Guest Post by Willis Eschenbach
[SEE UPDATE AT THE END]
Due to the recent posts by Lord Monkton and Nick Stokes, I’ve been thinking about the relationship between radiation and temperature. So I turned to the CERES dataset. Here is a scatterplot of the monthly global average surface temperature versus the monthly global average downwelling total radiation absorbed by the surface. The total radiation is the sum of the net solar radiation at the surface and the downwelling longwave radiation at the surface. I’ve removed the seasonal variations from the data.
Note that 3.7 W/m2 is the increase in downwelling longwave radiation expected from a doubling of CO2 …
When I saw that, I thought well, maybe the increase is small because there’s a lag between the absorption of the radiation and the warming. To see if that was the case, I did a cross-correlation analysis of the relationship.
No lag visible.
Now, I get busted regularly for drawing what I’m told are the wrong conclusions from the data that I present. So I’m just gonna say …
Comments?
——————————————————————————————————————
Me, I’m writing this from banks of the Kenai River in Alaska, one of my favorite spots in the world. When I got off the airplane, the aroma of the air was absolutely intoxicating. Summertime is short here but the days are long, and the air is full of the heady perfume of every plant and every animal growing and going at triple speed, making the most of the brief Alaska summer. Here’s what the sun is doing today this far north …
[UPDATE] Someone asked what temperature I’m using. I used the conversion of the upwelling longwave radiation from the surface. However, the answer is only slightly different if I use, for example, the HadCRUT surface temperature. Here is that result:
As you can see, there is no significant difference when I use the other surface temperature dataset.
My very best regards to all, may your days be as full of sunshine as mine,
w.
PS—My usual request: when you comment, please quote the exact words you are responding to, so we can all be clear about who and what you are talking about.
PPS—Bonus question. What latitude on the planet gets the most hours of sunlight per year?
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
About: Graphic nr. 1: Scatterplot Total Radiation and Surface Temperatures
WR: I suppose the high ranked months in the right half of the graphic are El Nino months. The higher moisture degree over larger areas of the Pacific creates both the higher downwelling radiation and the higher surface temperatures.
If the above is correct, I suppose that we are looking at water vapor’s down welling radiation, caused by natural variations (like El Nino, La Nina) that influenced the total surface area with a high water content in the lower troposphere. The larger the warm ocean surface area, the larger the area with high water vapor in the lower atmosphere that is absorbing surface radiation (near the surface) and the larger the area that enhances downwelling radiation.
The combination of El Nino effects with the effects of a temporary warmer Arctic (less sea ice, more water vapor in the air) brings us our present relatively warm era.
Water vapor is by far the most important greenhouse gas in the lower atmosphere. Something that will diminish the content of water vapor just above the surface will bring down surface temperatures. For example: colder ocean surfaces, more ice in the Arctic, more wind over the oceans that enhances ocean mixing and oceanic upwelling etc..
Wim, always good to hear from you. I, like you, think we’re looking mostly at water, both as vapor and as clouds.
Regards,
w.
Is your microwave 750 W or 1000 W?
It’s good to see yet another refreshing approach from Willis to the boring old subject of Global warming. Fresh approaches wake you up again after weeks of repetitive and meaningless ad hominems. Yes, I think that 0.5 C looks like a fair answer for a doubling of CO2 (3.7 w/m2).
Neat. That suggests negative feedbacks at play.
Al Gore once predicted a sea rise of twenty feet by the end of this century.When n I saw this I decided to make my own prediction,. At the time 80 kilometers sea surface were available. It predicted clrear sea rise of just under ten inches by the end if the century, This is ridiculousI said and I wanted to inform the journals Nature Science of this fact Both journals tgrew out my result and Al hgot a Nobel Prize for his lie. Checking today chows that all predictions are too high. My original prediction is likely to be closest to truth in the end. Arno Arrak
Willis,
“Note that 3.7 W/m2 is the increase in downwelling longwave radiation expected from a doubling of CO2 …”
No, it’s not. The 3.7 W/m^2 for 2xCO2 is the instantaneous increase in (straight) upward IR optical thickness from the surface through to the TOA. Or it’s simply an increase in upward IR absorption by the atmosphere from the surface.
RW, thanks for that detail. I was also confused on exactly what the 3.7 w/m2 applied to. When I used a conversion tool I found the warming from that amount of energy was on .68 C. To get 1 C of warming I needed ~5.45 w/m2 at the Earth’s average temperature (~15 C). In the chart above this amount of forcing would lead to around .55 C of warming.
sorry, I’ll fix my earlier post where I messed up the blockquote tags.
Could you explain what you mean by “downwelling” ( an awful term to describe radiative transfer ) ?
Since this is CERES I assume you have incoming SW and incoming LW minus backscattered SW and TOA LW emissions. ie total energy input. Am I correct in recalling there are still outstanding calibration issues in the total energy input this gives which is far in excess of a credible value and would be heating us up at a rate we know is not happening?
That is between the atmosphere and the surface. Since figure you cite is that after reaching a supposed new equilibrium state, the TOA budget should not change.
I’m not sure that what you plot has any bearing on the 3.7 W/m2 figure.
You don’t say what the temperature is , so I assume it is also a CERES measurement. That probably means it is “clear sky” surface temp. That obviously introduces a geographic bias and probably you need to consider what that implies about the graph.
On the face of it, I don’t think that graph tells us anything. At least what it represents has not been explained.
I find the zero lag very surprising but that may be accounted for by the above geographic bias, also is this surface ( ie water ) or near surface (5km) air “temperature brightness?
IIRC Spencer & Braswell published something a bit like this based on CESES about 5y back.
Greg June 9, 2019 at 12:46 pm
Sure. “Downwelling” means headed towards the earth, and “upwelling” means headed towards space. Why are those clear names somehow “awful”?
There are calibration issues, but they affect the accuracy, not the precision … and for e.g trends all we need is precision.
Read the damn thread. I’ve explained it once, and another commenter explained it once. I’m not gonna hold your hand, do your homework before uncapping your electronic pen.
I think you’re trying to say “I don’t know what the graph means despite explanations” …
Yeah, that’s lots of help …
Sadly,
w.
Thanks Willis. [Pruned] as usual in the face of even mild criticism.
Radiation does not “well” up and down it radiates: the clue is in the name.
Wells are full of fluid not radiation and the modes and reasons of movement are totally different. That is why it is an awful term.
You really think I’m going to read through a hundred or so mostly irrelevant comments to find out if maybe someone commented of where you got temperature from. Nothing like your usual, humble self to say :”heck I did forget to put that in my article, I’ll add note at the end. Good catch.”
… or the “where did you get the data for you graph in the article” questions.
So do they measure “downwelling longwave” from a satellite. If you were half as smart as you think you are you’d have a better reply to discussions about what the graph and its slope indicate, instead of avoiding the questions I raised by snarky remarks.
” IIRC Spencer & Braswell published something a bit like this based on CERES about 5y back.”
Well I don’t have the ref. to hand and you are as good as anyone at using google. So instead of expecting me to do your homework for you , go find it if you are interested. You may have 2 or 3 papers from those two authors, I’ll sure you’ll manage.
Greg June 14, 2019 at 12:42 pm
Greg, the sciences are full of “terms of art”. These are words which are used in a specialized sense. Often, they don’t make exact sense. “Upwelling” and “downwelling” are good examples. So is the “greenhouse effect”.
Now, you are free to whine all week about how the greenhouse effect doesn’t really work like a real greenhouse, or how downwelling has nothing to do with a well. But those are the terms that are used, whether you like them or not.
You really think I care whether you are willing to read the thread? Do your homework or not, I couldn’t care less. It’s not my homework.
w.
Greg, upon further thought I wanted to add that if you’d said, “Willis, could you add an explanation of where you got your temperature data it would make the post much clearer”, I’d most likely have said “Sure, I’d be glad to.” I’ve answered dozens of requests in that manner.
As to you saying you didn’t read the thread, oftentimes the real nuggets are in the thread. I can’t tell you how much I’ve learned from the commenters on my work. So you are throwing away valuable information. You can skip the junk withough reading more than a line or two of each comment.
Just sayin’ … that’s how I learn things.
Regards,
w.
I seriously doubt its legit physics to add SW + LW and convert the total as a surface temperature.
So do I, Macha … who do you think is doing that?
w.
Well, in terms of using a delta-Ts versus delta total surface-energy (LW+SW)-flux-absorbed in order to obtain an correlation coefficient (i.e., the first graph presented in the above article), I think the answer is obvious.
I think it is significant that the derived least-squares (presumed) linear fit of the data produces an equation that has a “b” term, in the linear equation form y = mx +b, that is significantly offset from zero (by my calculation by -38 C). I cannot yet determine the significance of this, but is of concern. It may be related to the fact that radiation should be governed by a T^4 scaling relationship (i.e., not linear) and for small delta-T changes this reduces to basically a 4* dT influence factor (binomial expansion) instead of a linear scaling.
Deeper thinking is required.
Gordon, and Macha, I added LW + SW. However, I didn’t “convert the total to surface temperature”.
Best regards.
w.
Why? IF – big “IF” there! – you assume all inbound radiation must equal all outbound radiation, and a single albedo and emmissivity and a single coef of heat capacity and coef of heat transfer exists for a solid, uniformly radiated object in a vacuum, then the equilibrium temperature will be proportional to the total of the inbound radiations.
If in a barren desert with typical sand, a forest of artificial dark green plants were placed on the surface, will the surface temperature be higher or lower?
Replace the artificial plants with real ones and add water. Now what happens?
There’s a reason why deserts are hotter at the same latitude and altitude than a tropical environment.
Convection rules the day, not radiation.
Don’t forget effective albedo over any given area of Earth. Effective albedo is a combination of both cloud coverage and ground surface coverage (e.g., sand versus vegetation versus open ocean water).
Unlike convection, albedo directly controls the amount of solar energy absorbed by the atmosphere and ground over any area of Earth during daylight hours.
To all the greenhouse gas faithful. I reckon there’s a Nobel prize for the person who, uses traditional science – such as controlled experiment and rigorous, repeatable observation – to conclusively prove the greenhouse gas effect and refute the Laws of Thermodynamics. Why are none of you interested in collecting your Nobel?
All this proof by statistics, lampooning ones opponents, argument from authority, telling us scientific and social parables, goes only so far. There will also be a core of scientifically-minded people who laugh at you.
What does 3.7W/m² physically mean? Does it mean an extra 3.7W/m² of energy is trapped? How much of it warms the surface? How much is due to CO2, water vapour, clouds? Give us precise numbers here (or at least clear error bounds). You can’t expect us to pay $200 trillion, world-wide to rewire our energy systems, based on your faith.
NB: prove: write it out in a form which can be falsified by some experiment or observation but such falsification fails.
Mark, please purchase a modern college textbook on thermodynamics. Also consult the writings of “skeptical” physicists. You are beclowning yourself.
I have to wonder about your scientific background, Mark. Please enlighten us as to you qualifications to question the conclusions of different branches of science.
I’ve looked at modern college textbooks on climatology. One told me computer model runs are experiments. No thanks, I don’t need your post-normal science. I’ll be a little circumspect on mere college textbooks in future. I doubt you’ve read any books on thermodynamics; textbook or otherwise.
UN IPCC climate models based on unproven assumptions are bunk. The assumption of 3X water vapor magnification of an unmeasurable amount of CO2-driven warming is bunk.
The assumptions that one may calculate global warming based on minuscule alterations in earth energy flows is bunk. We are unable to measure the theoretically-calculated minuscule changes in those massive energy flows.
There are too many other continuously changing climate metrics to tease out the impact of a minor change in CO2 forcing.
Now, focus on those facts and leave the physicists alone.
BTW, my college coursework included a year of thermodynamics.
Willis, did you do a similar analysis a couple of years ago? Title: Temperature and Forcing, July 13, 2017. Does this new analysis supersede that?
Willis,
“Note that 3.7 W/m2 is the increase in downwelling longwave radiation expected from a doubling of CO2 …”
Where is this value coming from? Can you elaborate? In my physic books and its spectral calculations you are looking at about 10 Kelvin temperature increase in the CO2 core radiation band to get this increase of down welling radiation.
It’s not an increase in downwelling radiation. It is the areally averaged decrease in outgoing radiation at TOA obtained by flooding a model with double the concentration of atmospheric CO2 and applying the RTE across all gridded points.
Outgoing Radiation is INCREASED by increasing CO2.
The satellite measurements have already proved it.
see
https://www.youtube.com/watch?v=gIhBEF94YlM
about 15 minutes in.
Yes it will.
It is radiation emitted from the Thermosphere (100-200Km up).
Not the Troposphere.
The Thermosphere receives energy from UV and particles in the solar wind (not from Earth).
The presence of CO2 absorbs, then emits some of that energy in the form of LWIR.
The more CO2 there, then the more LWIR emitted to space and greater cooling of the Thermosphere.
I wrote TOA to try to keep things simple. I was focused on explaining the tools used for computing the forcing attributable to CO2. The true picture is as always more complicated. Normally “adjusted forcings” are calculated at top of troposphere after allowing the stratosphere to stabilise, but keeping all other atmospheric variables static. The stratosphere instantaneously cools in atmospheric models when CO2 is doubled. It emits more LW outwards because of the increased CO2 and takes less flux from below – so it cools. The outgoing net flux at top of troposphere is decreased. It is the decrease in this latter that is averaged after stabilisation of the stratosphere and called the adjusted forcing (Fa). This is subsequently modified into what is called the Effective Radiative Forcing (ERF) in a step which I am not going to attempt to explain and which I do not support. Typically, the instantaneous change in net TOA flux (pre-stabilisation of the stratosphere) is larger than the Fa which is larger than the ERF. Yes, the thermosphere which is almost 100kms above the elevation of interest should cool if you add CO2 and keep everything else constant.
JB, you ask where the 3.7 W/m2 per doubling comes from. I’ve never seen a proper engineering style well-buttressed derivation of that number. I can only say that it is the number that is always used.
I’ve also never seen any uncertainty estimate for the value … but like I said, everyone uses it.
w.
Smacks of an evaluation is on order to find out.
This is very unlike you to take this number at face value and not question it.
JB, I didn’t say I didn’t question it.
I said I’ve never gotten a full answer.
Thanks,
w.
Thanks for the lag chart!
Is it all from the Ceres dataset?
I wonder how a satellite can measure downwelling infrared?
An other point is how the ground temperature is measured. Is that by radiation, then you could have some conflicts.
There seems to be some missing w/m2, because 0.38K means only 1.8w/m2 ekstra radiation from earth. The rest must be evaporation i think.
Svend, see here for the answers to your excellent questions.
w.
Willis wrote: “Note that 3.7 W/m2 is the increase in downwelling longwave radiation expected from a doubling of CO2 …”
3.7 W/m2 is the average decrease in OLR predicted for an instantaneous doubling of CO2 by radiation transfer calculations. The increase in DLR is only about 1 W/m2.
The difference would be going into warming the atmosphere the atmosphere until the 3.7 W/m2 imbalance had been negated by increased emission of thermal IR by the warmer atmosphere.
Enjoy those long days.
Frank,
There is no increase in DLR from an instantaneous doubling of CO2.
I clarified what it was here in this post:
https://wattsupwiththat.com/2019/06/08/radiation-versus-temperature/#comment-2719828
But Willis never replied. He’s not using the 3.7 W/m^2 figure correctly here and doesn’t seem to know what it is.
RW: Those doubled CO2 molecules must be pretty smart to be able to effect OLR, but not DLR.
You can go to the online MODTRAN website and see for yourself that doubled CO2 increases DLR and decreases OLR. The increase in DLR from clear skies is bigger than the “about 1 W/m2” I cited. However, if you add low clouds, doubling CO2 produces almost no change since most of the DLR photons are emitted from cloud bottoms.
I believe an early Ramanathan paper calculated an increase of 0.9 W/m2.
If you want a rational, doubled CO2 means that the average DLR photon after doubling arriving at the surface has been emitted from a lower altitude where it is warmer. This is analogous to the average OLR photon escaping to space being emitted from higher after doubling.
Frank,
There is no increase in DLR applied to the system for the instantaneous 2xCO2 case. The increase in DLR is (has to be) subtracted from the OLR decrease to conserve energy for the instantaneous case or calculation. So what is applied to the system (and the system is responding to) is simply an OLR decrease of 3.7 W/m^2
Frank,
Increased DLR and decreased OLR from 2xCO2 is a violation of COE.
Willis, is the CERES data sufficiently detailed to allow us to see the CO2 spectroscopic footprint? It would be interesting to see the amount of long wavelength radiation that is getting through in a real-world environment with water vapour in the atmosphere.
OK . . . to all WUWT readers that may wonder if they have the correct mental image of how the Earth radiates in the infrared band (particularly, radiation from global CO2 atmospheric content) and if this is accurately modeled over Earth’s surface area by a few linear (or even non-linear equations) run on a supercomputer . . .just check out this linked video. It’s been around for a while, but is still very informative and should be humbling to anyone attempting to calculate SWIR/LWIR from Earth’s surface directly to space and, moreover, from various atmospheric layers that are sending “downwelling” radiation to Earth’s surface.
Remember bright white area represents high IR emission to space while dark black areas indicate ~100% blockage/absorbtion of IR by water vapor and condensed water (i.e., clouds); peak IR imaging sensitivity is at 6.5 microns wavelength versus CO2 having a significant absorption/emission window at ~4-5 microns that is NOT blocked by atmospheric water vapor. The spatial and temporal variabilities are just wicked!
Enjoy: https://www.youtube.com/watch?v=f7QttjGu628
Dear Willis
You have better skills and tools than i have, so i would like to draw your attention to the surfrad data.
https://www.esrl.noaa.gov/gmd/grad/surfrad/
It is real radiation ( and other) data from the ground where humans live. Not up in the sky TOA.
They have only operated from 1995 and only a few places, but there must be some interresting connections to look at.
I wonder myself how the downwelling infrared depends on the upwelling. It looks that there are some connection, like constant difference og constant relation.
Best regards Svend Ferdinandsen
There is the Global Energy Balance Archive (GEBA). See article Martin Wild – The Global Energy Balance Archive (GEBA) version 2017: a database for worldwide measured surface energy fluxes https://www.earth-syst-sci-data.net/9/601/2017/essd-9-601-2017.pdf
Website GEBA and data acces: http://www.geba.ethz.ch/
“Data Availability and Exchange
The GEBA data are available at no cost for bona fide research.New users register by filling in the registration form”
Unsurprisingly, the uncertainty ranges are fairly large. It is instructive that UN IPCC CMIP5 do not do well in comparison.
There was some armwaving about the reasons for multi-decadal cyclical nature of brightening and dimming (increases and decreases in SW at the surface). They did not explain why the discussion was limited to the period beginning in the 1980’s.
It is interesting information.
Dave Fair: “Unsurprisingly, the uncertainty ranges are fairly large”
WR: Indeed, in 2017 the uncertainty ranges are still very large . The often wide uncertainty range is hardly known by anyone and hardly ever mentioned. For that an excerpt of the numbers and uncertainty ranges in the next table.
Some numbers of fluxes in the Global Mean Energy Balance diagram of GEBA, figure 4:
Best Estimate W/m2 Uncertainty Ranges W/m2
Solar absorbed surface 160 (154, 166)
Evaporation 82 (70, 85)
Sensible heat 21 (15, 25)
Thermal up surface 398 (384, 400)
Thermal down surface 342 (338, 348)
Thermal outgoing TOA 239 (236, 242)
Figure 4. Schematic diagram of the global mean energy balance of the Earth. Numbers indicate best estimates for the magnitudes of the globally averaged energy balance components (W m−2 ) together with their uncertainty ranges in parentheses, representing present-day climate conditions at the beginning of the 21st century.
Top-of-atmosphere fluxes are determined from the CERES satellite observations. Surface radiative flux estimates are derived from the CMIP5 model bias structure with respect to GEBA and BSRN observations as outlined in Wild et al. (2013, 2015). Adapted from Wild et al. (2015).
The diagram of the GEBA 2017 Global Mean Energy Balance:
Article: https://www.earth-syst-sci-data.net/9/601/2017/essd-9-601-2017.pdf
I’ve added an update to the head post showing the same results but using the HadCRUT surface temperature datase. Net result? No significant difference.
w.
Willis, this was very thought provoking and caused me to dig deeper and learn something new. So thanks!
What I’ve learned is that an increased greenhouse effect can be measured by looking at outgoing shortwave, which decreases with an increasing greenhouse effect.
See the following article from 2014:
The Missing Piece of the Climate Puzzle
http://oceans.mit.edu/news/featured-stories/missing-peice-climate-puzzle.html
“While one would expect the longwave radiation that escapes into space to decline with increasing CO2, the amount actually begins to rise. At the same time, the atmosphere absorbs more and more incoming solar radiation; it’s this enhanced shortwave absorption that ultimately sustains global warming.”
I recommend you try the same thing only looking at surface SW up this time vs. a surface temperature data set.
I tried this myself and for each 3.7 W/m^2 decrease in surface SW up, I found a 1.2°C increase in HadCRUT4.
For Berkeley Earth and GISTEMP, I found about a 1.5°C increase.
To be very specific, I used CERES “Surface Shortwave Flux Up – All-Sky”.