Guest Post by Willis Eschenbach
For a while now, I’ve been using a curious kind of scatterplot. Here’s an example. It shows the relationship between the surface temperature and the effects of clouds on surface radiation. Clouds can either warm or cool the surface, depending on location, time, and type. The phenomenon is called the “Cloud Radiative Effect” (CRE).
The amount of radiative cloud warming or cooling (the CRE) is measured in watts per square meter (W/m2). Positive means clouds warming the surface, and negative means clouds cooling the surface. Globally as an area-weighted average, clouds radiatively cool the surface by about – 21 W/m2.
Figure 1. Scatterplot, surface temperature (horizontal “x” axis) versus net surface cloud radiative effect (vertical “y” axis). Gives new meaning to the word “nonlinear”.
What are we looking at here? Well, each blue dot in Figure 1 represents a 1° latitude by 1° longitude gridcell somewhere on the earth’s surface. Each dot is placed horizontally with respect to its 21-year average temperature and vertically by its 21-year cloud radiative effect. The yellow/black line is a LOWESS smooth that shows the overall trend of the data.
And of particular interest, the slope of the yellow/black line shows how much the cloud radiative effect changes per 1°C change in temperature.
We can see from Figure 1 that clouds generally warm the coldest areas of the planet. Gridcells in the ~ 10% of the planet where the average annual temperature is below -5°C are warmed by clouds.
In warmer areas, on the other hand, clouds cool the surface. And when the temperature gets above about 25-26°C, cloud cooling increases strongly with increasing temperature. In those areas, for each additional degree of temperature, cloud cooling increases by up to -15 W/m2. This is because of the rapid increase above 26°C in the number, size, and strength of thermally driven thunderstorms in the warm wet tropics.
Here’s a video showing how the thunderstorms follow the warm water throughout the year.
Figure 2. Thunderstorm intensity is shown by colors (cloud top altitude is a measure of thunderstorm strength). Gray contour lines show temperatures of 27, 28, and 29°C.
From this, we can see that thunderstorms emerge preferentially over the hot spots, and they effectively put a cap on how far the temperature can rise in those areas. This is the reason that only 1% of the earth’s surface area, and virtually none of the open ocean, has an annual average temperature over 30°C.
With that as a prologue, since few people in climate use a gridcell-based scatterplot, let me discuss this kind of scatterplot. It has a very valuable property.
The value is that the method is looking at longer-term averages. In Figure 1, for example, these are the average temperatures that each of the gridcells has settled to after millennia. As a result, the gridcell-by-gridcell temperatures include all the possible various feedbacks and the majority of the slow responses to changing conditions.
And this allows us to answer questions like “what will be the response of the clouds if the temperatures warm slowly”? Alarmists would have you believe that the warming will be increased by the feedback of the clouds.
But Figure 1 tells a much more complex and nuanced story. The slope of the yellow/black line shows the change in CRE in response to a 1° change in temperature. If it slopes down to the right, it shows that the magnitude of the cloud-caused cooling is increasing with increasing temperature—the CRE is getting more negative, and clouds are doing more cooling..
There are only two places where the clouds act to increase an underlying warming. These are the areas in Figure 1 where the yellow/black line slopes upwards to the right. They are the 3% of the surface colder than -20°C, and the ~30% of the earth between 15°C and 25°C. These total about a third of the planet.
Gridcells at all other temperatures will have increasing cloud cooling as they warm, particularly the third of the globe that averages above 25°C.
Conclusion? Only a third of the globe has a warming cloud feedback and it is not that strong. Two-thirds of the globe has cooling cloud feedback, and in addition, the cooling feedback is far stronger than the warming feedback.
Thus, we can say that on average the cloud feedback is negative, not positive. An area-weighted average of the above data shows that globally, cloud cooling averages -3.2 W/m2 of cooling for each one degree C of warming. (In reality, the overall cloud response will be smaller than that, because the warmest areas of the earth where the cloud feedback is greatest are generally not going to warm much.)
Now, I’ve stated above that this method gives us the long-term answer after almost all of the various feedbacks, slow warmings, and adjustments have occurred. I’ve stated that this is not the short-term response of the clouds to surface temperature. It’s the long-term, basically steady-state response.
As a result, it can actually answer the question about the long-term response of clouds to 1°C of warming. And it can answer the question in detail, showing how cloud feedback varies from the poles to the tropics.
The only argument that I can see against this is that some slow thermal adjustment from the most recent warming hasn’t arrived yet. Possible, but here’s why that will likely make little difference—a rising tide generally lifts all boats.
In other words, if we have several nearby gridcells and one gets a slow residual thermal adjustment from recent warming, in all likelihood the other nearby cells will get a similar slow residual thermal adjustment as well.
And this will leave the slope of interest, the slope of the yellow line in Figure 1, pretty much unchanged.
Or at least, that’s what my logic said. However, I’ve always preferred data to logic. After some thought, I realized I could test this by taking shorter averages of the CERES data instead of the full 21-year average. I used 5-year averages of the same CERES data. For comparison, I’ve plotted them to the same scale as in Figure 1.
Figure 3. LOWESS smooths of the scatterplots of four selected subsets of the CERES data. Underlying scatterplot data not shown.
As you can see, the LOWESS smooth trend lines of all four gridcell scatterplots are so close that they cover each other up. This definitely shows that a gridcell scatterplot is indeed showing the long-term, all-inclusive relationship between the two variables of interest. It’s barely affected at all by the changes in CRE and temperature between the 5-year periods.
I’ll leave this here, and I will return to what I’ve learned from other gridcell scatterplots in the next post.
My very best to all,
w.
PLEASE: When you comment, quote the exact words you are responding to. I can defend my own words. I can’t defend your rephrasing of my words. Thanks.
Once again Willis highlights what really runs the climate(s).
Or more precisely, what DOESN’T run the climate(s).
Us.
I read the same article and got a different message.
Willis E. did not specify the initial causes of troposphere warming or cooling
He claimed that changes in clouds moderated those temperature changes.
It doesn’t really matter what causes the temperature changes in this argument.
He did show that the main premise of Climate Change fanaticism, that the small allegedly human cause warming will eventually be magnified by mechanisms, such as clouds – this conclusively shows that for most of the planet clouds act a damper not an amplifier.
So say, a doubling of CO2 would cause a ridiculous 3-5 degrees C over a hundred years, say, then the action of clouds would counteract that warming, especially strongly in areas that are above 25 degrees. In the range 25-30 C, cloud cooling increases by about 25W/m2 whereas supposed CO2 warming is only a few handful of watts.
I hope that CO2-global warming is actually true, for it would lead to a planet of LESS-extreme temperatures, equatorial regions unaffected and the poles a bit nicer for the planets and animals there. Like the way polar bears and arctic whale species have made such a huge comeback in the past 5 decades from just a small increase of temperature.
Again, the cloud radiative effect is not what you think it is, nor what the “science” usually claims. At least Gavin Schmidt is able to point out this basic issue..
Even if we only look at the LW part, there are 2(!) cloud radiative effects. One net, or single factor removal (SFR), one gross, or single factor addition (SFA). 60% of the CRE are overlapped with GHGs. This part of the GHE is caused by both clouds and GHGs simultaniously.
We can not, and that is important, take the net effect, compare it to the albedo effect of clouds (be it net or gross), and say clouds were cooling because the net CRE may be smaller than the albedo effect. This is not just logically illicit, it is equally unsupported by evidence. And of course the SFA CRE is larger than the albedo effect.
Indeed cloudiness and surface tempartures are positively correlated. Everywhere.
https://greenhousedefect.com/the-cloud-mess-part-2-something-spooky
The clouds are the condensed solid and liquid phase of water, emitting full spectrum IR, like all liquid and solid mass.
A transmission of heat from the surface to the cloud top in latent heat, condensed, then acting effectively as a blackbody radiator from there. Like a new surface. Not comparable to the radiation effects of gases.
This new surface (clouds) with a relatively more transparent view to space compared to the land/ocean surface, emitting full spectrum at a temperature around 273.15K. And we see this temperature is the peak Wien frequency in OLR observed from space.
Cloud radiative effect has been very confused.
Thx, but nothing you just wrote indicates you grasped what I am saying. And what I am saying is highly significant.
Your notions of spectral overlaps, gaseous absorption properties, and masking effects of cloud is fine and dandy in greenhouse vernacular. But it’s not a particularly insightful or useful description of what’s happening.
You are referring to Gavin like he’s a biblical prophet. That is an appeal to authority, where Willis is showing that authority is simply wrong by showing the actual behavior in the actual world in data, not theory or simulations.
I can disprove your argument and confirm Mr Eschenbach’s findings with just one Graph of actual data.
From Climate4You
http://www.climate4you.com/images/HadCRUT3%20and%20TropicalCloudCoverISCCP.gif
More cloud = less temperature, less cloud = more temperature, little change in cloud = little change in temperature
Sorry, but correlation does not mean causation.
Then explain me why in summer cloudy days are cooler then sunny days, and in the winter sunny days are bonecold, while cloudy days are much warmer.
that’s what Eschenbach’s plot is all about. And is is correct. You can average this out over time. I think you totally missed his point. It’s not about what radiation does in clouds, but about the direct blocking effect of clouds in the input output balance radiation and how it affects surface temperature.
you see that in the example of everyday weather.
it is an important double sided feedbackloop.
clouds are not watervapor. A lot of people make that mistake. A cloud is liquid water kept aloft by wind.
So there is more then albedo: there is also latent heat involved.
you are right if it would be water vapor only and radiation at the top of clouds.
but that’s not the surface temperature/radiation balance that the plots are referring to
It is not specifically Willis’ mistake, rather is one of many mistakes in “the science”. Willis is only just building his theories on it. And that is a problem almost all “criticals” have. They take a part of “the science” that seems to give them a leverage and start building theories right away.
Not just is “the science” largely wrong in such details, there is also plenty of misunderstanding on top of it. And that is why all these efforts are quite useless.
. 🤣🤣🤣🤣🤣🤣🤣
Leap directly to ad hominems.
No evidence, no proof,
No side-by-side comparison showing the alleged error.
Just a bald faced outright lie disparaging Willis.
You made the error, E. schaffer
And every regular WUWT visitor knows you made that mistake.
Willis is showing the actual observational data in a new way and analyzing what that teaches us, especially it teaches us how your prophets of doom are wrong in their *theory* and *simulations*.
Obersvational data, Hanover, 1950-2019
Hahahaha. One tiny spot. Willis’ data covers almost ALL THE PLANET.
Theory does not trump reality.
Exactly. Willis is showing actual observations, while Schaffer has no observations, just his/Schmidt’s pet theories. Ridiculous.
As I understand it from climatologists, it depends on the type of clouds — how thick they are and what altitude they are at. Some types of clouds have a net cooling effect, and other types have a net warming effect.
So, if cloud coverage changes, but the change involves a different mix of cloud types, that can change the net effect of the clouds.
Yes, but that doesn’t matter to this analysis, because it takes the observational data of all grid cells, regardless of what cloud type. As a grid cell warms, it will move further to the right along that curve, even if the types of clouds change, that change is represented by the other grid cells at that higher temperature, showing that the average of all cloud types at a given temperature amplify the cooling effect at a higher temperature, and increase the warming effect at lower temperatures… thus, clouds are a) on average always a moderating influence on temperature, and b) tend to cool at the equator and warm at the poles, which is good because most of the warming we see is at the poles, not the equator.
Where this is most important is to conclude: the decreased cloud cover in the latter 20th century can be blamed for a large amount of the increase in warming, and decreased cloud cover is largely blamable on human efforts to reduce aerosol emissions (which contribute to cloud formation). So, congratulations, warming IS human caused, but not by the stuff you blame it on. It is caused by your anti-pollution agenda.
Serious error in science comprehension:
“Indeed cloudiness and surface tempartures (sic) are positively correlated. Everywhere.”
That doesn’t mean that cloudiness is causing the increase in temperatures! Or else there would be rapid runaway effect! Increases in temperature, from whatever heat input increase, causes more evaporation, which leads to more clouds which blocks the most important heat source – The SUN – reducing any further increases. It’s a balance, an equilibrium that gets established.
And for all of the planet clouds are fighting runaway effects – warming the cold and cooling the warm regions.
This may be what you have read somewhere and what you chose to believe. But it is not how it is working. It ignores the reality that almost everything that blocks sun light is also blocking terrestrial IR. As a rule of thump just assume everything making the atmosphere more opaque will rather have a warming effect.
Again, this idea comes from “consensus science”, where it seems opportune to only look at absorptivity (or albedo respectively) while ignoring emissivity. This helps to vastly exaggerate the GHE and support the whole narrative of CO2 induced AGW.
I reality there is something similar however, but it works differently. The water cycle of evaporation and condensation constantly transports latent heat up into the troposphere. It is literally transporting heat away from the surface and cools it. Alternatively you may want to consider the reduction of the lapse rate due to latent heat, which equally means a reduction of the GHE. And indeed this is a very dominant negative “feedback”.
E. Schaffer October 14, 2022 10:21 am
Even if we only look at the LW part, there are 2(!) cloud radiative effects. One net, or single factor removal (SFR), one gross, or single factor addition (SFA). 60% of the CRE are overlapped with GHGs. This part of the GHE is caused by both clouds and GHGs simultaniously.
I have no idea what you are talking about, but it seems you misunderstand the surface net cloud radiative effect.
Surface net CRE is the difference at the surface between the total radiation (longwave plus shortwave) when there are clouds and when there are no clouds. It has nothing to do with single factor removal or addition. Just clouds, or no clouds.
It’s actually quite simple. Clouds have two opposing radiative effects. First, they reflect sunshine compared to no clouds, so they reduce solar radiation at the surface. This leaves the surface cooler than without the clouds.
On the other hand, they increase the downwelling radiation compared to no clouds. This leaves the surface warmer than without the clouds.
But there is NOTHING in there about single factor addition or removal.
Regards,
w.
Well, I am afraid so! You don’t need to be a fanboy of G. Schmidt, but the table I quoted above is taken from one of his publications. And it is one of the few exemptions, where “consensus science” at least tries to be accurate. I was aware of the problem well before I saw publication and was positively surprised it is at least featured there..
https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2010JD014287
There is NO seperate surface CRE (neither surface CO2 forcing or so..). Pls give up on “back radiation”..
No, it is definitely not that simple. We actually have 4 of which. On the LW side we have a net (SFR) AND a gross (SFA) LW-CRE. On the SW side we have the same issue. Just imagine a cloud over a snow covered surface. Then is the SW-CRE the amount of sun light actually reflected by this cloud, or is it the difference in albedo this cloud makes? Remember, most of sun light would be reflected by the snowy surface anyhow..
E. Schaeffer said:
I’m sorry, but that is demonstrably not true.
w.
I do not know what kind of data you have taken here. But I would very cautious not to confuse warming clouds with cooling rain. And yes, the two are naturally correlated.
So GCMs not doing clouds well is quite serious.
Worse than serious. See my old guest post ‘The trouble with climate models’.
They cannot do clouds at all. So clouds are parameterized (explained in post). And that drags in the attribution problem (natural variation vs AGW). And that is why the models have positive cloud feedback when WE shows again here that it is observationally somewhat negative. Thinking that thru, another proof that over the model parameterization period to best hindcast 30 years, there must have been natural warming that got dragged in and attributed to AGW.
Professor Mototaka Nakamura in his book says something similar about the failure of climate models.
He has little confidence in the GCM’s being able to make accurate predictions due to the inaccurate treatment of both oceans and clouds.
I think Willis has hit upon a major issue that needs to be addressed in the validity of the GCM projections/predictions.
From Fig 1 it appears that on 90% of the Earths surface clouds actually cool. And on the remaining cold ass regions it warms a little. Sounds like a pretty good system to me.
Slightly off topic.
Anyone know why the UAH site has not yet updated the September satellite temperature anomaly?
(It is usually published this many days into the following month.)
https://www.drroyspencer.com/2022/10/uah-global-temperature-update-for-september-2022-0-24-deg-c/
Thanks Luchezar!
I think my old link must have been failing to update.
Yes, but why not to publish and openly discuss the monthly UAH-reports anymore at WUWT? Who has changed the earlier approach? Maybe WUWT has not been satisfied with the monthly UAH-reports (I personally don’t think so) or Mr. Spencer has been “motivated” (rather disciplined) not to allow to publish for open discussion the monthly UAH-reports at WUWT? This policy change would be very noticeable.
Something from Germany to the issue “WUWT” due to the former monthly discussions on the monthly UAH reports. In a rather long blog discussion about the role of WUWT and on the contributions in it, the fraction of the CO2- and climate hysterics has stated explicitly the goal to reach a “Deutungshoheit” over climate realism, and an important task in diesem process would be to silence WUWT. As one reason for that exactly the publishing of the monthly UAH report and the open discussion about it has been explicitlt stated. I am not a native english speaker, so I don’t know the right term in english for “Deutungshoheit” (maybe “prerogative of interpretation”). Here is the definition of the term “Deutungshoheit” in German from wikipedia:
“Mit Deutungshoheit (auch Deutungsmacht) bezeichnet man das von einer Person oder Institution beanspruchte Recht oder die Macht, etwas allein und mit allumfassender Gültigkeit „deuten“ – und damit werten – zu können oder zu dürfen.[1] Treffen beispielsweise im Diskurs über eine Definition verschiedene Aussagen aufeinander, entscheidet die Deutungshoheit über deren endgültige Deutung. Der Wortteil „Hoheit“ impliziert hierbei eine ihr innewohnende Autorität, die Voraussetzung für ihre Akzeptanz ist.” (https://de.wikipedia.org/wiki/Deutungshoheit).
Conjecture: The disappearance of the monthly UAH reports and discussions from WUWT in the earlier form would not be accidental.
‘Say-so’
Authority
https://www.merriam-webster.com/dictionary/say-so
Willis,
Interesting data. What do you think causes the upslope of the LOWESS-smoothed line between 15C and 25C? (Or is that the subject of the next post?)
The base of an atmospheric column needs to be at or above 15C to reliably produce a level of free convection. That means the atmosphere partitions to a free convecting zone below the LFC and a dehumidifying zone above the LFC. It means convective instability can be supported.
Above a 22C surface, the atmosphere can have enough moisture to support cloudburst. Once that occurs, the most powerful column will be a mid level convergence zone so draws mid level moisture from the cooler zones. Surface at 26C are inevitably mid level divergence zones so have more sunlight. By 30C base temperature, the LFC is very close to the altitude of freezing so most cloud that forms and persists during and after cloudburst limits the surface sunlight and limits the surface temperature to 30C. The only ocean surface consistency warmer than 30C is near land masses that interfere with the cyclic instability.
The process is documented in detail here:
https://wattsupwiththat.com/2022/07/23/ocean-atmosphere-response-to-solar-emr-at-top-of-the-atmosphere/
Fascinating, maybe I have misunderstood, but not sure I fully agree with this:
It’s only a minor quibble, but between 15&25C isn’t the effect “reduced cooling ” rather than “warming” .it’s still in the “cooling” section of the chart.
Happy to be corrected.
True.
w.
…and the reason I made the point is because it strengthens your point even further. The conclusion must be that only place where clouds actually act to “increase the underlying warming” of the planet are the 3% below -20c.
If this is the case, and I can’t see any flaw in your argument, it seems to blow a big hole in everything that is currently argued.
…continuing to be picky, but hopefully helpful too, I think this “They are the 3% of the surface colder than -20°C” would be more accurately stated as “They are the X% of the surface colder than -5°C” .
I don’t know what the X% is but don’t imagine it’s much more than the 3%.
WE provided the area data in his chart. The smoothed line cross zero w/m^2 at about -4 C which appears to correspond to about 11% of the earth’s surface. I’d say this is a quite strong bit of evidence that cloud feedback is net negative and quite strongly so.
“Positive means clouds warming the surface, and negative means clouds cooling the surface.”
Neither clouds nor CO2 nor water vapour have the capacity to actually “warm”. They dont add energy. They’re passive effects that can only reduce the surface cooling rate. If there is a net energy increase in a region then that energy is coming from somewhere else, not the “clouds”.
That effect may be for example “clouds forming” where the latent heat of vapourisation is removed from one region and the clouds form in another region and release their energy there. Or perhaps its from ocean current energy transport. But the energy didn’t come from “clouds”.
It may seem nit picky that reducing cooling rate isn’t actually warming but this is good example of why its important to be precise about thinking.
True, but it’s a difference that makes no difference. Whether you call it “warming” or “adding additional energy to slow the cooling rate”, you end up at the exact same temperature.
w.
No difference mathematically, perhaps. But a world of difference (IMO) as to how to understand what’s actually going on.
The difference is between saying “A warms B” and “A leaves B warmer than it would be in A’s absence”.
I fear I see no “world of difference” between those two.
w.
Just a thought:
Never get into an discussion about the ocean becoming “More acidic” rather than “Less alkaline”. A pure waste of time…..
It is a very big deal when you begin trying to make more complicated gradients with time as a variable. Using averages for a fly over view is OK, but doing more complicated analysis requires accurate terminology and accurate math. Please don’t minimize the difference between reducing a cooling gradient and reversing a cooling gradient into a heating process.
Jim, the only difference between reducing and reversing a cooling gradient is magnitude. So I have no clue what you mean.
w.
What happens to the sign when Tcold becomes hotter than Thot when the gradient is (Thot – Tcold)?
The gradient goes from “+T/t” to “-T/t”, so it is more than just magnitude, the vector also changes direction.
Many folks forget that gradients are vectors with directions in the x, y, z coordinates, and includes time also.
How many people believe clouds bring energy to the party? They bring none.
TimTheToolMan October 15, 2022 1:43 pm
Sorry, Tim, but it’s unclear what you mean by “bring energy to the party”. Clouds absorb solar and longwave radiation along with latent heat of condensation, which is then radiated to their surroundings.
Is this “bringing energy to the party”? Because they are indeed a source of radiation.
w.
Willis write “it’s unclear what you mean by “bring energy to the party”.”
There is a big difference between being a source of radiation (everything is according to S-B) and being a source of energy. Clouds are not a source of energy. Clouds receive more energy than they radiate back to earth. They radiate upwards too.
The difference is important when contemplating how the atmosphere works IMO.
Willis, the addition of “%ge of the Earth’s surface” to the graph is a truly insightful stroke of genius on your part.
Don’t forget MODIS/CERES is in a sun synchronous orbit. Or more precisely a “daytime” synchronous orbit. So there’s a bit more to the story, i.e. night-time, and poleward darkness at certain times of year during daytime-type hours.
How does CERES measure surface temperatures? Presumably it is derived from another quantify.
What confounds everything, and this analysis, is turbulent fluxes in the boundary layer. Where net turbulent flux always opposes the direction of net radiation. Under sunlight net turbulent flux is directed upwards in relation to the surface, and in darkness net turbulent flux is directed downwards. Advection plays an enormous role. CERES mission is for radiation budgets, where TOA values are observed, and surface radiation flux is computed. This can say little of the actual temperature, as temperature is a property of the surface energy budget partitioning. This depends in large part in non radiative turbulent fluxes of sensible and latent heat. CERES virtual surface temperature is derived from a vertical profile of OLR and solar absorbed, but it is not a physical quantity.
Carlo, Monte October 14, 2022 10:53 am
I’ve converted the surface upwelling longwave radiation to the corresponding temperature. I’ve checked the temperature results extensively against Berkeley Earth, HadCRUT, and other temperature datasets. It’s accurate.
w.
Thanks, Willis.
Willis, isn’t this the same as saying that water vapor is a coolant in the warmer parts of the world?
Not exactly. Clouds are the main issue.
w.
Yes, it certainly clouds, but no water vapor, no clouds, no cooling. Isn’t it the case that water vapor drives your emergent phenomena cooling effect.
Nelson, a major quibble. Water vapor is in the gaseous state, containing the latent heat of evaporation. In which state it is indisputably a GHG. Clouds are in the liquid/ice phase states having given up that evaporation heat. So only in the loosest sense does your comment make any physical sense at all.
Rude, the creation of water vapor is the start of a complicated cooling process. Cloud formation is part of the process. As far as I know all polyatomic molecules in the atmosphere have a vibrational state in some part of the IR spectrum. Water vapor has a wide range of absorbtion, which helps provide buoyancy or convection. The point is that the net effect in the tropics is that water vapor drives net cooling, not heating. Figure 1 just shows the cloud effect. The latent heat absorbed at the surface that gets released when condensation occurs just adds to the cooling.
The argument is that water vapor may most importantly carry latent heat of evaporation from low altitude to high altitude, whereupon the heat is released in condensation—now well above most of the CO2 in the atmosphere, and therefore radiated to space regardless of CO2 levels.
Whether and how much water vapor absorbs/emits EM radiation is a separate issue.
Very nice work, as usual Willis. I would think that using annual averages would tend to smooth things out a bit, i.e., hide some of the clouds’ impacts and feedback effects. Would it be possible to produce two more versions of your Figure 1, the first using only 21 years of July data and the second using only 21 years of January data? I would think that would be illustrative, since you would then have max / min temperatures in the Northern and Southern Hemispheres, respectively.
WE, great post. I think it is peer reviewed journal worthy, because it shows the IPCC and climate models are observationally wrong about the sign of the cloud feedback. That is a BIG deal.
And, plugging your observational negative feedback number into the Bode feedback method of determining ECS (commented on several times previously here, using Lindzen’s 1.2C for the no feedbacks baseline) drops the estimated ECS value from 1.8 to something below 1.7C, much more in line with the several energy budget method results.
Thanks, Rud. Coming from you, that means a lot. I agree that I think I’m using an unexplored analysis method with a lot of value for this and other questions.
My main problem with writing for the journals is that I feel like I have to give myself a lobotomy to write in the dense, turgid giant-paragraph style the journals seem to prefer.
That, plus I’m always working so far away from the rest of the field that I rely very little on the work of others. For example, all of what you see above is straight out of my own imagination. I know of no one else doing this kind of analysis … which makes for a very thin and pallid references list …
But hey, if you want to collaborate, I’ll give you lead author status. You do the writeup and the nego with the journals, and I provide the analysis, data, and computer code.
Or if not you … anyone else interested?
w.
Journals only publish “THE Science” Willis.
Where pre-corroborated findings are accepted for peer review by pre-corroborating reviewers.
Yours and Rud’s work both carry the fatal flaw of originality & authenticity.
Many papers contradicting the consensus narrative are published. Take a look through
http://notrickszone.com/#sthash.cYDRqRGh.dpbs
to see many examples.
I couldn’t be lead author; am an Eagle Scout. And don’t want you to need a lobotomy. Let it stand as VERY high praise. I will add the IPCC/CMIP cloud sign problem you just proved to my growing list of simple to understand alarmist fails worthy of ridicule.
The climate science consensus is represented by the average climate model prediction — inaccurate for over 40 years. There is no chance of progress in climate science if one works with members of that consensus or tries to please them.
Most great science discoveries in past centuries were by individuals or small teams who REFUTED the existing consensus!
Oh, yeah, forgot to ask. What is the “Bode feedback method of determining ECS”? Links would be welcome.
Thanks,
w.
Explained on pp 284-285 of my ebook The Arts of Truth, in the long climate chapter Lindzen kindly reviewed before publication.
Short summary from which you can Google links. Bode worked at Bell Labs in the 1930’s. He proved that for any amplifier/attenuation circuit, the system gain/loss was equal to 1/(1-f) where f is the linear sum of all individual feedbacks in the circuit. Lindzen realized in 2011 that the Bode equation results in ECS if f is all climate feedbacks. My book used his Bode curve based on zero feedbacks =1.2C (Monckton’s equation yields 1.16C), given in his UK parliament talk early in 2012, which was then available on his MIT website. The Lindzen Bode curve is simply f on the X axis and delta degree C on the Y, where zero feedbacks f=0 gives 0 delta C. Page 285 reproduced Lindzen’s curve with his permission.
So, for example, if IPCC ECS is 3, then f is 0.67. Now IPCC AR4 said water vapor feedback (WVF) doubles zero feedback. So 2*1.2=2.4C or f=0.5. AR 4,5, and 6 all say the sum of everything except WVF and clouds is a net of ~0. So IPCC cloud feedback f is +0.17 (0.67-0.5).
Now Dessler in 2010, then McIntyre’s 2011 rework showed cloud f is essentially 0. And ARGO salinity has now shown that ocean rainfall is ~twice what is modeled, so half WVF f~0.25 which via the Bode curve gives an ECS of 1.8.
But cloud is a bit negative, not zero. So f=0.25 is too high. f=0.22 (0.25-0.03 per your w/m^2 estimate translated to -0.03f) gives ECS ~1.65, like the energy budget methods (I like the second Lewis and Curry paper best, because it incorporates responses to all the criticisms of the first.)
Regards to you and your gorgeous ex fiancée.
Thanks, Rud, appreciated.
w.
“AR 4,5, and 6 all say the sum of everything except WVF and clouds is a net of ~0″
Essentially an assumption, not a fact. The GCMs have been built with that assumption. They have no capability to do anything else. They’re designed to have stable control runs.
Fig. 3 of https://co2coalition.org/wp-content/uploads/2021/08/On-Climate-Sensitivity.pdf
ECS? Yeah, another unicorn argument? Adding temperatures is not really in line with an intensive property.
Even IF you were technically right, you are squarely on the losing side. Think about it if you can.
IPCC and their side have used ECS as a climate cudgel since 1988. So if it is a unicorn as you now assert, so are they.
That makes you a unicorn too, Rud.
No ECS proponent here, but unless you’re Christopher Monckton the concept isn’t based on adding temperatures.
As is illustrated in Fig. 3 of https://co2coalition.org/wp-content/uploads/2021/08/On-Climate-Sensitivity.pdf, the feedback explanation is usually based on adding forcings rather than temperatures. Now, the way in which forcings actually combine is somewhat more complex than the addition depicted in that explanation, so that explanation has its problems, too.
But the proposition that the ECS concept is based on temperature addition probably isn’t the best reason for dismissing it.
the feedback explanation is usually based on adding forcings rather than temperatures.
Same thing if it’s to do with the S-B equation. You cannot add forcings or else you could heat a room with a billion ice cubes.
Willis, I’d like to pile on with further praise here for your recurring productive use of subject-clarifying graphics that 2 1/2 years ago alerted this medico quite early in the Covid pandemic just who was most at risk for severe outcomes as experienced in northern Italy, soon after many Chinese fashion industry workers had returned on direct flights from Wuhan to Milan following Chinese New Year celebrations in the contagious company of their extended families. Once again that sense of proportion, little to be found elsewhere, was immensely valuable to me and others that I was able to extend it to and we all owe you a considerable debt of gratitude for your cognitive enterprise.
Water vapor is at complete saturation very close to 30C. It seems the strong change is correlated to that mechanism.
The available data does not support that statement.
What is the mass of water in a saturated atmospheric column with base at 30C?
Don’t fully understand the CRE. Literature says it is composed of two parts, blocking solar radiation from reaching the surface, and trapping radiation from the surface. The first I understand but not the second. Visible clouds I assume are composed of condensed water vapor with little GHE. If it is water vapor contained in the cloud that “traps” upwelling radiation and returns some to the surface then that would make sense.
Don, not WE. But see essay ‘cloudy clouds’ in ebook Blowing Smoke for an unraveling. In short, clouds are complicated. For incoming SWR solar, they provide albedo (reflection) so lower surface heat absorption. The surface up resulting IR is much more complicated, depending on cloud altitude and optIcal thickness. For example, very high thin cirrus (almost no optical thickness) actually warms (essence of Lindzens ‘adaptive iris’ paper). That is because cirrus (fine thin layer of ice crystals) is transparent to SWR but opaque to LW IR.
That’s fascinating, Willis, for several reasons. It has been obvious for years that clouds play a huge role in temperature, as you have often shown here. These results of yours, while not obvious or trivially obtained, do not seem beyond the reach of honest climate scientists. So why do we not see this feedback explication used to destroy the sky-is-falling, kill-economies-everywhere narrative that the climate scare pushers have shouted to the deafening point?
I think the answer is obvious. And I hope you will watch your back.
You can record temperatures but you cannot average temperatures. It is meaningless.
Temperature is an intensive property.
Similarly, in the S-B equation it is meaningless to average, add or subtract black body radiation fluxes because they are based on the temperature of a black body.
Did you know the average person has one testicle? That would make Hitler the perfect human. I don’t think so.
I agree with this but I do not understand how it relates to the post?
Globally as an area-weighted average, clouds radiatively cool the surface by about – 21 W/m2.
Each dot is placed horizontally with respect to its 21-year average temperature and vertically by its 21-year cloud radiative effect.
10% of the planet where the average annual temperature is below -5°C are warmed by clouds.
This is the reason that only 1% of the earth’s surface area, and virtually none of the open ocean, has an annual average temperature over 30°C.
Thus, we can say that on average the cloud feedback is negative, not positive. An area-weighted average of the above data shows that globally, cloud cooling averages -3.2 W/m2 of cooling for each one degree C of warming
It’s the use of the word ‘average‘.
But the shape of the curve would not change if there was no area average of temperature. The absolute value of the power flux would likely vary for a particular temperature but it is a made up number because it relies on back radiation in its calculation, which well-educated physicists know does not exist.
Average temperatures are still useful. For example, the time of year the average morning temperature drops below freezing is good for gardeners to be aware of. They aren’t really concerned whether it is an intensive or extensive property.
Like I said, meaningless.
Ja. Ja. I told you. Global warming is not possible because of certain mechanisms both on the sun and the earth.
An observation from the plot – average cloud over any ice free ocean surface is cooling. This may or may not be true but could be wrong because you are using make-believe surface energy fluxes and not working with measurable energy fluxes at the top of the atmosphere. You would be better served to look at measured top of the atmosphere fluxes for the exercise. They are real and not implied from radiance measurements scaled in W/m^2..
The causes –
The change in slope at 15C is due to the change of convective regime. Once the surface temperature exceeds 15C it can create a level of free convection and that enables convective instability – that is the temperature where the clouds begin to pop. Attached photo is over open water at 14C with general air mass moving from water at 15C. The clouds are just beginning to “pop”.
The tallest convective columns over open oceans cannot sustain more than 30C surface temp due to cloud persistence. They are always mid level convergence zones and upper level divergence zones over open ocean. These zones inhibit convective instability by drawing mid-level moisture from the adjacent cooler columns. Usually surface above 26C but less than 30C are mid level divergence zones so lose moisture to the tallest columns but gain heat due to upper level convergence from diverging warmer air from the tallest columns.
Few people appreciate the significance of the change in slope of the smoothed curve at 15C. That is the minimum surface temperature that can reliably support a level of free convection.
All life on Earth as we know it depends on that simple fact. Without the ability of the atmosphere to partition at 15C, the atmosphere would saturate and clouds would be ever persistent. The condition would not last long because the planet would become a snowball.
Attached is the relative humidity versus temperature over the global oceans. It is based on monthly data so there is the possibility of RH exceeding 100% due to response times involved. The key observation is that there is no point above 15C with RH at 100%.
Once convective instability can be supported, the atmospheric column only exceeds 100% RH during the instability (cloudburst) and that requires a surface temperature above 22C.
Thanks, Rick. You are quite correct. What you call “convective instability” I call the emergence of Rayleigh-Bénard circulation.
(From my post Emergent Climate Phenomena)
My best to you,
w.
The mention of the US President’s power, what happened back in the time of King George the 3 rd.
So you got rid of a King, & then gave his Power to the new President.
Fast forward to 1910 & Australia got the up-dated version of the s Westerminister system.
A pity that the US got the old version.
Michael VK5ELL
Thanks again Willis and lets hope that you and Rud can collaborate on a paper for a scientific journal ASAP.
Don’t forget that the Western countries have already wasted trillions of $ on their so called emergency or crisis or even Biden’s EXISTENTIAL THREAT.
So far this has been a mission impossible, but who knows you may be able to break through and stop the OECD countries further waste of 100s of trillions of $ by 2050 or whenever? All the best to both of you.
This was a very interesting analysis of the effects of clouds well backed up by the experience of cloud effect in the tropics and in the polar regions. Global Climate Models which fail to model clouds are useless.
I’ve looked at clouds from both sides, now.
You really don’t know clouds at all. 🙂
As I recall the scatterplots in excel are called XY plots. Very versatile. I used them at one time to map 3-D shapes into 2-D patterns, before CAD apps provided this.
Figure 1 pretty much says it all. The 3 x water feedback is not only wrong. It is out to lunch off the charts wrong.
I’m trying to imagine what would happen to Willis if he were a researcher in some alarmist aligned organization and he presented this report to his overseers. What do you think?
Hi Willis. I check into the site only infrequently, so usually miss your posts within the comment window. I’d like to comment on a few older posts.
In A Serious Question, you ask:
You forget: even that 1.2 W/m² at TOA isn’t actually Earth receiving any more power. It’s just a number reflecting a transient energy imbalance, until a 0 W/m² imbalance at TOA is once again restored.
You don’t need larger heat flows at TOA for radiation fluxes at the surface to rise.
What has to be conserved is HEAT flow, not RADIATION FLUXES. At all times, heat flows are conserved: Sa = 240 W/m² is absorbed, and an equal amount leaves. The surface absorbs around 160 W/m² from the Sun, and an equal heat flow (NRH + SLR – DLR) leaves, where NRH is non-radiative heat loss, SLR is surface LW radiation emissions and DLR is downwelling LW absorbed.
That’s what energy conservation requires.
Energy conservation places no limitations on how large SLR (and temperature) can be, with the same value of incoming sunlight.
People have an entirely false belief that temperature is determined by how much power enters the system. That’s not remotely true. Temperature is determined by the thermal response curve (heat flow vs. temperature difference) of the cooling system.
Consider a generic system like this:
In such a system, the temperature of Object B is ALWAYS determined by what temperature difference system C needs to be exposed to in order to transfer heat from B to D at rate P.
That’s the ONLY thing that determines temperature in this sort of system
So, you can change the temperature by changing the response curve of C.
Example: Consider a cabin warmed by a wood stove burning logs at a fixed rate. The temperature drop from inside to outside will be determined by the absolute thermal resistance or R-value of the walls. If you add insulation and increase the thermal resistance, the temperature will rise, although there was no increase in the heat being provided.
Here is an illustration (by me) of the process in which added insulation creates warming: https://ibb.co/7twY3ZB
The Greenhouse Effect just decreases the efficiency of radiant heat loss. Mathematically, it’s analogous to adding insulation to a house. There is no need for more external power. The retained power just builds up to a higher level.
Here is an illustration (by me) of the process in which adding GHGs creates warming: https://ibb.co/w6bHVQK
The idea that energy has to “mysteriously” be found from somewhere is leading you astray.
Greenhouse gasses simply decrease the efficiency of heat transfer through the atmosphere to space. That raises surface temperatures.
Bob, in a perfectly insulated cabin wouldn’t the air temperature just rise to the temperature of the wood stove. The wood stove surface temperature won’t change. Your analogy makes no sense to me.
Nelson, thanks for the question.
My argument was focussed on realistic scenarios in which the temperature of the cabin interior (maybe in the range from 32-77℉ / 0-25℃) is considerably lower than the stove temperature of 375-650℉ / 190-340℃.
Under those conditions, the heat transfer rate from the fire to the cabin interior doesn’t change much as the cabin interior warms. If you improve the insulation in the walls, the interior temperature will be warmer, exactly as I described. That will be true even though the fire isn’t putting out any more heat.
* * *
Your question takes that argument to it’s logical limit, looking at to whether there is a limit to how large a temperature increase could be produced by increasing insulation.
This question is somewhat academic, since neither a cabin nor the Earth has a temperature anywhere near the temperature of its heat source (the fire or the Sun, respectively).
But, I can address the question regardless, in case that would be helpful.
To be more rigorous, consider four variants on the scenario:
I think you’re really asking about variant #1, where the heat source has a fixed temperature.
When the temperature difference between the heat reservoir and the cabin interior is large, the heat flow into the cabin P will be fairly consistent. But, as the interior temperature approaches the temperature of the heat reservoir, so that ΔT approaches zero, the heat flow P = f(ΔT) from the reservoir to the cabin interior will drop towards zero.
So, I agree that the interior temperature can’t and won’t exceed the temperature of the heat reservoir—if that reservoir has a fixed temperature. But, that’s NOT because the heat transfer curve of the cooling system doesn’t matter, and doesn’t help determine temperature. It’s just that there is also a heat transfer curve for heat transfer from the heat reservoir to the object to take into account.
Temperature in general is determined by the balance between these two heat transfer curves.
* * *
When the temperature of the heat source is much higher than that of the temperature of the thing whose temperature we’re investigating, then the heating rate doesn’t vary much as the object temperature changes. In that regime, temperature is determined solely by the heat transfer curve of the cooling system and where the heating rate P₀ falls on that curve.
* * *
What I said about the temperature of the cabin being determined by is rigorously true in any variant in which the heat is generated at a fixed rate P. That is true in variants #2 and #3.
An electric heater or wood stove is NOT a heat reservoir with a fixed temperature. It’s not true that “The wood stove surface temperature won’t change.”
In variants #1 and #3, the surface of the electric heater or wood stove does not have any intrinsic temperature which caps how high the interior cabin temperature can get.
Both types of heaters are, once again, examples of the sort of A-B-C-D system I’ve described, except that:
So, as the temperature of the cabin interior rises, then the temperature of the heater’s case must also rise—unless the heat production rate inside the heater decreases.
For a wood stove, variant #4 is what really happens: the heat production rate will depend on temperature. I don’t think that case is worth analyzing in detail, but the usual principle applies: temperature is determined by the balance between the heat production curve and the cooling curve.
* * *
Bottom line: nothing in your argument is inconsistent with what I’ve said.
Is any of this making sense?
* * *
In the case of the Earth, we’re in variant #1, with the heat source being the Sun.
It’s true that heat from the Sun could never raise Earth’s temperature, or any other temperature, to be hotter than the temperature of the Sun.
But, we are in a regime where the temperature of the Earth (288K) is far lower than the temperature of the Sun ≈6000K. Because the temperature difference is so large, the rate of heat delivery by the Sun is essentially independent of changes in Earth’s temperature. (That’s only affected by changes in albedo.)
But, for a given rate of heat absorbed by Earth, Earth’s temperature is determined by the heat transfer curve of the cooling system (i.e., the Earth’s surface and atmosphere), specifying how much heat transfer to space there will be for any given average surface temperature.
* * *
This can be rigorously quantified, in the case of Earth.
If the atmosphere was transparent to longwave radiation, then the heat transfer curve cooling rate to space, OLR, would be specified by:
OLR = OLR_no_ghe = 𝜀σT⁴
However, given that the atmosphere is not transparent to LW radiation, the longwave emissions at TOA can be different than the LW emissions from the surface. So, in the presence of LW-absorbing materials in the atmosphere, the heat transfer curve cooling rate to space can be expressed as
OLR = OLR_ghe = 𝛽⋅𝜀σT⁴
where 𝛽 is a multiplier that is different than one only because of the presence of stuff in the atmosphere that absorbs/emits/scatters LW radiation. On Earth, at present, 𝛽 = (OLR/SLR) = (240/398) = 0.60 where OLR is outgoing longwave radiation to space and SLR is longwave radiation emitted by the surface.
If sunlight provides heat at a rate Sn (net absorbed insolation), then in the transparent atmosphere scenario
T_no_ghe = [Sn/(𝜀σ)]^{1/4)
But given the presence of LW-absorbing materials, the temperature is different, i.e.,
T_ghe = [Sn/(𝛽𝜀σ)]^{1/4)
The ratio of these temperatures is
T_ghe / T_no_ghe = [1/𝛽]^{1/4) = 1.14
Given T_ghe = 288 K, that means T_no_ghe = 288/1.14 = 253K. That’s a temperature difference of 35℃ due to 𝛽 being less than 1, which is possible only in the presence of an atmosphere that is not transparent to LW radiation.
(Note: that 35℃ is larger than the more commonly quoted value of 33℃ being the size of the GHE, because non-uniformity of temperatures on Earth make the transparent-atmosphere value 2℃ cooler, ie., 253K instead of 355K.)
The “longwave transmission reduction” parameter 𝛽 characterizes a shift in the heat transfer characteristics of the atmosphere. As 𝛽 decreases, the temperature rises.
(The only qualifier to that conclusion is that temperature could fail to rise if the albedo, a, shifts to compensate by reducing that net absorbed insolation Sn = Si⋅(1-a) for given incident insolation Si; or if Si itself shifts to compensate for the decrease in 𝛽).
There are no approximations involved in any of this. It’s an exact result, which must be true, given only (1) the Stefan-Boltzmann law, and (2) the fact that in equilibrium the energy in at TOA must equal energy out, i.e., Sn = OLR.
So, it’s a rigorous result that the temperature of Earth is set by 𝛽 and the net absorbed insolation Sn.
* * *
The only question this analysis does not address is: when the concentration of GHGs (greenhouse gasses) in the atmosphere increases, does that necessarily decrease 𝛽?
Instantaneously, the answer is “yes.” The “radiative forcing”, ΔF, that climate scientists calculate is
ΔF = -ΔOLR = -SLR⋅Δ𝛽
(since when calculating “forcing” nothing except OLR is allowed to change).
So, initially, Δ𝛽 = -ΔF/SLR.
But, as climate reacts to the initial change in beta, Δ𝛽, the value of 𝛽 can again shift. (That’s what the IPCC calls “feedbacks.”)
So, this analysis does not prove that increasing GHGs increases temperature (after feedbacks are taken into account).
But, it does prove that temperature is intimately tied to the efficiency with which surface radiative emissions are translated into radiative emissions at TOA.
* * *
I stand by the assertion that temperature isn’t determined by the rate of incoming heating alone. Instead, temperature is determined by the rate of incoming heating in combination with the heat transfer characteristic of the cooling path.
In particular, the temperature rise of an object (the case of the heater, the interior of the cabin, or the surface of the Earth) above the temperature of the heat sink (the air around a heater, or the air outside a cabin, or the space around Earth) will be the temperature rise that leads to a heat transfer rate for cooling that is equal to the heat transfer rate from the heat source.
𝛽 here is an empirically derived ratio. It does not have an inherent physical meaning. It could be imagined as opacity to LW, or transparency in the sense discussed.
It has been described here as
The conjecture is that
But in reality it is not what is observed in NCEP reanalysis. In fact the opposite occurs during periods of the most surface warming. In reality, over the long term, 𝛽 is best described as a constant. Where surface temperature is free to change irrespective of 𝛽.
Supposing that 𝛽 has any role in influencing surface temperature may be an artefact of imagination, and may just be an empirical consequence of the system. One which suggests the LW flux regime is simply along for the ride. Or, alternatively, that it is overwhelmed by other factors.
Some kind of regime shift is observed in the early 2000s, as the 𝛽 curve appears to change direction.
Thanks for your comment, JCM.
As encouragement for engaging with this long comment, I’ll foreshadow that, near the end, I offer a formula for Δ𝛽_forcing which I believe offers an opportunity to directly, empirically test the hypothesis that radiative forcing is impacting Earth’s climate.
Yes, 𝛽 is an empirical parameter. But it’s a well-defined quantity which can be used as the basis of rigorous scientific reasoning.
That is NOT a “conjecture.” A slightly refined version of that statement is a rigorous, inevitable conclusion based only on the Stefan-Boltzmann law and the equilibrium requirement that OLR = Sn, where OLR is the total power of outgoing LW radiation and Sn is the power of absorbed sunlight.
If Earth’s surface had a uniform temperature T, then the rigorous conclusion would be:
T = Tₓ / 𝛽^(1/4)
where Tₓ = [Sn/𝜀σ]^(1/4)
Given that the temperature of Earth’s surface varies, the rigorous conclusion is what I might call the “master temperature formula” (MTF):
⟨T⟩ = Tx / [⟨𝛽⟩’⋅𝛾]^(1/4)
where
⟨x⟩ denotes the global average value of x
⟨𝛽⟩’ = ⟨𝛽⋅SLR⟩/⟨SLR⟩ =⟨OLR⟩/⟨SLR⟩
𝛾 = ⟨T⁴⟩/⟨T⟩⁴ ≥ 1
SLR = 𝜀σT⁴ = surface LW emission rate
Tx = [⟨Sn⟩ / ⟨𝜀⟩’ σ]^(1/4)
⟨𝜀⟩’ = ⟨𝜀⋅T⁴⟩/⟨T⁴⟩
The factor 𝛾 would be 1 if Earth’s surface temperature was uniform, but is greater than 1 because of variations in temperature (e.g., between the tropics and polar regions).
This result is a rigorous, inevitable consequence of the Stefan-Boltzmann law and the requirement that ⟨OLR⟩ = ⟨Sn⟩.
As the MTF indicates, the average temperature ⟨T⟩ depends on net absorbed solar irradiance ⟨Sn⟩, weighted average emissivity ⟨𝜀⟩’, the LW transport reduction factor ⟨𝛽⟩’, and the “temperature variation emissivity boost” factor, 𝛾.
If you’re not seeing a clear relationship between changes in ⟨𝛽⟩’ and changes in ⟨T⟩, that means that one or more of the other 3 factors (⟨Sn⟩, ⟨𝜀⟩’, 𝛾) is changing.
The value of 𝛾 changes during times like El Niño and during AMOC oscillation. As the poles warm more rapidly than the tropics, this decreases the value of 𝛾.
So, it’s very possible that some average surface temperature changes are due to changes in 𝛾 rather than change is 𝛽. That doesn’t mean that 𝛽 doesn’t continue to be a key factor in the overall equation for temperature.
That might be the case if the idea of 𝛽 influencing temperature was simply a hypothesis based on observations. But it’s not. It’s a conclusion based on rigorous theoretical analysis based on two simple physical principles: the Stefan-Boltzmann law and the equality of energy in and energy out in thermal equilibrium.
It’s not “simply along for the ride”, in the sense that it is ALWAYS one of the multiplicative factors that determines temperature.
It is, however, true that changes in other factors could in principle mitigate changes in 𝛽.
Your data is very interesting. But, you’re NOT plotting 𝛽. How do I know? The values are too high. In 2015, 𝛽 had a value of 0.60 (IPCC AR6 WG1 p948). What did have a value of 0.67 in 2015 was 𝛽_clearsky, the value when no clouds are present. So, you plot apparently shows 𝛽_clearsky.
My claims about the influence of 𝛽 on temperature relate to 𝛽 = 𝛽_allsky. The quantities 𝛽_allsky and 𝛽_clearsky have different values, and those values won’t necessarily change together if the distribution of clouds changes over time, as it has during that period.
So, we don’t yet have in this conversation any information about how 𝛽 has actually changed over time.
* * *
But, if 𝛽 changed over time in a manner similar to 𝛽_clearsky, would that indicate a problem?
Let’s look at how I would have expected changes in 𝛽 and T to relate over this period. As I’ll derive at the end of this comment:
ΔT = T⋅[Δ𝛽_forcing/𝛽 + ΔSn/SLR – Δ𝛽/𝛽 – Δ𝜀/𝜀 – Δ𝛾/𝛾]/4
or equivalently
Δ𝛽 = Δ𝛽_forcing – 𝛽⋅[ΔSn/SLR + Δ𝜀/𝜀 + Δ𝛾/𝛾 + 4ΔT/T]
where
Δ𝛽_forcing = -ΔF/SLR
From 1960 to 2020, it’s claimed that forcing due to CO2 has increased by ΔF ≈ 1.4 W/m² corresponding to Δ𝛽_forcing = -0.0035. Temperature has increased by about 1.0℃ so that -4𝛽ΔT/T ≈ -0.008. The sum of these is about -0.012.
Your chart shows 𝛽_clearsky at about 0.655 in 1960, going to about 0.670 in 2010 and dropping back to around 0.655 in 2020. That a change Δ𝛽_clearsky ≈ +0.015 from 1960 to 2010, -0.015 from 2010 to 2020, for a net change around 0.
Is that compatible with my equation for Δ𝛽? Certainly. It’s not a problem, given that:
So, whatever the evidence shows about changes in the observed 𝛽, that doesn’t mean that 𝛽 isn’t related to temperature. It just means that changes in things like cloud cover and the global temperature distribution are also happening.
* * *
Simply plotting 𝛽 over time and not seeing a trend doesn’t mean that changes in 𝛽 don’t drive temperature.
But there is a way of checking this!
To see if changes in Δ𝛽 are playing a role, what one needs to do is to calculate and plot Δ𝛽_forcing where
Δ𝛽_forcing = Δ𝛽 + 𝛽⋅[ΔSn/SLR + Δ𝜀/𝜀 + Δ𝛾/𝛾 + 4ΔT/T]
Measuring and plotting the empirical value of Δ𝛽_forcing (equivalent to -ΔF/SLR) offers a test of the hypothesis that radiative forcing is affecting climate:
* * *
APPENDIX: DERIVATIONS
Relationship between Δ𝛽 and ΔT when no forcing
Let’s look at relationships in the absence of any forcing.
I’ll use my rigorous non-uniform temperature result, by use a simpler notation in which Sn, T, 𝛽, 𝜀 actually mean ⟨Sn⟩, ⟨T⟩, ⟨𝛽⟩’ and ⟨𝜀⟩’. Then raising my MTE to the 4th power yields:
T⁴ = [Sn/𝜀σ]/𝛽⋅𝛾
𝛽⋅𝛾⋅𝜀⋅σT⁴ = Sn
This implies changes should look like
Δ𝛽⋅𝛾⋅𝜀⋅σT⁴ + 𝛽⋅Δ𝛾⋅𝜀⋅σT⁴ + 𝛽⋅𝛾⋅Δ𝜀⋅σT⁴ + 4⋅𝛽⋅𝛾⋅𝜀⋅σT³⋅ΔT = ΔSn
or
SLR⋅[Δ𝛽/𝛽 + Δ𝜀/𝜀 + Δ𝛾/𝛾 + 4ΔT/T] = ΔSn
ΔT = T⋅[ΔSn/SLR – Δ𝛽/𝛽 – Δ𝜀/𝜀 – Δ𝛾/𝛾]/4
Δ𝛽 = 𝛽⋅[ΔSn/SLR – Δ𝜀/𝜀 – Δ𝛾/𝛾 – 4ΔT/T]
Relationship between forcing ΔT and Δ𝛽
The result above assumed equilibrium. A GHG forcing instantaneously decrease OLR by an amount ΔF, which is equivalent to saying 𝛽 changes by an amount Δ𝛽_forcing = -ΔF/SLR. This shifts the system out of equilibrium so that the energy balance equation at TOA is
Sn – OLR = ΔF = -SLR⋅Δ𝛽_forcing
However, when the system changes, the total change in 𝛽 need not stay at that value. Instead
Δ𝛽 = Δ𝛽_forcing + Δ𝛽_response
where Δ𝛽_response obeys the equations described above. The net result is:
ΔT = T⋅[Δ𝛽_forcing/𝛽 + ΔSn/SLR – Δ𝛽/𝛽 – Δ𝜀/𝜀 – Δ𝛾/𝛾]/4
Δ𝛽 = 𝛽⋅[Δ𝛽_forcing/𝛽 – ΔSn/SLR – Δ𝜀/𝜀 – Δ𝛾/𝛾 – 4ΔT/T]
where
Δ𝛽_forcing = -ΔF/SLR
Filling in numbers
Let’s look at how radiative forcing has changed over the period you charted, from 1960-2020. Forcing nominally changes as
ΔF = (5.35 W/m²) ln(1 + ΔC/(278 ppm))
From 1960 to 2020 the concentration of CO2 has changed from about 317 ppm to 412 ppm. So, ΔF nominally changed from by a net amount ΔF = ΔF2020 = ΔF1960 ≈ 1.4 W/m².
This implies Δ𝛽_forcing = -(1.4)/(398) ≈ -0.0035.
There was about a 1℃ temperature rise from 1960 to 2020. SNR changes at a rate of about 5.5 W/m²/℃.
So, all that means I would expect over that period
-4𝛽ΔT/T = -4⋅0.60⋅1/288 = -0.008
The forcing and temperature change contributions to Δ𝛽 sum to -0.012.
No, the plot is of all NCEP area weighted grid cells. I have calculated LW upward flux from surface temperature with an emissivity of 0.97. If you want to get closer to your value of 0.6 simply set emissivity to 1 for all annualized data. Year 1960 then starts with an all-sky 𝛽 of 0.62 based on NCEP reanalysis area weighted grids.
The linear equations you have posited assume a stable static atmosphere.
Δ𝛽 is observed with a positive slope for the period of record, not one of -0.012.
Your direct forcing derivations may be valid, but it only resolves a partial system. This yields virtual values that are not observed in nature. In reality there is a thermodynamic response.
It is the same pitfalls of climatology clearly demonstrated in your thorough responses. Radiative physicists with the audacity to believe the system can be reduced in such a way as presented. The reality is much more fascinating.
JCM: No, the plot is of all NCEP area weighted grid cells.
Interesting. I’ll need to think through the implications.
Critically important: Is the plot showing:
The latter would correspond to what I’m focused on. That’s probably what the plot shows, but I have to ask to be sure what I’m seeing.
((b) is also equivalent to the SLR-weighted average of (OLR/SLR))
I’m pretty sure no such assumption is present. Why would you believe that to be the case?
I see that. I’m going to need to think through the implications.
As I noted in another comment, I added forcing to the equation as a last-minute addition, much less thought through than the rest, and I think I got that bit wrong.
I don’t understand what either of those sentences is intended to mean.
The plot corresponds to (b) (area weighted average OLR)/(area weighted average SLR).
I suspect that in reality Δ𝛽 is a stationary value, one falling between the plotted values and that presented through your formulations.
This yields virtual values that are not observed in nature. In reality there is a thermodynamic response.
My intent is to illustrate that the thermodynamic response to the radiative forcing term, such as an increased or different distribution of convection, appears to neutralize any supposed change to 𝛽 by radiative forcing alone. Without considering the thermodynamic response, any derivations of opacity or what-have-you are incomplete.
One must also be careful with their conceptualizations of the empirical term 𝛾: “temperature variation emissivity boost” factor. The dynamic atmosphere not only advects heat in the horizontal plane, but also advects and transmits in the vertical. This serves to resist net radiative forcing by IR active gases, in addition to direct planck response at the surface.
The atmosphere does not sit flat and idle in response to forcing by IR active gases, it may become more instable, or change arrangement of dissipation process in numerous ways, instead of increasing surface temperature. This is supported by observation. The specific mechanisms are unknown.
As a PS we must always consider that the surface temperature really has increased, and yet 𝛽 does not follow from forcing by IR active gases hypothesis as a causal mechanism. And so there may be much to discover. That’s what makes this all so much fun.
The question then becomes what mechanisms could cause an increase to surface temperature without a corresponding decrease of the 𝛽 ratio.
Thanks for clarifying.
Not sure what the term “stationary value” means in the context. Do you have a reference or a quick explanation?
Now that I’ve removed he explicit erroneous radiative forcing term, the equations are compatible with radiative forcing potentially not causing a change in 𝛽 after the system responds.
The way my analysis works, all of that is accounted for.
The term 𝛾 reflects the actual distribution of surface temperatures, however it got that way, and yields the actual total surface emissions.
Then, the term 𝛽 reflects the actual transmission qualities of the atmosphere, how a given SLR value yields a particular OLR value. So, the equations ought to be valid regardless of any thermodynamic responses.
Of course, my equations (with the erroneous forcing terms removed), don’t tell us whether or not those effects resist net radiative forcing as you suggest.
One question I have, though, about that hypothesis: The atmosphere allowed the effects of IR gasses to produce around 35℃ of warming. Why, then, would the atmosphere say “enough is enough” and no longer respond to increased levels of IR active gasses and associated net radiative forcing?
My suspicion is that our concepts of the greenhouse effect are relative to the dry adiabatic lapse rate, such that once the temperature lapse due to IR radiation equilibrium exceeds the dry adiabatic lapse rate, instability occurs.
And as long as we’ve been observing, the temperature lapse rate due to the IR radiation equilibrium is always steeper than the adiabatic one. The more IR active gases, the steeper the radiation lapse rate, and the sooner surface dissipation process kicks in.
I don’t claim to have all the answers, but this is how I make sense of it.
Research into “lapse rate feedback” apparently shows that changes in lapse rate do suppress the effects of radiative warming in the tropics, but amplify the effects of radiative warming in polar regions.
The effect of lapse rate feedback globally is apparently thought to produce a net amplification.
my hunch is that is unphysical but that’s all I can offer at this time. It is important to recognize no feedback operates in isolation, so picking out the net effect of one in particular is practically very difficult. It must also be recognized, if we’re honest, that most available literature is undertaken with the aim to justify a LW radiative forcing first and foremost. My approach is to flip the whole thing upside down and to see whether humanity could have any direct impact on non-radiative dissipation at the surface first and foremost. I think this is under-researched. These processes may well impact the net radiation side of surface balance, including both SW and LW parameters. Perhaps there is a happy medium to be found. cheers.
I have a tentative new hypothesis: those thermodynamic adjustments you refer to DO cause radiative forcing to not show up in beta—but to instead show up in gamma, which also causes warming, but in a way that preferentially warms the poles.
Oops. The way that forcing factors in was a last minute addition, and not nearly as carefully though through as the rest of the analysis. I think I got it wrong, when I include forcing as an explicit term in the equations in the way that I did.
Which then invalidates the assertion I made that I could offer an empirical way of measuring radiative forcing. So, I think I was wrong to say the equation I provided offers a way to validate or falsify the hypothesis that radiative forcing is playing a role in changes in Earth’s climate.
So, please discard the terms and words related to forcing. The remainder should be valid.
I continue to work on sorting all this out…
The problem that will never go away is that warming near the surface can result from hindered convective heat transfer up through the atmosphere, as much as that resulting from hindered radiation transfer. In fact, the two are bound in the lapse rate and one is inextricable from the other. Two things can happen with a human caused radiation perturbation, a temperature increase near the surface, or increased convection. Nobody has resolved this dilemma.
The more IR active gases in the air, the steeper the temperature lapse rate, and the sooner instability occurs. As soon as the temperature lapse rate becomes greater than the temperature lapse rate by adiabatic expansion of dry air, convection begins. This delivers large magnitudes of heat and mass higher in the atmosphere to regain balance. In an area with unlimited water this process is endless.
When looking at possible human caused perturbations to these processes, it is reasonable to see if humanity could have any part of limiting available moisture that drives this dissipation process, such as over 5 billion hectares of the eroded and somewhat desertified terrestrial surface. It is in a dry area where the radiation transfer is the limiting process. In a place with ample water, non radiative transfers will dominate.
Let me reiterate what Nelson said.
In a thermodynamic system being analyzed like this you must use cooling gradients. However you look at it, there is a gradient of heat loss from the stove to the final point. If you look at the limits, if you had an insulator that had an infinite resistance, this means the temperature would reach a static equilibrium temperature equal to the stove. It simply can not exceed the temperature of the stove. There is only so much energy available INSIDE the system. For the system outside the stove to get warmer than the stove means you would need more energy from somewhere that would warm the stove.
So what is the conclusion? Cooling is the direction of heat flow, i.e. hot to cold. You can minimize the cooling gradient but you can’t reverse it when dealing with a constant source. If you could reverse it, we would have perpetual motion machines.
Hi, Jim. Please take a look at my rather thorough response to Nelson.
I agree that the temperature of the cabin cannot exceed the temperature of a heat source (if the temperature of that heat source is fixed) and the temperature of the Earth cannot exceed the temperature of its heat source, the Sun.
However, I don’t get how that fact is relevant to my argument.
I’m saying that decreasing the efficiency of cooling will raise temperature. I haven’t claimed that there is no limit to how high that temperature could get.
If the heat source inside a house has a fixed temperature, as the amount of insulation on a house increases, the interior temperature will keep increasing, but the size of those increases will get smaller and smaller as the temperature asymptotically approaches the temperature of the heat source.
Why are you arguing about whether the temperature can become larger than that of the heat source? None of the examples I’m talking about are in that regime:
So, how is your point relevant?
Maybe the problem relates to some nuances in what I previously said.
I did NOT say that the temperature is determined ONLY by the thermal resistance / cooling heat transfer curve. If you thought that was what I was saying, I can see why you might think I believed that increasing thermal resistance could raise temperatures to be absurdly high.
But, what I said is that temperature determined by the thermal resistance / cooling heat transfer curve AND the heating rate.
Maybe it will help if I spell it out more explicitly:
Do you see anything wrong with this argument?
[Side note: if the heat source is a thermal reservoir, f(T) will decrease as T approaches the temperature, Tr, of the reservoir; if the heat source is an electric heater without a thermostat, f(T) will be constant over a wide temperature range; if the heat source is a badly-designed nuclear reactor, f(T) will increase as T increases.]
Once again, I agree with your statement, but fail to see how it is relevant to anything I’ve asserted.
So, in both situations, heat always flows from hot to cold, and there is no risk of a perpetual motion machine.
However, in both situations, it’s true that the equilibrium temperature, T, is the temperature for which the heating rate, f(T), is equal to the cooling rate, g(T).
If the cooling system is changed to make cooling less efficient, so that g(T) shifts to g'(T), where g(T) > g'(T), then balance point shifts and the equilibrium temperature T will increase.
In the case of house insulation, the temperature drop across a wall is ΔT = R⋅P where R is the insulation R-value and P is the rate of heat loss. So, if an electric heater is producing heat at a fixed rate P₀, increasing the insulation R-value by an amount ΔR will increase the temperature inside the house by an amount ΔT = ΔR⋅P.
This logic does not involve any claim that the thermal gradient is reversed or the 2nd Law of Thermodynamics is violated. Both before and after adding insulation, heat flows only from hot to cold.
I don’t have a big quarrel with your posts.
One is a pet peeve of mine about using the term making something warmer. A way too many folks translate that into higher temperatures than normal. It slows cooling, rather than actually causing higher temperatures. A large number of people think the earth is going to go up in flames because of this usage of language. We need to be careful to explain that Tmax temps are not out of control. That is why this post from Willis is important.
Please don’t use the sun’s temperature to discuss heat. That temp only determines the Planck temperature and how much EM is transmitted from the sun. At 1 AU from the sun, the amount of EM power the earth receives is much attenuated.
Thanks for your time in making detailed descriptions of what you have done to evaluate this old climate on earth.
Tsun is very useful in entropy budgets
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2002RG000113
Glad to hear it.
I agree that term “warms” or the phrase “making something warmer” can be ambiguous and often lead to serious misunderstandings in discussions of climate.
In a context where there is a steady supply of heat, a net slowing of cooling actually DOES cause higher temperatures.
That’s exactly what I’ve been illustrating with my examples above.
I’m not sure what you mean by “Tmax temps”?
I admire Willis’s work. But, in my evaluation, his logic often contains serious flaws.
Sometimes I’ve invested heavily in exploring his thinking. For example, I did a lot of modeling and then wrote Thunderstorm World: A Model to Explore Ideas from Willis Eschenbach. That work demonstrated that some of the patterns Willis has noticed are (a) very real and (b) they’re understandable as arising from known physics—and (c) they don’t mean quite what Willis seems to think that they mean.
In the case of this current essay, I again believe Willis’s logic is interesting but seriously flawed. Though, at the moment, I don’t have the time to construct a clear counter-argument and support it with evidence.
I agree that heat flow and temperature are different things, and it can cause trouble if people don’t keep the two ideas as distinct things.
I bring up the Sun’s temperature only when I encounter an argument that says “you can’t make something hotter than the heat source” (which is the argument you appeared to be making). When that sort of argument arises, it seems relevant to me to point out that the Sun is the heat source for the Earth, and there is no risk of anything being made hotter than that.
Do you think it’s inappropriate to mention the Sun’s temperature in that particular context? Maybe your comment about something never getting hotter than the source wasn’t meant to refer to the situation on Earth?
That’s mostly true.
There are situations where the Sun’t temperature matters at a practical level.
If you build a “solar concentrator” which focusses sunlight tightly to heat something at the focus to a high temperature, you can to a significant extent undo the attenuation that results from the Earth being 1 AU from the Sun.
In that particular context, it can be useful to understand that it’s impossible to concentrate sunlight so much that the temperature at the focus of the solar concentrator could exceed the Sun’s temperature.
Similarly, no matter how powerful a Greenhouse effect any planet had, that GHE could never make the planet’s surface temperature as hot as the surface of the Sun.
So, in that sense, the temperature of the Sun can become relevant.
Thanks for engaging in a thoughtful way.
As I mentioned in my other comment, I’d like to use this opportunity to comment on a few older posts.
* * *
In Greenhouse Efficiency, you considered as a metric the “greenhouse multiplier” GHM = SLR/OLR where SLR is surface longwave radiation emitted and OLR is outgoing longwave radiation.
I’ve recently been able to derive rigorous analytic solutions to certain aspects of planetary temperatures based on a similar insight. For what it’s worth, I’ve found it to be more consistently helpful theoretically – if less intuitively appealing – to use the reciprocal, 𝛽 = OLR/SLR, which I call the “longwave transmission reduction” factor.
The main advantage it that one can average 𝛽 geographically in a much more meaningful way than one can average 1/𝛽, at least in terms of the physics I’m unpacking.
* * *
In The Multiplier, first you say:
I think you’re creating unneeded/unhelpful trouble by focusing on “surface warming” (absorbed SW and LW) which in my notation is (Sa + DLR).
The surface energy budget is (Sa + DLR) = (SLR + NRH) where SLR is surface emissions and NRH is non-radiative heat transfer. Rearranging that, SLR = Sa + DLR – NRH. In terms of changes, that’s ΔSLR = ΔSa + ΔDLR – ΔNRH.
The actual claim is that an imbalance of 3.2 W/m² at TOA leads to 1℃ surface warming, which by SB leads to 5.5 W/m² increased surface emissions (if 𝜀=1 and T=288K). [The 1.2 W/m² figure involves feedbacks that take centuries to manifest; 3.2 W/m² is a short-term figure.] Your logic asserts that this “requires” 5.5/0.8 ≈ 5.9 W/m² “at the surface.” If I understand you correctly,, that means you think (Sa + DLR) needs to increase by 5.9 W/m².
To which I might respond… so what?
As I noted above, energy is conserved if and only if ΔSLR = ΔSa + ΔDLR – ΔNRH. So, (ΔSa + ΔDLR ) can get as large as you like, as long as ΔNRH also changes appropriately, and there will be no energy imbalance.
By focusing on (ΔSa + ΔDLR) I think you’re creating the appearance of a problem where no actual problem exists.
There is a coherent narrative if one focuses on surface emissions; looking at (SW + LW) absorbed doesn’t lead to any coherent narrative at a theoretical level, as far as I’ve been able to discern.
* * *
IPCC 2021 AR6 WG1 says doubling CO₂ leads to forcing of 3.78 W/m² (p. 945) which leads to a TCR (70-year timescale) response of 1.8℃ (1.4-2.2℃); or an ECS response (requiring centuries of feedback) of 3℃ (1.5-4℃) (p1007).
The number you are quoting corresponds to the ECS, which requires centuries of feedback. The short-term number that would be most relevant is the “Planck response” of 3.2 W/m²/℃. (I recently figured out how to analytically derive the formula for that response, so I understand it reasonably well.)
* * *
As I explained, the near-term response is more like 3.2 W/m² of forcing at TOA yields 1℃ at the surface which leads to increased emissions of 5.5 W/m². In your terms, that’s a multiplier of (5.5/3.2) = 1.7. That’s pretty close to the multiplier that you’ve already acknowledged as existing. (And, there is no reason to expect the marginal change rate to equal the average change rate, so that can and does account for some difference.)
* * *
In A Serious Question, you write:
All that matters is the “lost as longwave radiation” value, which corresponds directly to temperature. Any value for that can be accommodated, from an energy conservation viewpoint, by adjusting the other flows. There is no physical mechanism to enforce anything like a fixed ratio between the amount of “absorbed radiation” and what is “lost as radiation.”
I may have missed an IPCC ECS range revision in recent years, but I do not recall a range that matches the one in your comment. And the ECS range you cited: +1.5 to +4.0, would not seem to average +3.0
IPCC 2021 AR6 WG1 says doubling CO₂ leads to forcing of 3.78 W/m² (p. 945) which leads to a TCR (70-year timescale) response of 1.8℃ (1.4-2.2℃); or an ECS response (requiring centuries of feedback) of 3℃ (1.5-4℃) (p1007).
This is the current IPCC and CMIP6 ECS range:
The current ECS range from the IPCC AR6 report (2021) is from +2.5 degrees C. to +4.0 degrees C.”
“The current ECR range for the latest CMIP6 batch of climate computer games (2022) s even larger, from +1.8 degrees C. to +5.6 degrees C.”
Thanks. You’re right, my comment reflected a typo. The ECS listed by AR6 WG1 (p.1007) is: Central value 3℃; likely range 2.5-4℃; very likely range 2.0-5.0℃.
I think they should add more decimal places:
Howz about +2.5637 to + 4.2861 C. ?
If you’re gonna guess, why not four decimal places?
Hi Willis, one more point to respond to.
I had failed to track that your Greenhouse Multiplier was defined as (in my notation) SLR/Sn where SLR is surface longwave emissions Sn is net absorbed sunlight.
I’ve been using 𝛽 = OLR/SLR as my metric (where OLR is outgoing LW radiation), and i had mistakenly thought your metric was 1/𝛽.
So, it’s interesting to see your post A More Accurate Multiplier. In that post, you adjust for advection. Though I’m not 100% certain how you did that adjustment, the ways that I would think to do that would lead to you basing your multiplier on an “input” that is numerically equal (or close to equal) to OLR. So, it appears that you rediscovered the metric I’ve been using (or the inverse of that metric).
For climate engineering this means one would need cloud engineering across the whole globe. Green alarmists are living in cloud cuckoo land.
Is this article saying the following?
There is global warming
The warmer troposphere holds more water vapor
More water vapor in the troposphere leads to more clouds
More clouds block some incoming solar energy
As a result, changes in clouds should limit the water vapor positive feedback effect of a warming troposphere
This would be a logical explanation for why there were limits to global warming trends in the past 4.5 billion years — clouds saved us from runaway global warming from a water vapor positive feedback.
Over a decade ago that was my theory for why our planet never had runaway global warming, even when CO2 levels in the troposphere were up to 10x higher than today.
The reason we don’t have runaway warming on earth is because 1) we have oceans and 2) the planck response is larger than total feedbacks, by a factor of about 1.5x or so.
The reason we don’t have runaway global warming is because the Greenhouse Effect is bollocks dreamt up by climate change alarmists and lukewarmists.
That’s not a rational response. Do you have anything reasonable to contribute? The Planck response is about 3.2 W/m^2/C. Total feedbacks are about 2.1 W/m^2/C, give or take. Whatever the precise value, it’s smaller than the Planck response, so temperatures stabilize.
While your description of the Greenhouse Effect is somewhat true, I don’t think it was “dreamt up” so much as a misunderstanding of the energy flow in the atmosphere and radiation models. It all looks rational from a static viewpoint.
In reality, once saturation of surface radiation frequencies occur, then there’s no more energy to warm the planet. The claim that additional CO2 reduces IR flow to space is based on seeing increases in downwelling IR from the models. However, this is only part of the process.
When you increase CO2 you must also increase upwelling IR. And, the increase must be higher than the increase in downwelling IR due to the changing density of the atmosphere as up go up. Fortunately, this cooling effect is balanced by slightly enhanced absorption due to pressure broadening,
The net of this is the temperature is almost completely independent of CO2 concentration.
The warm temperatures so evident over the region encompassing Indonesia, Philippines and Papua New Guinea is the antipode to the Chicxulub impact event.The primary heat source is geothermal.
Willis,
Using your Figure 1 I did a quick plot of % cumulative surface vs temperature. It is a bit crude because the bin size is 10C but it certainly does suggest there is no need for complex climate models to determine temperature.
I also did a PDF.
https://freeimage.host/i/Zt4exs
Willis,
I would like to repeat my two graphs with higher resolution if the raw data for your fig 1 is available. By year if possible to capture trending.
The PDF especially because it is so skewed. Assumption of a normal distribution may have skewed results.
PM email if convenient
Thanks, Ferd
Added a trendline to your data. Almost perfect fit with poly order 4. While this might simply be S-B, would expect to see cos(lat) solar.
Willis,
The most obvious response to me is that you did not calculate or estimate ANY feedback in any meaningful sense of the word. A feedback by definition is how much a warming signal is amplified or dampened by some feature of the climate system – in this case, clouds. So if GMST increases by 1 C, then a positive feedback of 0.2 W/m^2/C would amplify that signal by 0.2 W/m^2.
What it looks like you did is plot CRE with temperatures inferred from the CRE – a simple S-B calculation – to plot 21 year averages of CRE with inferred temperatures. It shows how CRE changes with temperature inferred from CRE regardless of the GMST anomaly. It does NOT calculate to what extent a global warming signal is dampened or amplified by clouds. At the very least, for that you’d need two curves at two different GMST anomalies to see how the curve CHANGES with increasing GMST. You didn’t do that. More interesting would be two plots, one of 1980-2000 and one of 2001-2021 and then compare the two.
In an actual estimate of the cloud feedback published in a scientific journal, this is what you you get. “We are able to constrain global cloud feedback to 0.43 ± 0.35 W⋅m−2⋅K−1 (90% confidence), implying a robustly amplifying effect of clouds on global warming and only a 0.5% chance of ECS below 2 K.” https://www.pnas.org/doi/full/10.1073/pnas.2026290118
Scott J Simmons October 15, 2022 1:25 pm
That is exactly what I did. I showed that as temperatures increase by 1°C, over part of the temperature range clouds have positive feedback (warming) and over the rest of the range they have negative feedback. I also gave the global area-averaged value of that feedback.
I did nothing of the sort. I did NOT infer temperatures from CRE, nor do I think that’s even possible.
Instead, I calculated the temperatures from the upwelling surface longwave. However, the graph does not change in any meaningful amount if I use the Berkeley Earth temperatures instead.
Clearly, you didn’t understand what I did. Which is not a problem … until you try to lecture me based on your misunderstandings.
Finally, you say:
Yes, I know that. It’s a central tenet of their alarmism, because without that, their house of cards collapses. AS I SAID IN THE HEAD POST:
You’re free to take that on faith. After all, they “use data of 52 GCMs from the Coupled Model Intercomparison Project phases 5 and 6 [CMIP5/6]” and who can possibly argue with 52 climate models that don’t even agree with each other?
I trust my analysis of the actual observations over any number of climate models, but hey, that’s just me …
That’s not what you did. You did NOT calculate how the curve you plotted would be altered by a 1 C change in GMST. You plotted how CRE changes with 1 C changes in inferred local temperatures. That’s not a feedback calculation. A feedback is in units of W/m^2/C, where C is a change in equilibrium GMST.
You could have used Berkeley Earth, but instead you used the S-B equation. But again, the X axis is still local temperature, NOT a change in GMST, so it’s not an argument against how the cloud feedback amplifies warming from a change in GMST.
You didn’t supply any shred of evidence that there’s a collapsed house of cards. You plotted a graph of CRE with inferred local temperatures. You did NOT show how that graph would be altered by a change in GMST, which is what you’d need to do show that a cloud feedback is negative.
But your analysis simply was not an analysis of any cloud feedback. It was a simple plot of temperature from the S-B equation with CRE. It was not an estimate of a W/m^2 change per 1 C increase in equilibrium GMST.
From the paper “Here, we develop a statistical learning analysis to calculate an observational constraint on global cloud feedback that significantly improves on previous estimates and does not require high-resolution simulations or observations.”
In this case, laughably, they use no science whatsoever. Past performance is no predictor of future performance. And yet this is what they claim constrains the result.
“ Clouds can either warm or cool the surface.” It takes 540 calories to change one gram of water at any temperature to one gram of water vapour at the same temperature, (latent heat.) As moist air rises it cools and the relative humidity increases. At the cloud condensation level, the vapour in the cloud condenses to liquid water, releasing 540 calories per gram. All this heat raises the temperature at the cloud condensation level. The rate of heat transfer from a warmer body to a colder body is proportional to the temperature difference between them.The Earth cools more slowly under cloud than under a clear sky. Cooling more slowly is not the same as warming up. To warm the surface, the cloud would have to be warmer than the surface. This would violate the principle of conservation of energy.
That’s not how conservation of energy works. If you put on a jacket, its temperature is less than 98.6 F, and yet it keeps you warm. It slows down the rate at which IR light escapes your body. In climate system, the GHE can keep the surface warmer than its effective temperature without violating conservation of energy.
I think this is a big deal.
<I> Well, each blue dot in Figure 1 represents a 1° latitude by 1° longitude gridcell somewhere on the earth’s surface. </I>
Doesn’t a “gridcell” represent more square miles near the equator than near the poles? And, so, when you say, <i> Two-thirds of the globe has cooling cloud feedback, </I> I wonder if that’s 2/3rd of the grid cells or 2/3rds of the area.
a 1° latitude by 1° longitude grid cell is 60 nautical miles on the north-south axis. the east-west dimensions are cos(latitude) x 60 nautical miles. this gives a reasonable approximation of what is otherwise a complicated subject.
pouncer, you’re correct about the differing size of the gridcells. I take that specifically into account via area-averaging.
And when I speak of “two-thirds of the globe”, that always means actual area, NOT 2/3 of the gridcells.
w.
Willis writes “In warmer areas, on the other hand, clouds cool the surface. And when the temperature gets above about 25-26°C, cloud cooling increases strongly with increasing temperature. In those areas, for each additional degree of temperature, cloud cooling increases by up to -15 W/m2. This is because of the rapid increase above 26°C in the number, size, and strength of thermally driven thunderstorms in the warm wet tropics.”
Also I think its worth noting you’re effectively hiding the incline by ignoring those few, presumably land based areas that show an average surface temperature above about 35C. The graph, if it was plotted, looks to reverse and go suddenly back to about/above 0 W/ms2 CRE. This appears to be around 15% of the earth’s surface too so not at all trivial.
Better to be thorough than evasive.
Tim, I am not hiding one damn thing, nor am I “evasive”. Those are scummy, scurrilous personal attacks that unfortunately you seem to specialize in. I invite you to gently place them a goodly distance up the far end of your alimentary canal.
You will be thoroughly ignored by me until such time as you can ask a scientific question without including any more of your false vile invective.
I am an honest man. It seems you are unfamiliar with the breed. I don’t hide things. I don’t evade questions. I tell the truth as clearly and transparently as I can, and I work to answer all cogent interesting questions.
But I’m not here to be your damn whipping boy, in that regard you are more than welcome to osculate my fundamental orifice.
w.
Instead of being overly sensitive you might have explained why your first graph doesn’t go all the way to the right and cover the last 15ish% of the planet.
But your response says it all really.
Tim, without a scrap of evidence you claimed I was hiding facts and evading issues. I’m an honest man, and I won’t stand for some charming anonymous internet humanoid like you making false accusations about me. I tell the truth as best I know it, and I admit my mistakes when I make them.
And now you have the insufferable arrogance to claim I’m the bad guy for objecting to your nasty behavior?
Really? You insult me without either reason or evidence, and I’m wrong for pointing out that you’re acting like an ℁soyl?
Like I said, if you want to discuss the science I’m your man. Stick to the science, leave the insults out, we can make it work.
But I won’t be your whipping boy in the process. Maybe that works on your friends. It doesn’t work on me. Keep a civil tongue in your head, and we can talk science.
In hopes of better days,
w.
Willis writes “without a scrap of evidence you claimed I was hiding facts and evading issues.”
Without a scrap of evidence? You didn’t plot the graph for what appears to be about 15% of the planet. Its right there in plain sight.
Twice now you’ve had the chance to explain why but instead have gone off at me because I’ve called you out for failing to plot all the data.
Tim, I don’t generally explain things to people who start out by insulting me.
However, I’ll make an exception. You said:
Yes, there are gridcells with 21-year average temperatures above 35°C.
They make up 0.009% of the earth’s surface.
0.009% …
NINE THOUSANDTHS OF ONE STINKIN’ PERCENT, and you’re making slimy false allegations about me because my graph didn’t include that meaninglessly small part of the planet?
Piss off, and don’t come back until you can keep a decent tongue in your mouth.
w.
Three things
The plot finishes half way through the 33% section of the earth so you need a better graph if that represents 0.009% of the earth.
Like it or not, you hid that data in your graph which should have shown it even though it would have visually weakened your argument on thunderstorms.
I’m the one who should be offended, not you. You’re a nasty piece of work at times.
Willis writes “Yes, there are gridcells with 21-year average temperatures above 35°C.
They make up 0.009% of the earth’s surface.
0.009% …
NINE THOUSANDTHS OF ONE STINKIN’ PERCENT, and you’re making slimy false allegations about me because my graph didn’t include that meaninglessly small part of the planet?”
Actually the graph stops at 30C, not 35C so its a LOT more data that’s been hidden. Tell me, why are you claiming you only missed the data above 35C?
Ah, my bad. The amount above 30°C is 1.0%.
w.
Okay…so…Water Vapor is the strongest greenhouse gas (as has been exclaimed here and as has been widely known since—-you know—-atmospheric physics) AND it regulates temperature almost immediately…..
As opposed to the depths of the oceans eating the CO2 induced warming!
Okay….okay….so how can they regulate water vapor and make money from that boogeyman?!
My basic observation is there are two kinds of clouds. Clouds over land, and clouds over sea. Perhaps more likely, clouds transported over dry land, and clouds evaporated from the local sea.
My growing up in low humidity Northern California, and my time in low humidity interior Alaska my observation is that terrestrial clouds over a low humidity land provide insulation between terra and sky. At night, this means terrestrial clouds act as a blanket to hold the heat over the land, keeping the land from radiating that heat to the clear cold night sky. In the day, terrestrial clouds over dry land shield the land from warming, thus keeping the land cool.
Dear Willis,
Very interesting and thought-provoking. Excellent article.
One question: you mention that these data come from grid cells of 1 degree latitude by 1 degree longitude. One degree of latitude is always the same; 60 nautical miles. However, 1 degree of longitude varies, from 60 nautical miles at the equator declining steadily with latitude until, at the north and south poles, 1 degree of longitude approaches zero.
So, do the grid cells have constant area? Or do they decrease with increasing latitude?
Best Regards,
Dale McIntyre
Thanks, Dale. The gridcells are indeed different in area, so it’s necessary to area-weight all averages.
Best regards,
w.
Hi Willis. After offering a few comments on older posts, I’m ready to address the current one. As is often the case, I find your CRE scatterplot analysis admirable, intriguing―and flawed.
I believe that your core premise relies on invalid logic, and that your empirical test offered no clear evidence to validate that premise.
In what follows, I’ll offer specific explanations of both these assertions.
This is the premise I’m referring to:
I’d like to explain:
1) A gridcell-by-gridcell analysis does not include all the possible feedbacks because every gridcell functions in a particular global context; it there is truly global warming, then that will put gridcells into a context that none of them has ever experienced before. That’s a general statement, but I can spell out exactly what I mean. I’ll risk being verbose so that you can follow my logic without the with a reduced chance of it not making sense:
Thus, another gridcell with the target temperature we’re moving towards will NOT in general have been operating in equivalent conditions. So, it’s NOT logical valid to assume it can provide a valid prediction.
Your procedure seems based on the premise that atmospheric behavior depends only on absolute temperature. But, that’s not true. Thermodynamic behavior of the atmosphere also depends on relative temperatures, not just absolute temperatures. After a global temperature change, there will be brand new combinations of relative and absolute temperatures. And that invalidates the logic of your process.
2) Regarding your empirical test… I’m glad you thought to try to do a test, but at the level of what you’ve shown us (up close the data might reveal more) the test is NOT sensitive enough to assess the validity of your hypothesis.
In short, simply eyeballing the curves and saying they look “really close” is highly misleading and proves almost nothing regarding your argument. Without careful examination for differences as small as even 0.1℃ of shift to the right, and better yet, a test of statistical significance, I think you’re fooling yourself if you think your well-intentioned test has thusfar validated your hypothesis.
If you do investigate whether or not you can rule out key parts of the curve having shifted, I’d be very interested to learn what you discover.
It’s been fun to think about this.
Thanks for your consideration, and for your ongoing investigations.
Thanks, Bob. You say the analysis method is flawed because it’s not taking into account a host of things other than available solar power that affect the temperature.
But in fact, I’ve included every one of those other things. For example, selecting for where the temperature is 20°-21°C, we get a range of gridcells from around the planet, hundreds of them. The only thing they have in common is the temperature. They have different clouds and seasons and all that stuff.
And despite that, they don’t appear randomly at any value. Instead, there’s a complex but very clear relationship between temperature and CRE shown in Figure 1—a pattern that includes all known variations of the other variables affecting the CRE.
Regarding your second question, you point out that it might shift to the right or left by 0.4°C. I agree, but it’s what I call a “difference that makes little difference”.
Remember, I’m looking at the whole world. What’s of interest to me is the overall shape, the weighted average, and the way it drops almost straight down at the highest temperatures. Your shifts don’t affect those questions.
w.
True. But, unfortunately, those are the wrong questions to validate your hypothesis.
To the contrary. Your intuition might tell you tiny shifts like that aren’t important. But math says that such a shift would entirely destroy the validity of your thesis.
Suppose that for a segment of your LOWESS smooth curve, it’s well fitted (in some units) by:
CRE = +2⋅(T – 10)
You have a grid cell at T=11℃ in 2003. It’s CRE value is in the center of the distribution, so CRE = + 2⋅(11 – 10) = +2.
By 2018, the temperature of that grid cell has risen by 0.4℃ to T=11.4℃
You predict CRE will change by ΔCRE = 2⋅(0.4) = 0.8
BUT, an undiscernable shift of 0.5℃ happened between your first curve and your 4th.
So, NOW, the new curve for CRE is
CRE = -2⋅(T – 10.5)
For our gridcell with a new temperature of T=11.4, that means
CRE = -2⋅(11.4 – 10.5) = 1.8
That means ΔCRE = 1.8 – 2 = -0.2.
Your method predicted ΔCRE = 0.8 but actually ΔCRE = -0.2.
That tiny, undiscernable shift has the power to make your predictions entirely wrong!!
* * *
With respect, I think you are engaging in what I call “sloppy verbal reasoning.” In my experience, that’s a practice that can justify just about any conclusion a person sets out to obtain. What it’s not good at is identifying reality.
I don’t trust my own sloppy verbal reasoning or anyone else’s, because it’s so extremely likely to be wrong. Part of the problem is that such verbal reasoning is typically full of ambiguous vague concepts, unexamined implicit assumptions, and unjustified logical leaps. When I was younger I did a lot of mathematical proofs, and I learned how tiny a logic error it takes to make one’s conclusions worthless.
So, when I want to check some logic that matters, I try to get precise about meanings, and if at all possible put what I can of it into some mathematical form, to see what that reveals.
* * *
You say that for those gridcells, “The only thing they have in common is the temperature.” I guarantee you that’s not true, though it might seem that way superficially.
The statistician in me is screaming at how obviously and wildly erroneous that logic is.
Let me offer both a concrete argument and an abstract argument. The abstract one might be harder to “get”, but to me it’s compelling.
* * *
A] Here’s a concrete argument. NONE of those gridcells have, in the last few centuries, experienced an atmosphere with a CO2 concentration 2X higher.
That increased CO2 concentration is NOT simply equivalent to increasing sunlight. CO2 introduces a dynamic where every time the lapse rate changes, that changes the efficiency of emissions from the surface reaching space. That’s NOT something that happens when sunlight increases.
None of your data for gridcell behavior in recent decades includes gridcells which have experienced a lapse-rate vs. heat transfer dynamic that strong!
* * *
B] Here’s an argument that may seem abstract, but which I’m deadly serous about:
The problem is, all those gridcells are drawn from the same probability distribution for properties. If something like radiative forcing changes the probability distribution, then all bets are off. Your prior information is potentially worthless.
When you change the probability distribution, then that “complex but very clear relationship between temperature and CRE” WILL shift. That is almost a certainty, for the vast majority of possibly changes in probability distributions.
The trend you’re seeing isn’t just about the grid cells you’re looking at but the probability distribution they are drawn from.
As someone who messes around with probability and statistics from time to time, it’s as clear as the nose on my face that it’s astronomically unlikely that your hypothesis could turn out to be true.
You’s acting as if all probability distributions were the same, just because it seems like the grid cells you look at from the current distribution subjectively seem to cover “the full range of possibilities.” That can’t possibly be true. Different probability distributions lead to different outcomes, and I can’t imagine an argument that would convincingly justify an assertion that the probability distribution won’t change.
As an example, maybe that “very clear relationship” you discovered for some segment has the form: CRE = 5 + 0.01⋅T ±2.
But, unbeknownst to you, that’s because the 5 in that formula is is really 0.1⋅X where X was a variable that used to have an average of 50 and a range of ±20.
Now, after radiative forcing, X averages 80±20. So now,
CRE = 8 + 0.01⋅T ±2.
Suddenly, everything you thought you know about CRE and T is wrong.