μ to 0.3μ) as well as the strong water vapor absorption in the near-IR (scattered wavelength areas between 0.8μ to 2μ). Water droplets (fog and clouds) also absorb some Sunlight directly. Thus, at least during the daytime and early evening, there is an external source of thermal energy for the Atmospheric electric blanket that distinguishes it from a purely passive blanket.
But, the main reason I mentioned the electric blanket is that, for the first time, I realized how it really works. It has long bothered me that our electric blanket is above both our top sheet and a regular blanket, which means half its heat is lost going upwards, and the rest has to go through two layers of insulation to get down to our skin to warm it. Though it is quite obvious, it only recently dawned on me that the downward heat is in the form of radiation, which passes through the top sheet and blanket relatively freely.
Bottom line on the electric blanket analogy: A regular blanket works primarily because it limits air conduction/convection and only secondarily because it warms up and thus re-radiates some of our body heat back down. The GHGs in the Atmosphere work by absorbing and re-radiating longwave energy, both down towards the Earth and out to Space, which reminds me more of an electric blanket.
Smooth, approximate “blackbody” curve: Yes, surface temperatures do vary considerably, but, if you look at the dashed lines in the Petty Fig. 8.1 plot above, you will see that ther is not that much difference between the 300K (peaks around 17μ) and 245K (peaks around 21μ) lines, other than their amplitude. Both are smooth curves that cover the ~7μ, ~10μ, and ~15μ regions.
When the Earth is heated by shortwave or longwave radiation, the surface absorbs and re-radiates that energy as thermal radiation in a relatively smooth curve that approximates a “blackbody”, with characteristic contributions in the ~7μ, ~10μ, and ~15μ regions. Since the downwelling radiation from the Atmosphere is not a “blackbody”-like smooth curve, because it has less representation in the ~10μ region, the transaction of energy from the Atmosphere to the surface, and back, results in a relative amplification of the ~10μ region. The effect of this relative amplification is shown in the satellite-observed radiation from the top of the Atmosphere where the ~10μ region is relatively over-represented.
mkelly says:
March 10, 2011 at 12:24 pm
Phil. says:
March 10, 2011 at 10:53 am
If I said surface I misspoke as I know the air isn’t going to heat the surface. My point was and is that -33 C which is used as the starting point for GHG effect may not be correct and that 0 C (STP) is a more proper place to start with any effect of GHG’s.
As I told you before STP is an arbitrary point with no physical significance at all.
Again with the angry. I never said I had all the answers about gas physics I said I was asking questions and there was a difference of opinion.
And I told you that facts are not something you can have opinions about, perhaps you should start listening to the answers to those questions.
But now I know that, “At any given temperature they will not emit more that given by S-B at that wavelength. ”
So does this say I can use Wein Displacement Law for a temperature find the wavelength and figure the energy being emitted?
At any temperature a distribution of wavelengths will be emitted (S-B distribution), Wien’s Law will tell you the the wavelength of the peak of the distribution. It will give a different value if applied to the spectrum in the frequency domain, a cause of confusion above.
Glickstein’s analogy of the “greenhouse effect” to an ELECTRIC blanket is totally incorrect. Electric blankets operate on an INDEPENDENT source of energy to produce additional heat. In stark contrast, GHGs can only operate on the energy thermalized mainly at the surface, producing none on their own. It’s sad to see such basic physical confusion being perpetuated here on WUWT.
Massimo PORZIO says:
March 10, 2011 at 1:24 pm
“What I’m arguing is that the increased absorption pit at 15um due to the doubling of the CO2 seen at the satellite nadir view doesn’t mean that the radiation is backscattered or held by the atmosphere, that energy just exits the atmosphere under different angles not seen by the satellites “eye”.”
Over the arctic looking down from 20km the bottom of the pit conforms perfectly to a 225K blackbody curve while outside the pit the curve fits perfectly to a 265K curve. Clear arctic air is very dry so water vapor has a minimal effect. 265K is -8C which is the surface temperature of the ice. Yet in the 15um window the ice appears to be -48C. What we are seeing in the 15um window is the so-called emission altitude. Dry adiabatic lapse rate is 1 Kelvin per 100 meters. The emission altitude is 4000 meters. This is the height at which the CO2 in the atmosphere has absorbed all the 15um radiation available upwelling from the ground and what remains is thermalized radiation from the atmosphere at that height.
Were we to reduce the 265K curve to a temperature such that the total reduction in volume underneath the lowered curve is the same volume as the hole then we would have the equilibrium surface temperature absent ALL atmospheric CO2. I believe that works out to about 15C. The thing of it is that the first 100ppm or so of CO2 does the lion’s share of the work so by the time we get to the surface equilibrium temperature difference between 280ppm and 560ppm CO2 (a doubling) the increase is only about 1.0C.
This takes us round to my hypothesis that the MOST IMPORTANT climate function of CO2 is in raising the average surface temperature of the earth from -23C (no greenhouse gases – the average surface temperature of the moon) to -8C. At -23C the earth is a snowball and positive feedback from the exceedingly high albedo of ice keeps it that way. Once CO2 is in the atmosphere to help melt some snow and ice this lowers the albedo dramatically where it matters the most (at low latitudes where the sun is the strongest) and additionally starts pumping water vapor into the atmosphere from the liquid water surface which accelerates the melt and we are back to a water planet instead of an ice planet. Once we have a water planet the low albedo of the global ocean and the higher absolute humidity keeps it a water planet (barely, with a few excursions to a snowball once every few hundred million years).
So CO2 serves as “kindling” to light the fire which melts the ice and turns what would otherwise be a frozen world into a liquid water planet. The water cycle, once activated takes over the show. And water limits the maximum temperature through negative feedbacks of evaporation and cloud formation.
During most of the earth’s history CO2 has been up around 2000ppm. The current 380ppm is dangerously low which probably has a lot to do with why we’ve been in an ice age for the past 3 million years.
Despite even 2000ppm CO2 the earth has still entered ice ages once in while and a few times severe enough to freeze the entire planet. It’s a bit of a puzzle what could possibly kick the earth out of a snowball episode but after about 10 million years it does indeed melt. The most popular hypothesis for the melt is that when the earth is frozen over all the normal CO2 sinks are gone. The global ocean doesn’t absorb any CO2 and there are no forests that grow on glaciers taking in CO2 and no chemistry between CO2 and ice forming carbonate compounds. But volcanoes don’t stop belching out CO2 so over the course of millions of years atmospheric CO2 rises and rises and rises until its greenhouse effect starts driving back the ice at an accelerating rate and viola – a water world is born again and the water cycle caps the maximum temperature at a point nice and comfy for living things from pole to pole.
Bomber_the_Cat says:
March 10, 2011 at 12:01 pm
I recently voted for this as being the best scientific blog but, in response to this article, we still get people who say that back radiation cannot happen or wish to divert the discussion to some strange theories about the temperature on Venus.
But there is a serious problem with Ira’s article!
As expressed before (6:10AM); his blackbody curves are showing a peak at about 18 micron when they should be peaking at about 10 micron for a 300K blackbody. So, the graphs don’t make sense
Now Joel Heinrich at 2.57 AM and Phil at 8.21AM, March 10, say this is because Ira has converted wave number to wavelength.
But this does NOT cut the mustard. This is a perfectly acceptable thing to do, albeit ending up with a non-linear scale rather than a linear scale. In fact, the
Source Data itself includes the wavelength scale, so it is nothing to do with Ira’s conversion.
On the ‘wave number’ scale the peak of the blackbody curves should be around a wave number of 1000 – which they are not! They peak around a wave number of 600 – which is WRONG!
So the whole article and its conclusions are WRONG!
Now, can any mathematician or physicist please explain to me why I am mistaken? What am I missing?
What you have in your head is the spectral radiance wrt wavelength whereas what’s plotted in Petty’s paper is spectral radiance wrt wavenumber, as I said above you can’t linearly transform them and they don’t have the same shape nor a maximum in the same place. Both are correct they’re just showing the data in a different form, as George has explained above chemists tend to prefer one form, physicists another.
If you use Wien’s Law to determine the maximum there are two forms, a frequency form and a wavelength form, the peak frequency does not correspond to the peak wavelength using c=λν.
Thanks Cal! You have correctly interpreted my intentions – many of my more reasonable environmentalist friends are losing faith in the runaway …. tipping-point … so-called scientific consensus. However, they are hesitant to join a Skeptic group that appears to them to be bending over backwards in an opposite and equally un-scientific direction. Only a minority of Skeptics fit that description, but our Warmist antagonists use these unscientific statements to push their side.
My consensus position as a Lukewarmer-Skeptic, which I believe is the most tenable from a scientific viewpoint given our current state of knowledge, is that, yes Virginia, there is AGW – there has been a net warming trend over the past century or so, and human activities have contributed a bit to it, but it is not any kind of crisis and may actually turn out to be beneficial.
Conversion of ~7μ and ~15μ thermal energy to ~10μ happens in the emission-absorption-re-emission transactions between the Atmosphere and the surface, where, in the upward direction, all three regions are represented more or less according to a smooth “blackbody” curve. However, as all the above graphs show, ~7μ and ~15μ are relatively over-represented in the downward direction. That disparity results in a relative amplification in the ~10μ region. That this is true, desipite all the other effects (convection, precipitation, etc.) , is shown in the graphs of radiation to Space, where ~10μ is relatively over-represented, compared to a smooth “blackbody” curve.
Please let me know if I am mistaken, but it seems to me that the blackbody radiation-spectrum gives an upper limit on downward emmissions from the atmosphere. I understand that if the atmosphere gave such a spectrum, that would would imply that it actually absorbed everything and re-emitted isotropically.
It looks like where CO2’s absorbtion and emmission are significant, the blackbody-spectrum has already just about been reached. Does this imply the small difference between current emmissions and those of a blackbody gives a low upper limit on the potential effect of additional CO2 on the temperature?
If only there was a FAQ for some of these issues! Then we could refer people to previous discussion rather than rehashing all the discussions like:
* What is “heat”? (confusion abounds about how “heat” relates to energy flow, energy content, temperature, of “thermal IR”. As long as different people use different definitions, people will continue to talk past each other.)
* Can energy flow from cool objects to warm objects? (yes!)
* Can NET energy (heat, as defined in thermodynamics) flow from cool objects to warm objects? (no!)
* Is “the greenhouse effect” the same as how a real greenhouse works? (not at all, so any arguments based on such analogies are pretty much useless from the start.)
* Do GHGs “heat” the ground? (no, by the definition of “heat” above) Do they “radiate IR energy” to the ground? (yes) Do they “increase they temperature” of the ground? (definitely, compared to an atmosphere with no GHGs) Do they “warm” the ground? (as I would use the word, my answer is yes)
What else would need to go in that FAQ?
Thanks for your comment, Phil. The Petty plots have both wavenumber and wavelength, and I preserved his wavelengths in my simplified curves when I reversed the plots horizontally. The reason I think in wavelength is that my experience with IR is based on near-IR 1.06μ laser rangefinders and far-IR ~10μ Forward-Looking Infrared video sensors. However, the conversion between the two measures is simple:
Wavenumber = 10,000/Wavelength
The reason for the difference in peaks has nothing to do with the use of Wavenumber or Wavelength.
As Phil noted, Petty plots Radiance.
My earlier posting plots Spectral Intensity. (It is from http://en.wikipedia.org/wiki/File:Atmospheric_Transmission.png.
From Wikipedia:
The units of the spectral intensity are energy/time/area/solid angle/frequency. In MKS units this would be W·m-2·sr-1·Hz-1 (watts per square-metre-steradian-hertz).
The SI unit of radiance is watts per steradian per square meter (W·sr−1·m−2).
Notice the difference? That explains why the peaks are different and why the longer wavelengths plot higher in Radiance than they do in Spectral Intensity, where they are normalized.
Dave Springer says:
March 10, 2011 at 4:30 pm
“So CO2 serves as “kindling” to light the fire which melts the ice and turns what would otherwise be a frozen world into a liquid water planet.”
Don’t have time for a lengthy discussion, but the thermal mass of CO2 is way too small for it to play the “kindling” role that you suggest. A CO2-free Earth might be slightly cooler, but would not be frozen, because convection would nevertheless heat the base of the atmosphere to similar temperatures, thereby sharply reducing the radiative loss from the surface. It’s the PRESENCE of a dense atmosphere, rather than its chemical composition, that matters most thermodynamically. This is entirely lost in treatments that look only at radiative heating of the atmosphere.
when it is too hot; gall bladders just do not know hot and cold.
George E. Smith says:
March 10, 2011 at 3:46 pm
“Guess how a lot of people who are dumb enough to try and climb K2, end up dying up there. No they don’t smash their limited brain capacity on a glacier 4,000 feet below them; nor do they freeze their arse off at -40C/F deg.
They COOK their insides (gall bladder included). That coat (which of course couldn’t possibly warm them), stops their skin from freezing or even goose bumping, so their skin cannot get rid of all the excess internal heat energy generated from the decay of hot dogs, and drinking too mcuh 5-hour energy.”
——————————–
This must be why the Canadian woods are filled with people running around harvesting bear gallbladders from the poor creatures that got cooked from the inside.
John Marshall says:
March 10, 2011 at 2:04 am
I keep hearing this. It gets boring.
Cool Body radiates x in all directions, including towards Warm Body. Warm Body radiates 2x in all directions, including towards Cool Body.
Warm Body cools more slowly than if Cold Body were not present
Overall heat transfer is not from Cool Body to Warm Body, so that 2nd law has not been broken, and nobody (pun intended) goes to Thermodynamic Jail.
CommieBob,
I am not asserting that a cold object cannot radiate EMR to a warm body. Only that the net flow of energy will only go from a hot to cold.
Since you are an EMR expert (no sarc) , I do have a question that I have been wrestling around with, but have been unable to find an answer.
Any other EMR experts please feel free to chime in. None of my texts or classwork deal with this (at least through grad school), as it is a combination of thermo, general relativity, and quantum mechanics.
When an object drops from an excited state and emits a photon, that photon is then absorbed somewhere else in the universe by another object. To that photon, all distances are zero, time does not exist, and the photon transfers its information instantly. The photon doesn’t exist between between points (GR). Is it possible for that individual photon to transfer information from a less excited molecule (or electron or whatever) to one more excited? The answer isn’t the obvious yes that I thought initially. In fact, blackbody radiation would still work form a cold source to a warm source because of all the random motion and complete statistical range of excitation levels of the particles. I know there are some super smart people trying to tie in thermo with QM, QED, QLG, so it would seem if it ties in, then the photon would only be able to transmit information from an excited state to one less excited.
In other words, since this is all happening instantly from the perspective of the photon, can a photon “tell” an excited object to become more excited as it “tells” the less excited object to become even less excited. If it is all based on probabilities of states (QM), then the answer should be yes, even though this goes against the grain of the Thermo underpinnings.
Thanks!
RJ says:
March 10, 2011 at 4:17 am
NO. They would just cool down more slowly, so the overall temperature of the human would increase slightly until equilibrium (energy gained = energy lost) is reached. This assumes they are living (which is not a given in a container of CO2, BTW), and is generating a constant amount of energy from that living process.
I really fail to see why something so obvious is so hard to get. When you add a blanket at night, do you ‘cook’? No. Your temperature just increases until equilibrium (energy gained = energy lost) is reached. Try it!
Jer0me,
OK then, let’s see you falsify this.
☺
Joe Lalonde says:
March 10, 2011 at 4:49 am
I think a lot of convection is due to differing temperatures. This is why you get a lot of wind when clouds come over on a sunny day. Also why the wind (assuming few clouds) goes toward the sea (which is warmer) in the first part of the day, and way in the latter part. Simple temp.
No. 1. there is no such thing as centrifugal force. It is simple inertia. Centrifugal force, does exist, however, and that is what stops everything flying off. In this case, gravity. The rest of the concept you put forward makes no sense to me.
Well, between the tropics, anyway. It differs through the year.
That is a false assumption. Just because sunlight is shown does not mean the rest of it stops when the sun goes down.
This might be too simplistic… but what I would like to see are some specta showing the measured (not calculated or theorized) downward radiation from the sun as measured at the TOP of the atmosphere. I find it amazing that the sun emits so little in the IR wavelengths.
Anyway, if there are good measured downward spectra at the top of the atmosphere, and good measured downward spectra at the bottom of the atmosphere, then the difference between the two should clearly show what’s being re-emitted (downward) by the atmosphere… shouldn’t it?
So… what do the data show?
Back radiation is like what a diver sees when he is on a wreck 100 feet down with visibility of only 50 feet. He can’t see the surface but he can certainly see. Its like a blue glow that comes from all around. If we could see infrared and not visible, thats what we would see.
The energy is still there and keeps on traveling. The stationary null that you describe is matched by a stationary peak somewhere else.
What you describe is similar to a standing wave in a cable. You get peaks and nulls. As you travel along the cable, the RF voltage varies in a predictable manner. The two waves continue to exist and travel along the cable and get to their respective ends where they reflect or are absorbed. http://en.wikipedia.org/wiki/Standing_wave
In three dimensional space, you get a predictable pattern. If I arrange two antennas wrt space and phase, I will get a particular pattern in the resulting field. No energy is created or destroyed. It is, however, redistributed. In the case of two antennas the pattern is basically based on hyperbolas. In other words, if you plot the nulls (or peaks, but nulls are easier to measure) you get hyperbolas.
“”””” Ira Glickstein, PhD says:
March 10, 2011 at 5:21 pm
Phil. says:
March 10, 2011 at 8:21 am
… the Petty data is spectral radiance (mW/m^2.sr.cm-1) plotted vs. cm-1 and is correct, the transformation to a plot in terms of wavelength is non-linear (and is why it’s misleading for Ira to have reversed the axis on his plots, should leave it in wavenumbers).
Thanks for your comment, Phil. The Petty plots have both wavenumber and wavelength, and I preserved his wavelengths in my simplified curves when I reversed the plots horizontally. The reason I think in wavelength is that my experience with IR is based on near-IR 1.06μ laser rangefinders and far-IR ~10μ Forward-Looking Infrared video sensors. However, the conversion between the two measures is simple:
Wavenumber = 10,000/Wavelength
The reason for the difference in peaks has nothing to do with the use of Wavenumber or Wavelength.
As Phil noted, Petty plots Radiance.
My earlier posting plots Spectral Intensity. (It is from http://en.wikipedia.org/wiki/File:Atmospheric_Transmission.png.
From Wikipedia:
The units of the spectral intensity are energy/time/area/solid angle/frequency. In MKS units this would be W·m-2·sr-1·Hz-1 (watts per square-metre-steradian-hertz). “”””
Notwithstanding ANYthing that you may find on wikipiddle, the units of “spectral Intensity” ARE NOT energy/time/area/solid angle/frequency.
“INTENSITY” (fancy word for candlepower) has NO MEANING for anything but a POINT SOURCE; so with /area/ in there the intensity (spectral or not) would be infinite.
The /area/ for a source of non zero area, converts Intensity (spectral or ot) into RADIANCE ore STEARANCE which some snobs like to use.
)
As to the per frequency for the “spectral Intensity” or “spectral Radiance”, it could be either per frequency (or wavenumber, or per wavelength; so long as it is stated; either /cm^-1 or /micron.
I like to use per wavelength; because I much prefer to use a normalized Black Body Radiation plot, which is totally universal; and acknowledges that the Planck Radiation Function is a function of the single independent variable lambda.T which immediately points to the Wien Displacement Law.
So a normalized plot plots on a horizontal scale (usually logarithmic) whose variable is (lambda.T)/|lambda.T|max; and the Y-axis, is spectral radiance/ |spectral radiance|max and usually is plotted on both a linear or a logarithmic vertical scale.
So the BB curve typically goes from say (0.1) t0 maybe (100) in X and (0) to (1) in Y or 1.0 down to 10^-5 on the logarithmic plot.
An excellent graph can be found in “Modern Optical Engineering” by the late Warren J Smith, who was for years with Infra-Red Industries in Santa Barbara California.
Now yes you can reformulate all of this in terms of frequency or wavenumbers; and then Wien’s Displacement law becomes T/f is constant.
And yes a frequency based plot acknowledges the Einstein relationship E = h.nu; but as I have already noted, neither a wavelength or wave number graph is very intuitive (on the spectral interval) since one form favors the UV, and the other favors the far IR, as far as viually indicating relative importance. There are simply huge wave number ranges associated with those miserable looking spectral radiance values in the H2O bands; but the very prominent looking CO2 band (in the wave number pictorial) has a measly 100 cm^-1 for spectral bandwidth.
Note that 25% (almost exactly) of BB radiation occurs at shorter than the spectral peak Wavelength on a per micron spectral radiance plot; I have no idea how much lies below the peak on a per wave number plot; but you can easily work it out.
As I said some people like their BB plots in numbers of Photons, not in direct energy or power units.
But nyet on the spectral Intensity; the earth surface or atmosphere is not a point source.
For the pedantic, the actual IRRADIANCE on a surface due to a finite area source (on axis) differs from that due to a perfect point source by less than 1% for a source to surface distance that is ten times the source diameter.
Also there is no such thing as a point source of anything (physical). For EM radiation the boundary conditions for Maxwell’s Equations cannot be satisfied for a point source; and for the Quantum Physicists, a point source by virtue of the Heisenberg uncertainty principle must have an infinite spectral bandwidth; so the energy or power in any finite wavelength or frequency range would be zero so it would be undetectable at any wavelength or frequency.
I don’t have a problem with Ira’s use of both the wavenumber and wave length horizontal scale, as Ira said, that makes no difference, except as Phil pointed out, the wavelength conversion scale from wave number would be non-linear. But just keep the per wave number or per micron units straight in the sepctral Radiance units, and then everything is kosher.
Peter Humbug solved that dilemma by simply leaving out the per frequency altogether; but I’ll give him a typo yellow card on that.
Fred Souder says:
March 10, 2011 at 6:06 pm
Is it possible for that individual photon to transfer information from a less excited molecule (or electron or whatever) to one more excited?
Yes provided that hν for the photon matched the energy difference between the two higher energy levels (in a Q branch for example).
Ira
Thanks for your further comments. I now understand why you suggested an electric blanket as analogy. However, I remain unconvinced that it is in fact a superior analogy. In fact in some ways, I liked the comment by Alan McIntire March 10, 2011 at 6:08 am comparing the position with traffic flow over a section of road with accidents, road works whathaveyou. My response to that is that yes at peak times, there are snarl ups and the journey time is increased, however, late in the evenning and at night, there is less and less traffic indeed during the night no traffic such that all the cars eventually exit the highway and the road is completely free come morning. All the traffic has disippated akin to all the energy received by the Earth finally finding its way out to space so that there has been no effective build up of energy.
I digressed, reverting to your electric blanket. Lets for the sake of argument assume that you are right with respect to the power source (although I have concerns regarding the lack of independent power source), one of the problems I have is how this all works in the real world (as opposed to in isolation in a laboratory), by which I mean that other atmospheric processes are at play which may overwhelm the effective influence of effective heat transfer from the sky to the Earth. In particular, convection. Further, does the energy absorbed by say CO2 thermalise before it has had an opportunity to radiate.
Say that you normally make your bed by creating a sandwich of a sheet, electric blanket and ordinary blanket. When the electric blanket is plugged in, this keeps you nice and snug. Now consider that you have a 4 poster bed but without any curtains. Instead of your electric blanket being placed on top of the sheet, it is suspended at the top of the 4 poster frame say about 3 feet above the matress. You now only have the sheet and the blanket directly over you. When the electric blanket is switched on, the down radiated heat is not sufficient to warm you, nor does the down radiated heat sufficiently prevent heat loss being radiated from your body. With this set up, you are cold at night. Even in the stillness of your bedroom, convection overcomes you feeling any real benefit from the down radiated heat. I am not saying that the electric blanket has no effect, only that it is overcome by other processes.
The fact is that radiation is weak compared to convection. I have gas central heating running radiators. When the system is on, even if I place my hand within 1/2 inch of the face of the radiator (ie., place my hand parallel with the radiator close to the surface of the panel), I struggle to feel the heat. Yet, if I place my hand several feet above the radiator, I can still feel the heat. Of course, if I touch the radiators they are very hot to touch and unpleasant to keep your hand there for any length of time.
I must admit that I struggle to see that the amount of energy that the Earth is said to receive from back radiation is more than the energy it receives from the sun in circumstances where the source of that back radiation has ultimately all come from the sun (if one ignores geothermal energy being produced from the centre of the Earth). If that were true, on a sunny day, there would not be a substantial difference in temperature between the temperature in the sun and under a cloud. At night, our bodies could be warmed and bask in the radiant energy from the sky which would be safer than sun bathing, and the photo electric cell would not be tuned to capture sun light but rather research would be directed at tuning it to work on back radiation which is available 24 hours a day and whether it is sunny or not. It appears to me that the heat transfer from the Earth from the ground to lower atmosphere is dominated by other processes (particularly the water cycle) which completely overwhelm the effects of back radiation.
I am not saying that there is no ‘greenhouse’ effect (how I dislike that expression) but rather in the real world, the effect of back radiation may be considerably less than is speculated (and I use this word since there is no body of empirical observational data backing up the claims) and I consider that the effect of any change in CO2 concentratioons above present levels is likely to be extremely modest. Indeed, Miskolczi could well be right on how it all works.
PS. Whilst I do not necessarily agree with various points you make, I liked your post and it certainly has generated many interesting comments and views.
PPS. I think you should consider carefully whether you wish to run with the eklectric blanket since this in my opinion in many ways confues the issue. Would not a better analogy be a one way mirror which is losing some of its mirrored surface such that it allows most of the incoming sunlight to pass through but only allows some of the LWR emitted from the Earth to pass back through towards space and reflects some part of the LWR back down towards the Earth.
I really like Dr. Glickstein, but I think he’s way off base with his ideas about a cold, rarefied gas heating the earth’s surface. Trapping radiation? Really? You might delay it’s escape to space by a few microseconds, but storing heat energy? Back radiating? Unicorns?
Think about this. You have a heated bowling ball and a passive bowling ball. Suppose you could flip a switch and the make the passive ball instantly colder. What happens to the radiation between the balls?
A) Stays the same.
B) Increases.
C) Decreases.
richard verney says:
March 10, 2011 at 7:00 pm
Would not a better analogy be a one way mirror which is losing some of its mirrored surface such that it allows most of the incoming sunlight to pass through but only allows some of the LWR emitted from the Earth to pass back through towards space and reflects some part of the LWR back down towards the Earth.
I’ve used this as an example to show how the recycled IR can exceed the initial solar radiation.
Consider a blackbody illuminated by 300W/m^2 solar radiation via a dichroic mirror which passes 100% of the solar wavelengths but reflects 50% the IR (BB) incident on it. At equilibrium the temperature of the BB has increased so as to radiate 600W/m^2 of IR, 300W/m^2 of which is returned to the surface. If the reflectivity of the mirror is increased to 66.6% then the BB radiates 900W/m^2 of IR, 600W/m^2 of which is returned to the surface.