How Thunderstorms Beat The Heat

Guest Post by Willis Eschenbach

I got to thinking again about the thunderstorms, and how much heat they remove from the surface by means of evaporation. We have good data on this from the Tropical Rainfall Measuring Mission (TRMM) satellites. Here is the distribution and strength of rainfall, and thus evaporation, around the middle of the planet.

trmm annual average evaporative coolingFigure 1. Evaporation in W/m2 as shown by rainfall data from the TRMM. It takes about 80 watt-years of energy to evaporate a cubic metre of water, so a metre of rainfall per year is equivalent to an average surface cooling of 80 watts per square metre. The TRMM satellite only covers from 40° North to 40° South.

I have held for some time that the global surface temperature is restricted to a fairly narrow region (e.g. ± 0.3°C over the 20th century) by the action of emergent phenomena (see references at the end of the post). Chief among these emergent phenomena are tropical thunderstorms. My hypothesis says that when the tropical surface temperature goes over a certain threshold, that thunderstorms emerge to put a firm cap on the temperature by cooling the surface.

Thunderstorms cool the surface in a number of ways, but the main cooling method uses the exact same mechanism used by the refrigerators that keep our food cold. Thunderstorms use a standard evaporation/condensation cycle. In one part of the cycle the working fluid evaporates, cooling the surroundings. In another part of the cycle in another location, the working fluid condenses. For a refrigerator, the working fluid used to be some form of Freon, nowadays it’s some other fluid. For thunderstorms, the working fluid is water. When it evaporates at the surface, it cools the local area, and the heat is moved from the surface to the clouds and on upwards.

Now, for my hypothesis to be correct, the number and intensity of thunderstorms needs to increase quickly as temperatures go above a certain temperature threshold. In addition, the change in the resulting evaporation needs to be quite large in order to successfully control the system.

With that in mind, I made a scatterplot of sea surface temperature versus thunderstorm evaporative cooling. Figure 2 shows that result.

scatterplot oceanic evaporation vs temperatureFigure 2. Scatterplot of 1° x 1° gridcell annual average ocean-only thunderstorm evaporative cooling on the vertical axis, in watts per square metre (W/m2) versus 1° x 1° gridcell annual average sea surface temperature on the horizontal axis. 

As you can see, the thunderstorms are clearly functioning to cap the temperature. When the ocean surface gets hot, thunderstorms form and exert immense cooling power. There are some points worth noting about Figure 2.

First, the red area shows the tropics. This part of the earth is important because it is not only about 40% of the planet’s surface. In addition, just over half of all the energy absorbed by the surface of the earth is absorbed in the tropical regions of the planet. As a result, the regulation of this large amount of incoming energy by albedo control is crucial to the overall energy balance of the planet.

Next, these are annual averages. However, they are also daily averages. But during the day/night cycle in the tropics, the evaporation is by no means constant. At night the evaporation is small, some tens of watts per square metre. During the day, on the other hand, evaporation is quite large, hundreds of watts per square metre, because the strong tropical sunshine evaporates the water directly, plus the thunderstorms are largely a daytime phenomenon. This means that the peak hourly thunderstorm evaporative cooling is on the order of twice the average values shown above, up to about 600 W/m2.

Next, the cooling effect of the thunderstorms is not applied blindly or randomly. The thunderstorms only form as and when the local area is above the temperature threshold. This means that the cooling effect, which can be up to 500-600 w/m2, is always located where it is most needed, on the local hotspots.

Finally, here is the most important consideration. The timing and the amount of thunderstorms are NOT a function of greenhouse gas forcing, or of solar forcing, or of volcanic forcing, or of any other kind of forcing. As Figure 2 shows, they are a function of temperature. As long as the surface-atmosphere temperature difference is large enough, thunderstorms will form even at night when the radiative forcing is quite small.

This means that their effect will be to maintain the same temperature, regardless of reasonable-sized fluctuations in the amount of forcing. Clouds don’t know about forcing, they form and disappear based on local conditions.

And this is simply one more piece of observational support for my hypothesis that emergent phenomena regulate the temperature and maintain it within a fairly narrow range.

My best to everyone. Here in California, we have about 150% of the usual snowpack in the Sierra Nevada mountains. When it was drought, it was said to be the result of global warming … and of course, now that there is heavy snow, that is also said to be the result of global warming.

Buckle your seatbelts and keep your hands inside the vehicle, it’s gonna be a long, uphill struggle to get rid of this madness …

My best to all,


My Usual Request: If you disagree with me or anyone, please quote the exact words you disagree with. I can defend my own words. I cannot defend someone’s interpretation of my words.

My Other Request: If you think that e.g. I’m using the wrong method on the wrong dataset, please educate me and others by demonstrating the proper use of the right method on the right dataset. Simply claiming I’m wrong doesn’t advance the discussion.

Some Of My Previous Posts On The Subject:

The Thermostat Hypothesis

Emergent Climate Phenomena

Cooling and Warming, Clouds and Thunderstorms

Tropical Evaporative Cooling


TRMM Data is here, see the bottom of the page for the NetCDF file.

CERES Data is here, I used the EBAF dataset.

5 1 vote
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
January 8, 2016 11:08 am

Hi Willis.
A small point, but you said “… thunderstorms only form as and when the local area is above the temperature threshold.”
I would agree with this statement much more strongly if the word “only” was replaced with “largely”, “mostly”, or “primarily”.
Because there are other mechanisms for thunderstorm formation that purely by surface heating. Some of the others are converging surface winds, cold core lows aloft, frontal boundary lifting, and orographic lifting.
In each of these cases, it is likely true that the storms that do form will be more numerous, or longer lasting, or rise to greater altitudes when the surface air which is entrained into them starts off warmer or more humid (because in both of these conditions there is more energy available).
Thunderstorm height may be very important in large scale global feedbacks, as the higher the heat from the surface is transported, the more readily/completely it may be transferred to space.
Thanks for another interesting article!

Reply to  Menicholas
January 8, 2016 11:17 am

Indeed, Willis and Menicholas, like a pan of boiling water on the stove, adding more energy does not change the temperature of the water. The energy is still in the system, though. Only radiation of photons to space can truly cool a “global” temperature.

Reply to  Bernie
January 8, 2016 11:38 am

Not just radiation but also solar reflection from cloud tops which should be greater.

Reply to  Bernie
January 8, 2016 12:12 pm

I guess the argument is that thunderstorms transport heat from the surface to high in the atmosphere, where the greenhouse effect is less.

Reply to  Bernie
January 8, 2016 12:47 pm

but if the pan ISN’T boiling, then adding more energy DOES change the water temperature…and last time I checked, we didn’t have any boiling seas…so what was your point?

Reply to  Bernie
January 8, 2016 12:53 pm

pete, I didn’t intend for you to interpret my comment as an analogy. It’s just a familiar example of a natural mechanism that provides limiting behavior through a discontinuity.

Richard Petschauer
Reply to  Bernie
January 8, 2016 1:30 pm

Yes, but latent heat from condensation during cloud formation warms the clouds which increases the radiation to space. The warming is evident with the warmer part of the cloud rising and producing the cloud tops. And at this altitude, the low water vapor content means the atmospheric window passes about 75% of IR radiation to space, up from about only 25% from the surface on a clear day.

george e smith
Reply to  Bernie
January 8, 2016 4:47 pm

Richard Petschauer
January 8, 2016 at 1:30 pm
Yes, but latent heat from condensation during cloud formation warms the clouds which increases the radiation to space. …..”””””
I’m of the opinion that latent heat does NOT warm diddley squat.
Water vapor molecules floating around in the atmosphere have more energy, than water molecules floating around in the ocean; that extra energy is the latent heat of vaporization.
Those atmospheric water vapor molecules are going to remain as water vapor molecules so long as they retain that extra energy. As the water vapor rises, the ambient Temperature falls and since the water vapor molecule is in collisions with the air molecules, that extra energy is thermalized and distributed around to the other air molecules, which are in turn in collision contact with even colder molecules at higher altitudes.
Only after all that excess (latent) heat is extracted, by air molecules which remain colder than the water vapor molecule (in terms of where its energy is on the Maxwell Boltzmann distribution tail) will the water vapor molecule be able to deposit on some substrate with other H2O molecules and form liquid water, or ice crystals.
The surrounding atmospheric air molecules never become hotter than the water vapor molecule or else it will stop cooling and losing its latent (excess) heat.
I don’t know where the notion that condensing water vapor hits the atmosphere with a blast of heat and warms it comes from.
When steam from boiling coffee hits your skin, and scalds you, it does dump a lot of heat into your skin (590 cal/gm) but it never raises your skin temperature above that of the steam. Your skin is at 37 deg. C (98.6 deg. F but if the steam is at 70 deg C you will really feel it even though your skin will not reach 70 deg’ C.

Reply to  Bernie
January 8, 2016 7:02 pm

George says:
I’m of the opinion that latent heat does NOT warm diddley squat.
…..As the water vapor rises, the ambient Temperature falls and since the water vapor molecule is in collisions with the air molecules, that extra energy is thermalized and distributed around to the other air molecules….

It seems to me something is getting warmed. Not the water vapour in the cloud tops, but the other gases which must be mixed in there.

Robert B
Reply to  Bernie
January 8, 2016 8:03 pm

George E Smith,
In very small droplets of a few molecules, these are the surface molecules of the droplet so even if large droplets of water should spontaneously form, they don’t unless nucleated by the presence of something that can make the droplet stable while still small.
The gaseous water molecules don’t lose thermal energy to the surrounding air to become liquid molecules of water. The initial small droplet is much warmer than the air and like the heat transfer from surfaces to air, most of it is through emission of LWIR. The droplet will then grow with warming of the droplet when another molecule condenses and that energy being lost through emission of LWIR.

Reply to  Bernie
January 8, 2016 9:22 pm

Not sure how to reply to George E Smith’s reply, but latent heat release when water condenses out of a rising airmass and removes water from the cloud does warm the local air parcel and the best example of this is the fohen wind blowing over mountain ranges. Water removed from the air on the windward uplift side warms the air compared to its surrounding environmental air mass. When this air falls back down the lee side the air warms anomalously to its environmental airmass. I experienced this first hand just a couple of weeks ago driving from Christchurch NZ to Mount Cook. Christchurch broke its December temp record 36 deg C. Yet when I arrived in Mount Cook cloud was streaming over the mountains and evaporating rapidly as it decended with occasional rain.

Reply to  Bernie
January 9, 2016 6:55 am

Search for comment below by gymnosperm.

george e. smith
Reply to  Bernie
January 10, 2016 1:35 pm

I have described one view of the evaporation process (of water) several times at WUWT; I’ll try again.
My thought experimental setup consists of H2O molecules in pure water. The possible presence of any other sort of molecule simply complicates things, so this experiment assumes there are none such.
A molecule in the bulk, a m, cm, mm, whatever deep is attracted on all sides by other water molecules for some reason. I called it Van der Waals forces; Phil has told us it is Hydrogen bonds. I don’t understand what Hydrogen bonds are, although I have heard and read of them; I’m not a chemist. I’m happy to accept Phil’s statement. so H bonds it is.
So an individual H2O molecule has no net tendency to go anywhere, and they vibrate around, with an average kinetic energy per degree of freedom, that is determined by the water Temperature.
Those water molecules are much closer together than are the air molecules immediately above the water surface, which are also at the same Temperature as the water if (heat) energy is not being transferred between water and air.
At the water/air interface surface, the H2O molecules have neighbors around and below them, but not above, so there is a net downward pull on a surface H2O molecule by all those hydrogen bonds, and that prevents an H2O molecule from wandering off into space. We call the effect surface tension, and it has the effect of trying to keep the surface flat; whatever flat means at the molecular level.
The KE of the water molecules has some Maxwell-Boltzmann like distribution, which crowds most molecules towards the low energy end of the distribution, with a long decaying tail at the high energy end.
Every now and then, a H2O molecule gets an upward velocity at one of these higher energy values, and pulls away from the surface, breaking its hydrogen bond restraint. This should slow down the escaping molecule, and it might immediately get knocked back down into the water.
The further out of the MB distribution tail, a molecule is, the higher its KE and the greater the probability it will escape with a high enough velocity to get away from the surface.
But the further out on that tail the fewer water molecules there are.
So below some “escape” energy there is low probability of escape, and above some high tail energy there is a low probability of finding many molecules.
So there clearly is some high energy out on that tail, where there are a lot of molecules, and they also have plenty of energy with which to escape.
Consequently, the H2O molecules that escape from the surface tension and drift away from the surface, also have a net KE per degree of freedom that also has a peaked distribution which is at a much higher energy than that of the bulk water molecules.
So the water vapor that is escaping from the water, has a higher Temperature than the water bulk temperature. It might be much higher.
Now the air (N2/O2/Ar) is at the same Temperature as the water because of conduction between them, which won’t allow A HIGH Temperature gradient, so the escaping “steam” finds itself in collisions with air molecules that are much colder, and the water vapor quickly re-establishes thermal equilibrium with the air Temperature, thereby losing all of its excess KE .
By the time the water vapor molecules have moved far enough from the water surface; maybe microns or less, to escape recapture by the water and its hydrogen bond attractions, collisions with air molecules has removed all of its excess energy so it has cooled to the same temperature as the air, which has some Temperature lapse rate with altitude, and as the lighter H2O molecules rise in the atmosphere they too cool and remain at the same Temperature as their local air.
So this process has transferred heat energy from the water and delivered it to the gases of the atmosphere, over and above the mean energy appropriate to the water/air Temperature that would be, in the absence of evaporation, and yes we would describe this as warming/heating of the atmosphere.
BUT! This all happened within microns of the water surface; not at some km altitude.
In the bulk of the humid atmosphere, the H2O molecules are at the same Temperature as the air molecules, and they all cool together as they rise in altitude, at whatever the lapse rate is.
So now what happens if some H2O vapor molecules collide with each other and maybe a few of them. Why don’t those hydrogen bonds make them grab onto each other and remain in contact to form a water droplet.
So now we have to think about that surface tension again.
ST manifest itself as a surface area contracting force of so many newton per meter. A water film that terminates on a wire say one meter long, exerts a pull (per surface) of t newton on the wire, or 2 t for a double sided water film.
If we have a water droplet of radius (r) or a bubble of radius (r) in water, the surface tension (t) tries to reduce the surface area and shrink the bubble or compress the water droplet. Well the hydrogen bonds in the droplet are doing the sucking, that is manifested as (t) but the end result is that the water droplet or the bubble in water, must have an internal pressure that exceeds the ambient pressure either in the ater or the air around the droplet.
How much is that excess internal pressure ?
It’s a 4-H club exercise to show that the excess pressure is 2.t/r newton per meter squared. For a soap bubble with two water surfaces it would be 4t/r.
OOooops ! I see a glitch .
If the water droplet needs an internal excess pressure of 2t/r, then that would require infinite pressure for a zero starting radius droplet.
Now the evaporating water surface was mostly flat, so the internal excess pressure is zero, but to form a coalescing droplet at near zero radius, requires a much higher excess pressure, so it is much easier to evaporate than to condense. In particular the water vapor temperature has to get much lower to reduce the tendency of the intra-molecular collisions KE to just blow the molecules apart.
I’m afraid, I still don’t see how H2O molecules in the gaseous phase can remain well above their local air Temperature and convey excess energy to some high altitude. It seems like they did all their warming (of diddley squat) within microns of the evaporating surface.
But I’m eager to learn differently from someone with a better understanding of it than I do.

Reply to  Menicholas
January 8, 2016 12:33 pm

Menicholas, my thoughts exactly. Plus the effect of wind blowing the evaporated moisture around, sometimes thousands of kilometres away. It doesn’t just rain on the ocean. The evaporation cools the ocean, and the clouds and rainfall can cool land as well, or suppress evaporation over water in other areas. Another factor limiting thunderstorm development would be higher pressure areas, when the surface cools via evaporation but the moisture is blown somewhere else and few clouds form locally. This is seen in the dry season in monsoonal regions (at least north Australian waters) and even in the supposedly wet season in El Nino years.
As always, a stimulating article. Thanks, Willis.

Reply to  Menicholas
January 8, 2016 12:55 pm

It’s also not entirely true that increased SST will cause an increase in thunderstorms. This is seen during El Nino, when the cloud cover actually decreases in the eastern Pacific, and a decreases overall worldwide. The opposite is true for La Nina.
I think this has to do with El Nino reducing nutrients in tropical waters, restricting plankton growth (especially diatoms), which in turn reduces dimethyl sulfides in the air, reducing the nuclei for cloud formation.

john harmsworth
Reply to  RWturner
January 8, 2016 3:09 pm

Not sure I agree with this analysis. Just because sea level evaporation increases doesn’t mean that cloud formation will increase in the same region, as that will depend on RH/ temperatures at higher altitudes as well as prevalent winds at different altitudes carrying away the water vapour that results from the surface evaporation. e.g.- B.C. is at some distance Northward from the el nino mass but is receiving heavy snowfalls this winter.

Reply to  RWturner
January 8, 2016 7:10 pm

RW Turner, John Harmsworth,
Note Willis has calculated the energy moved from rainfall data.
So, on the proviso the TRMM database is accurate, that water vapour has evaporated, has formed clouds, and has fallen as rain.

Reply to  RWturner
January 8, 2016 9:09 pm

I think you have Nina and Nino cloud amounts reversed. The OLR decreases (cloud increases) over the Eastern tropical Pacific during El Nino because the warmer SSTs induce more convection and cloud cover. Save some satellite pictures this year and compare them with a year from now when the inevitable cooler La Nina Phase kicks in.

Reply to  RWturner
January 9, 2016 11:24 am
I was mistaken by confusing stratus cloud cover with cumulus, which I suppose is the only cloud type that pertains to this essay. There is a negative correlation with marine stratus and stratocumulus to SST, not a negative correlation with cumulus.

john harmsworth
Reply to  Menicholas
January 8, 2016 3:02 pm

Pretty much agree with all this. The way I look at this is essentially, the climate/ weather system of earth is a heat engine. All the movement of air and water is caused by the heat input to the system, with the end result being heat output to space. More heat input mainly accelerates the engine. Water is clearly a primary mode of transport for this heat, and based on what is evident from our knowledge of water evaporation and rainfall, it is obvious that more heat causes more rainfall on a pretty immediate basis. Unfortunately, it is also obvious that the field requires this thorough explanation and even this will probably prove to be insufficient. I’m not a scientist. I’m just a refrigeration tech with a lot of design experience in psychrometrics but I know the heat capacity of humidity would stagger most people and I believed from early on that water vapour was key to this question. This is very good work and needs to be seized upon and stressed until the talking heads acknowledge it’s importance.

george e smith
Reply to  Menicholas
January 8, 2016 3:49 pm

Well I have no disagreement with Willis or his hypothesis.
I have often cited Wentz et al SCIENCE July 13 2007 (I think) “How much more Rain will Global Warming bring ? ” and its consequent presumption of cloud cover modulation and ultimately surface solar energy negative feedback, as a basic mechanism of earth Temperature regulation.
It seems to me that Willis’s process, is a part of that negative feedback loop.
I don’t think earth’s temperature has much to do with GHGs at all (except water).
An atmosphere without water would launch the mother of all radiative forcings

Ray Boorman
Reply to  Menicholas
January 10, 2016 7:16 pm

Menicholas, the storms formed by the mechanisms you mention would be very rare in the tropics, even rarer over the oceans, which is where the cooling effects Willis is describing take place.

Steve McIntyre
January 8, 2016 11:09 am

Willis, what is the response of thunderstorms to the current Nino in the Nino 3-4 area? It seems to max out a little over 29 deg C.

Reply to  Steve McIntyre
January 8, 2016 1:08 pm

Is the “max out at a little over 29C” referring to SST’s in the Nino 3-4 area or Figure 2 in Willis’s posting? If the former, then figure 2 gives a good indication as why SST’s max out at 29C.
I was intrigued by an earlier WUWT article by Willis on SST’s versus thunderstorm activity. Crunched some numbers relating to steam table data and found out that some where between 25C and 30C, the buoyancy for air (inverse of density) at 100% RH went from being driven by air expanding due to temperature (roughy linear) to being driven by the increase in water vapor (roughly exponential). Also noted that the NWS considers SST’s of 26C needed for tropical cyclone intensification.

Owen in GA
Reply to  erikemagnuson
January 8, 2016 2:00 pm

I thought the “max out” value that Willis has used in the past was about 31C. But I may be misremembering

richard verney
Reply to  erikemagnuson
January 8, 2016 3:11 pm

Whatever the physics is, it is universal across the globe, and whilst the majority of the oceans in the tropical regions do not have a SST exceeding 30 to 31degC, this is not because evaporation places such a cap on ocean temperatures (as Willis initially suggested in one of his ARGO posts). It is far more complex and there are many processes at work which act to effectively restrict SST in the majority of the oceans exceeding that temperature. Currents in 3D being the main one, and winds being another, and of course evaporation also plays a part. If evaporation capped temperatures at around 30 to 31degC, one would not see swimming pools with summer temps of around 36degC (my swimming pool often has such a temp in late July), nor would one see any oceans such as the Red Sea, the Gulf of Mexico etc regularly supporting SSTs in excess of 31degC..
However, as I pointed out in the ARGO thread there are dozens of seas and gulfs which regularly have surface temperatures exceeding 32 degC and often up to 34 to 36degC. I recall posting a comment on the ARGO thread where I listed at least half a dozen port authorities currently that day listing highs exceeding 32degC.
That said, the higher the surface temperature the more evaporation is driven, the earlier the clouds begin to form etc.

Reply to  erikemagnuson
January 8, 2016 8:16 pm

richard verney – A single swimming pool is way too small to create a thunderstorm. On land, quarter section plowed fields can generate enough convection to form small cumulus clouds, but it takes a large area to generate the convective flow needed for a major thunderstorm.

Steve McIntyre
Reply to  Willis Eschenbach
January 8, 2016 3:50 pm

that’s an interesting perspective on the Nino area, to say the least.

William Larson
Reply to  Willis Eschenbach
January 9, 2016 10:47 am

Mr. Eschenbach–
“…but the previous El Ninos (1998, 2003, 2007, and 2010) are clearly visible.” I can track, based on the above figure, with this statement regarding the El Ninos of 1998, 2003 (I suppose) and 2010, but 2007? I find your emergent-phenomena concept compelling, to say the least, so I want to agree with everything you say that has to do with it, but my eyes see nothing of note for 2007. Am I out of it, or what?

Reply to  Steve McIntyre
January 8, 2016 9:52 pm

This is what the IRIS interferometer measured from 1100 km over Guam in clear skies vs the top of thunderstorm anvils back in the day. The cloud top temperature tracks the 215K Planck function like it is on rails except for blurbs of higher temperature radiation in the CO2 and ozone bands. Shades of an Antarctic spectrum.comment image
Unfortunately, it seems clear skies do a much better job of radiating to space in the atmospheric window than thunderstorms do, probably because water is such a versatile GHG. Somehow the cloud top temperature seems to be constrained to the lapse rate, as 215K is pretty chilly.
On the other hand, there is evidence that energy gets transmitted to the stratosphere by Ninos, monsoons, volcanoes, etc. We know that thunderstorm anvils occur because they flatten out against the stratospheric inversion. Maybe there are tunnels of conduction there?
Not trying to be difficult but we have to deal with all the data.comment image

Reply to  gymnosperm
January 9, 2016 6:59 am

Is your spectral chart daytime or nighttime?

Reply to  Bernie
January 9, 2016 9:27 am

Think it is daytime. Think the graphic is from Petty. Don’t recall any discussion of day vs night.
Interesting question though. Does the temperature of the stratopause change at night?
I’ve been thinking the spectrometer may be reading some sort of “cool skin” on the cloud that belies the internal temperature. But if that skin is persistent, the effective radiative temperature does not change.

Reply to  gymnosperm
January 10, 2016 5:01 am

Unfortunately, it seems clear skies do a much better job of radiating to space in the atmospheric window than thunderstorms do, probably because water is such a versatile GHG. Somehow the cloud top temperature seems to be constrained to the lapse rate, as 215K is pretty chilly.
I think that there is a subtle error here. Without clouds, what you are ‘seeing’ is of course largely the ground radiating. With clouds, of course its the cloud tops, which of course are pretty cold, the adiabatic cooling haven taken care of that.
215k is not chilly in comparison to deep space, which is what counts for net radiation interchange between two ‘black bodies…’
If cloud tops are not losing heat, and if the radiation is so much better from the ground, what happened to the albedo? Or is that radiating/reflecting in a frequency that that graph doesn’t show? And how come warm wet air goes up and cold rain comes down?
I would absolutely believe those curves as night time curves, because we know that clouds make it warmer at night, but by day? When that spectrum is not just deep infra red, but massively up to the visible and UV.?
I am not questioning the data, but does it mean what you say it does?

Reply to  Leo Smith
January 10, 2016 10:41 am

I’m struggling with that too. As I mentioned there is very good evidence for monsoon and enso energy transfer to the stratosphere. How do they do it if the cloud tops are radiating at a lower temperature than the stratosphere above?
Albedo (reflection) of high energy solar light is not thought to affect the temperature of the reflecting body, but if it did, the effect would be to warm it.
Crazy instant idea: what if the apparent energy transfer to the stratosphere were just the cloud top reflectance forcing the solar SW to make two runs through the ozone?

Reply to  gymnosperm
January 10, 2016 11:19 am

” How do they do it if the cloud tops are radiating at a lower temperature than the stratosphere above?”
The temperature of molecules at high altitude is high, but the density has to be very low, so it doesn’t matter.
From the ground, the ir window reads 100F or more colder than surface temps, which has to include all molecule average temp in that column of air. At cloud tops, there must be more loss to space. I think there should be a strong ir signature from water vapor losing energy to become liquid water, while some must be reabsorbed a lot can dump straight out that window to space.

Reply to  micro6500
January 10, 2016 6:35 pm

The following is a widely known old school spectrum from Petty that was taken with the same instrument from a (U-2?) at 20 km and from the surface of the Arctic ice looking up.comment image
You are right. In the atmospheric window the U-2 sees something like the surface at 288K and the guy looking up sees the chilly Arctic tropopause at like 160K. Bear in mind that ice has very low vapor pressure and surface humidities can be as low as CO2 concentration.
How can this be? These instruments measure photons. Photons radiate equally in all directions. It is easy to see how the upwelling from the surface (if as I suspect that surface is the ice which is a surprisingly good blackbody) is going nowhere but up. However, the radiation at the tropopause goes both ways.
It must boil down to intensity. The surface intensity is very high, but it is running away from the sensor looking up.The tropopause intensity is very low (barely above zero), and winds up being lunch money for the sensor looking through it.
Not sure where this leaves us regarding cloud tops. Maybe the same thing, but the cloud tops are way above zero. Something like 30 radiance units for the cloud tops vs 100 for clear skies…

Reply to  micro6500
January 10, 2016 7:21 pm

Just thinking the high altitude sensor can’t really be seeing the ice itself as it is hard to imagine 15 C ice. Just another mystery. A melted skin?

David Dibbell
January 8, 2016 11:17 am

Nice work! Water vapor is the superhero of climate modulation. As such it has a CAPE. “Convective Available Potential Energy” in meteorology. This old-school mechanical engineer appreciates the immense refrigeration system with which the atmosphere rejects just the right amount of heat.

January 8, 2016 11:19 am

You get additional radiative cooling in the tops of those thunderclouds that return cold rain and hail to the surface.

Gloateus Maximus
January 8, 2016 11:21 am

How can ocean temperature not be a function of heating by the sun, and possibly of undersea volcanism?

richard verney
Reply to  Gloateus Maximus
January 8, 2016 3:24 pm

Some people think that DWLWIR plays a role, but they do not explain what processes are at work which effectively sequester the DWLWIR energy that is absorbed in just a few microns to depth, thereby dissipating and diluting that energy by volume, at a rate fast enough to prevent the oceans being boiled off from the top down.
Some people consider that ocean overturning (which does not operate 24/7 since it is a diurnal event) and other slow mechanical processes such as wind, waves and swell mix the absorbed LWIR to depth but since these are slow mechanical processes, and do not operate effectively when weather conditions are calm say BF2-3 and less, these people never explain how these processes can overcome the consequences of DWLWIR being fully absorbed in a wafer thin volume of water (fractions of a human hair width) before that energy if absorbed would drive evaporation.
Fortunately, for us, solar is absorbed at depth and over a large volume and hence the energy is dissipated and diluted by volume. IF the absorption of solar was the same as that of LWIR, the oceans would have boiled off long ago and we would not be living on this water world of ours.

Gloateus Maximus
Reply to  richard verney
January 8, 2016 3:30 pm

Naturally, quite so. But relatively small changes in solar activity can produce climatic effects over decades, centuries and millennia. On the time scale of tens and hundreds of thousands of years (and probably shorter scales as well), insolation variation based upon Milankovitch cycles produce larger effects.
Most of the heat that is discharged during El Ninos comes ultimately from the sun. Thus El Ninos should and do occur with greater frequency and amplitude during periods of higher solar activity.

Reply to  richard verney
January 8, 2016 4:23 pm

El Nino conditions and episodes increase during solar minima!

Reply to  richard verney
January 8, 2016 4:30 pm

I’ve heard this ‘IR cannot heat the oceans’ meme a number of times. But never from a researcher.
The ocean is constantly losing heat through evaporation. Add heat from IR and yes, it will disappear as evaporation – but an equivalent amount of heat that was already in the ocean will, as a result, NOT be lost. So it does cause heat to build up. Conceivably the amount of heat accumulation in the ocean will be less than the increase in IR, due to increased evaporation, and conceivably for this reason solar could be more (causing a greater proportion of ocean heating vs evaporation as compared to IR). I’m out of my depth now. But IR does heat the ocean, you can be sure.
The idea that IR cannot heat the oceans also fails the sanity check. If not increased downwelling IR, what has caused the buildup in ocean heat content? I’ve heard about decreasing cloud cover but that applied to the 1980-2000 period or so, not the last 15 years (when OHC buildup has continued).

oebele bruinsma
January 8, 2016 11:23 am

Thanks for the highly interesting article Willis.
Always ” thought” that forcings were an administrative term.

January 8, 2016 11:31 am

Willis you say: “The timing and the amount of thunderstorms are NOT a function of greenhouse gas forcing, or of solar forcing, or of volcanic forcing, or of any other kind of forcing. As Figure 2 shows, they are a function of temperature. ” (my emphasis added)

Are you implying that temperature is not a function of solar forcing?

Gloateus Maximus
Reply to  Michael Palmer
January 8, 2016 11:40 am

IMO submarine volcanism also is a factor in ocean temperature, even of the surface, as is of course average depth. For instance, in the hottest part of Cretaceous, the accelerating breakup of Pangaea caused thermal expansion of the oceans, driving them up onto the continents, as in the shallow epicontinental seas of Europe and North America so comfy to giant marine reptiles. Tropical ocean temperature then was 37 degrees C (98.6 °F) or more, even with the sun producing almost one percent less power.

Gloateus Maximus
Reply to  Willis Eschenbach
January 8, 2016 11:56 am

While the Early Cambrian might have been cool as a hangover from the Snowball Earth, its average temperature has been estimate much warmer than now, ie c. 21 °C
(7 °C above modern level).
IMO your analogy overlooks the possibility of turning up the thermostat, as happens when the sun is more active both in its irradiance and magnetic flux. The sun has been more active coming out of the LIA than it was during it. SST also reacts with a lag to accumulated solar heating on annual to decadal time scales, which affects the changes in wind force and direction which drive the ENSO, for instance.

Steve Fraser
Reply to  Willis Eschenbach
January 8, 2016 12:27 pm

Willis…I’d be interested to hear your thoughts on the contribution of surface wind to the rate of evaporation.

Reply to  Michael Palmer
January 8, 2016 5:57 pm

Palmer: I understood him to be saying that *given* temperature, solar forcing as such doesn’t have any predictive value for timing & amount of thunderstorms. I certainly didn’t see anything there saying that solar forcing had no effect on temperature.

richard verney
January 8, 2016 11:36 am

My hypothesis says that when the tropical surface temperature goes over a certain threshold, that thunderstorms emerge to put a firm cap on the temperature by cooling the surface.

When I was at school, this is what I was taught about the pattern of weather in the tropics. This was one of the characteristics. Just like in some areas monsoons being another. No one at that stage was addressing global warming or climate change, and therefore no one suggested that this process/pattern was the or even a control knob which kept centennial temperatures at a level of say +/- 0.3 degC, but it was well known and accepted that was simply the pattern of the day; the temperature increased, clouds formed, thunderstorms emerged, and the day cooled.
A similar course of events usually happens to end warm spells in the UK.. There may be a few goods days of weather, the week heats up, and then the good weather is brought to a sudden end with a thunderstorm , Usually, in time for the weekend so that the planned for BBQ is ruined. Following the thunderstorm, the air is fresh and cool, and temperatures return to ‘normal.’

Brian j in UK
January 8, 2016 11:42 am

I lived in Zambia for five years and the year was divided into two, re weather – wet season 6 months and 6 months totally dry no rain. The wet was a major thunderstorm every day about 2:30pm lasting about 1 hour, very heavy torrential rain. Within minutes of the rain ceasing the roads would be drying up, steam rising from the tarmac. Within half an hour the whole place was dry. I read somewhere at the time that 70% of the rainfall was immediately transpired back into the atmosphere. This was at least 1000 miles from any ocean. As it did so the temperature noticeably cooled – so it was ok for our local cricket matches starting after work at 4pm!
Just an observation.

john harmsworth
Reply to  Brian j in UK
January 8, 2016 3:16 pm

Perhaps cricket causes cooling? Gives me the chills just thinking about watching it : )

January 8, 2016 11:47 am

See this happen all the time on the bottom edge of the tropics where I live – hot day often ends with a thunderstorm and cooler afternoon.

brian j in uk
Reply to  Willis Eschenbach
January 8, 2016 12:36 pm

Willis, I read your contributions with v great interest. I also lived in Sydney for many years and same thing there. V hot day or days then a line of thunderstorms would come up – known locally as a “Southerly Buster” with torrential rain and enormous lightning activity – sometimes lasting many hours. Result – temperature would drop as much as 15 degF in as many minutes. Great relief all round!! We thanked all that was holy for those Southerly Busters. I know this is only anecdotal, but thunderstorms cool locally wherever they occur. So if they occur over a very wide area as you maintain then massive cooling will result.

john harmsworth
Reply to  Willis Eschenbach
January 8, 2016 3:23 pm

Just as a side note on this. It is always apparent here in Western Canada in the winter time that snowy days are typically warmer. Below approximately -30C it very rarely snows and is usually bright and sunny. The phase change of water vapour to snow ( ice) is 970 btus per lb water vapour to liquid sate+ 144 btus per lb liquid to ice with the heat given off helping to keep surface temps slightly higher. Unfortunately, some of that heat ( most? ) escapes to space. Seems to me we have pretty good numbers worldwide on precipitation in different forms so we should be able to calculate total heat transfer.

george e smith
Reply to  Willis Eschenbach
January 8, 2016 4:55 pm

When I lived in St Louis county Missouri, in the Summer time, every Sunday afternoon at 3-4PM there was a thunderstorm, and me and my buddy would go out in a corn filed and set up our cameras to record the lightning strikes.
We gave up the practice, when on one of those occasions, a big tornado went through the cornfield, and tore it up something fierce.
The nearest ocean to St Louis County, is Lake of the Ozarks.

January 8, 2016 11:52 am

A watt is a power unit, 3.412 Btu per English hour or 3.60 kJ per metric hour. The sensible heat of dry air is 0.24 Btu/lb-F, of liquid water 1 Btu/lb-F. Water evaporates/condenses latent heat at constant temperature at about 1,000 Btu/lb. The heat in moist air (see psychrometric charts, programs) is a combination of dry air sensible heat and water vapor heat, dry at the sensible heat value and the water vapor holding over 1,000 Btu/lb.
The water vapor cycle is the thermal gorilla that runs the climate, CO2/GHG RF is a flea on that gorilla’s butt.

gary turner
Reply to  Nicholas Schroeder
January 8, 2016 1:41 pm

More likely it’s a pimple on the butt of said flea.

Reply to  Nicholas Schroeder
January 8, 2016 2:26 pm

Willis said 0.8 Wa/m2 equals the heat required to evaporate a 1mm thick layer of water. So 2xCO2 means about 5mm increase in yearly rainfall / evaporation would eat the heat flow. But. How much the temperature rises to do that? A little? Or more? And the atmospheric water vapour, I thought it is not increasing atm?

January 8, 2016 11:54 am

P.S. It’s relative humidity that drives evap/cond more so than temp.

Tom in Florida
Reply to  Nicholas Schroeder
January 8, 2016 1:00 pm

That’s probably why you do not see thunderstorms develop over desert regions.

Gloateus Maximus
Reply to  Tom in Florida
January 8, 2016 3:21 pm

Desert thunderstorms do occur but with less frequency, as during the Arizona monsoon season. They are a major cause of hazardous flash floods.

Reply to  Tom in Florida
January 8, 2016 4:41 pm

Lived in Reno, NV for some 8 years. We had thunder storms in the summer without seeing a cloud. They would start brush/forest fires.

Reply to  Nicholas Schroeder
January 8, 2016 2:27 pm

Or wind multiplied with RH?

john harmsworth
Reply to  Nicholas Schroeder
January 8, 2016 3:26 pm

Relative humidity is a function of both temperature and absolute humidity so can’t exactly discount temperature.

Reply to  Nicholas Schroeder
January 8, 2016 3:55 pm

Rel Humidity also plays a role in night time cooling, when it increases as it cools, once it in the 80’s and 90’s, the hourly rate of cooling slows. I’ve not sorted out if it’s just the latent heat of condensation contributing to a constant rate of energy loss, or whether the larger water molecules slows the rate of energy loss in the IR bands.

January 8, 2016 11:56 am

Very nice analysis. There is a compounding knockon effect beyond SST thermoregulation, Lindzen’s adaptive infrared iris (BAMS 2001) which he showed regulates regionally over multiple days via tropical cirrus impacts. More/strongerTstorms, more rainout, less moisture detrainment from the thunderhead ‘anvil’, so less cirrus (dryer upper troposphere), so more cooling. Cirrus are are ice, so ‘warm’ since opaque to outbound infrared LWR while transparent to incoming SWR. Bjorn Smith’s paper last year showed that by adding adaptive iris alone, model sensitivity was moved halfway to observational. Judith and I did back to back posts on it over at CE. That Without including the additional Eschenbach heat flux effect.
Dai’s 2006 J. Climate paper showed precipitation is undermodeled. Wentz’s 2007 Science paper showed by about half in the tropics. Strong literature support for this solid looking theory highlighting fundamental climate model flaws.

Quinn the Eskimo
Reply to  ristvan
January 8, 2016 12:27 pm


john harmsworth
Reply to  Quinn the Eskimo
January 8, 2016 3:28 pm

With a chance of rain!

JJM Gommers
January 8, 2016 12:02 pm

This mechanism controls nicely the temperature in the tropical region, but there is heat dissipating to the higher latitudes with an amplifying effect on the temperature. The ultimate heat sinkholes are the poles. Incremental heat effects like the anthropogenic will become apparent at the polar regions. There is no measurable change at Antarctic and minor change at the Arctic because the majority of the anthropogenic contribution is in the Northern Hemisphere. Is the conclusion justified that the anthropogenic contribution to the climate is far less than the proponents of AGW are telling.

john harmsworth
Reply to  JJM Gommers
January 8, 2016 3:33 pm

Most of the heat dissipated to the poles would be “attached” to water vapour so an important effect would be increased precipitation which would reduce the realized temperature effect. Perhaps this is why the actual observed temp rise is vastly lower than the models predict and why Antarctica is gaining billions of tons of snow mass. Arctic snowfall mainly occurs over water I would expect so quantity is probably largely unknown?

Wim Röst
January 8, 2016 12:02 pm

“Buckle your seatbelts and keep your hands inside the vehicle, it’s gonna be a long, uphill struggle to get rid of this madness …”
Posts like this will help to make ‘the uphill struggle’ much shorter!

Thomas Edwardson
Reply to  Wim Röst
January 8, 2016 1:26 pm

Are we there yet?
Are we there yet?
Are we …

January 8, 2016 12:12 pm

So how do lapse rates fit into the picture or can they be considered to be Constant?

Reply to  Alf
January 9, 2016 1:54 am

You need a T-phi diagram to find out.
On this one, the atmosphere is stable, until the surface temperature hist 35ºc. Then the dry adiabatic line (DALR green sloping left) hits the CMR line from the dewpoint temp (purple line sloping right). At that point, the air condensates, and follows the saturated adiabatic lapse rate (SALR green dotted that bows right and left).
The basic premise is that the air will stop rising when either the DALR or SALR hit the bold temperature line (red). But as soon as cloud formation starts, the air gets a warming boos from the latent heat of condensation. This is why the SALR line is almost vertical, and suddenly the rising air pocket is not goind to hit the stabilisation red temperature line until 150 mb (44,000 ft). Presto, you have a thunderstorm.
So yes, the lapse rate is key to thunderstorm formation.

Reply to  ralfellis
January 9, 2016 2:07 am

Here, I have filled it in.
My purple pocket of air rises up the DALR line and hits the red temperature line, and comes to a halt at a low level. ie: there is no deep convection and no thunderstorm.
But when the surface temperature reaches 37ºc, my blue pocket of rising air hits the CMR (my orange) and condensates, before hitting the red line. So it then follows the SALR all the way to 44,000 ft. So a very little extra surface warming, can cause a massive change in cloud top altitude.

Reply to  ralfellis
January 9, 2016 3:14 pm

Good posts Ralph …
However the Cb will not top-out at the where it crosses the ELR – it has momentum. You need to calculate the CAPE (in my experience your T-Phi exhibits massive CAPE and will (given correct wind profile) a supercell/tornados.
A T-Phi/SkewlogT has energy equaling equal areas, and so for the area enclosed by your SALR and ELR from the Normand’s point, the cloud could well rise an equal area above the 150mb level into the Strat.
Follow the SALR on up until a line drawn horizontally from it back the the right until reaching the ELR again and you have your cloud (Cb or thunderhead top).

January 8, 2016 12:28 pm

It’s interesting to watch anvil clouds form and disappear over Lake Erie during the summer months. It’s the big anvil clouds you have to be wary of. Small anvils no problem.

Reply to  Barbara
January 8, 2016 12:52 pm

tell that to Wilie E

Reply to  ferd berple
January 8, 2016 4:54 pm

By no problem. I mean small anvil clouds are not so dangerous in regards to thunder storms.

Reply to  Barbara
January 8, 2016 4:48 pm

‘Weird Clouds Look Even Better From Space’
Has photo of anvil clouds from space. Location – West Africa.
Photos of Great Lakes anvil clouds are on the internet.

Reply to  Barbara
January 8, 2016 6:56 pm

Interesting collection of satellite photos of clouds in your link above. Thank you.

Smokey (can't do much about wildfires)
Reply to  Barbara
January 9, 2016 12:43 am

Personally, I don’t care to be standing under clouds made of ANY size anvil. I’ll leave that to Wile E. Coyote, Super-Genius, and welcome to it!

David Dibbell
January 8, 2016 12:55 pm

In support of the points made above by other commenters, here is a quote from a NASA web article on its EarthObservatory website (Lindsey, 2009) “At an altitude of roughly 5-6 kilometers, the concentration of greenhouse gases in the overlying atmosphere is so small that heat can radiate freely to space.” 5-6 km is about 18,000 feet. Thunderstorms top out typically much higher than that, reaching over 50,000 feet in some cases. Every thunderstorm launches a massive burst of upward/outward longwave radiation unhindered by the overlying atmosphere!

Wim Röst
Reply to  David Dibbell
January 8, 2016 1:07 pm

I think a very important quote! Thanks!

Wim Röst
Reply to  Wim Röst
January 8, 2016 6:48 pm

Ristvan, thanks. “There are two altitudes for the two GHG, and they vary wirh latitude”. Can you give us some examples of their altitude in the tropics and on higher latitudes?

Reply to  David Dibbell
January 8, 2016 3:41 pm

Actually,this altitude can be calculated more prescisely from CO2 concentration, surface specific humidity plus lapse rate, and local surface barometric pressure. There are two altitudes for the two GHG, and they vary wirh latitude. Essay Sensitive Uncertainty. Models don’t do any of this. But if they did, they would show that the CMIP5 models were falsified by this excellent cited observation.
Simplified, models have to parameterize convective processes like Tstorms because their computationally constrained minimum grid scales are several orders of magnitude too big tomeven attempt a bad mthematical simulation. Guest post here last year. And, per IPCC, the parameterization left out natural variation. So they must–and do– run hot by a lot.

Wim Röst
Reply to  ristvan
January 8, 2016 6:49 pm

Sorry, took the wrong ‘reply’:
Ristvan, thanks. “There are two altitudes for the two GHG, and they vary wirh latitude”. Can you give us some examples of their altitude in the tropics and on higher latitudes?

Alan Watt, Climate Denialist Level 7
January 8, 2016 1:01 pm

Several interesting (to me anyway) things. First it looks like tropical latitudes take longer to respond to heat than extra-tropical ones, if I’m interpreting your scatterplot correctly. It looks like NH and SH regions produce significant evaporative cooling at 15-20°C, while the tropical areas hardly register at 20°C and really take off around 25°C. I can’t think of why that might be. Why wouldn’t the onset of evaporative cooling occur at approximately the same temperature in all regions?
Second, echoing Steve Mcintyre’s comment, I would expect there to be an El Niño signal in the TRMM data. Does it go back far enough to include the 1998 El Niño?

January 8, 2016 1:02 pm

You have explained—with elegant charts and prose—that t-storms have a governing (thermostat) effect on local temperatures but it is not clear to me what this has to do with the question of climate “forcings.”
It is obvious and established fact that thunderstorms form when hot moist air is lifted. If conditions are right, the lifted air is cooled by expansion, so the lifted moisture condenses to form cloud particles, and the resulting latent heat of condensation causes further lifting. The fact that the local environment is cooler just before, during, and after a thunderstorm forms has been known since well before man had evolved sufficiently to invent the word “cold.”
In the satellite temperature record of the lower troposphere one can clearly see the effect of volcanic forcing from the eruptions of El Chichon and Pinitubo on global temperature. That these eruptions were the cause of global cooling is further backed up by measurements of atmospheric optical depth.
However, tropical thunderstorms (presumably a dearth of them) were not able to overcome cooling caused by those volcanic eruptions. The global temperature did not recover until the had dust settled. It is not unreasonable to conclude the t-storm thermostat would also not be able to overcome any CO2 forcing on global temperature.
On the other hand, I know of no scientifically credible evidence to support the idea of runaway AGM. We have already conducted a 100 year experiment and we find that the current global temperature is probably no warmer than it was in the recent Holocene past.
Humankind has prospered because of our use of fossil fuels. Yet, all the global warming claimed by alarmists would, only just, be detectible by the average human being if it occurred instantly. No human could detect such a small temperature rise occurring slowly over 135 years (for obvious reasons).
Media hype about increases in dangerous weather, droughts and floods is not substantiated in any scientific record that I have ever seen, and I’ve seen quite a few.
Future global warming is likely to be similar to past warming, mild and mostly beneficial for Earth’s biosphere, which we are an integral part of.

Reply to  Willis Eschenbach
January 8, 2016 11:11 pm

If we compare Global Stratospheric Aerosol Optical Thickness (link below) with UAH Lower Tropospheric Temperature, we can clearly see the effect of the eruptions of Pinatubo and El Chichon. First on optical thickness, then on global temperature.
El Chichon in 1982 caused a notable increase in global optical thickness in 1983. The global temperature bottomed out in 1984 and didn’t recover until the dust had settled back to background levels in 1987.
With Pinatubo, global temperature fell immediately (1992) and didn’t recover for four years.
In your “Spot the Volcano” post you mention Mt. Redoubt in Alaska (2009) but that eruption had no notable affect on global optical depth, partly because it occurred in the far north, and partly because there was little ash and sulphur dioxide ejected into the upper atmosphere.
If we look at eruptions that eject large amounts of ash and sulphur dioxide into the upper atmosphere, sufficient to cause a significant increase in optical thickness, they have a strong global cooling effect and the effect lasts until the dust settles. This is evidence that the thermostat effect of thunderstorms, while real, is not sufficient to overcome the cooling effect of aerosol forcing.
For El Chichon and Pinatubo, the cooling effect is small, 0.25°C and 0.30°C respectively, but it’s clearly real. They occurred early in the UAH record and there have been no significant eruptions since, which makes me wonder of if the lack of recent volcanos might be the major cause of all the warming trend in that record. Was global temperature suppressed by volcanos in the 1980’s and early 1990’s so that what appears to be a warming trend is just a recovery back to normal? In other words, did a lack of volcanos cause global warming?
I will email my chart to Anthony and ask him to forward it to you. I would appreciate further discussion with you if you are in the mood.
One subject that currently fascinates me is that global warming as evidenced in the UAH record isn’t global at all. Almost all the warming is in the norther extra-tropics and polar regions. There seems to be a close correlation with the AMO and hurricane tracks, which can carry heat from the tropics up the gulf stream “super highway” and into the Arctic Ocean. The Atlantic has a large water connection to the Arctic Ocean while the Pacific has only a small connection at the Bering Strait.
GISS Global Stratospheric Aerosol Optical Thickness site:
GISS Global Stratospheric Aerosol Optical Thickness DATA

Man Bearpig
January 8, 2016 1:28 pm

Does anyone know how much energy in MWatts is released in a single lightning strike? and;
How many tonnes of CO2 would it take to make the same amount of energy ?

Reply to  Man Bearpig
January 8, 2016 5:00 pm

MW is power, not energy. Btu is energy. MW is 3.4E6 Btu/h.

Crispin in Waterloo
Reply to  Man Bearpig
January 10, 2016 12:03 pm

Your first question is answerable, and the number is very large. It is not a useful number but there is such a number. Suppose the voltage was 100m and the current (typical) is 20,000 amps. The duration is not relevant because asking for MW automatically introduces time which is independent. The answer is about 2 million MW but the time interval is short. Multiple to time in seconds (as a decimal) by the 2 million MW and you get MJ.
The second question is not answerable. CO2 doesn’t make MW.
More important is the fact that the energy in the lightning bolt mostly comes from friction in the atmosphere so it is not a net heating factor. It just concentrates energy from a diffuse source to a concentrated point. If we could tap lightning, it would cool the atmosphere, eventually, unless it was all turned into heat, not radio waves, deformed materials or high energy chemicals.

Stephen Richards
January 8, 2016 1:38 pm

Willis did you see Steve Mc’s question.

January 8, 2016 2:09 pm

Fascinating. I wish I knew what makes the depths of the ocean so cold. There seems to be even more refrigeration going on there. Do you suppose the mid-atlantic ridge is a ridge due to heat?

Reply to  HankHenry
January 8, 2016 4:14 pm

It is actually pretty simple conceptually, even if the details are horribly complicated. It starts mainly in the Arctic. Google thermohaline circulation.
In a nutshell:
The annual Arctic winter ice buildup cause of subfreezing atmospheric temps means the new sea ice exudes brine. (Ice is freshwater). Resulting briny cold (near freezing) surface seas are denser thanks to the extra salt, so sink to the ocean bottom. Now that ‘thermal engine’ flow displaces other deep water around the globe, based on ocean bottom ‘shapes’. Until it reemerges hundreds of years later as nutrient rich upwellings like off the western Americas.
Thus is the briny deep always near the freezing point of seawater. Actually, about 0C- 2C compared to seawater freezing point about -2C. Ignore depth pressure considerations in this rough sketch explanation.

Reply to  ristvan
January 8, 2016 4:56 pm

It seems peculiar and surprising that the energy that drives this thing is found at a polar extreme. Is there some kind of latent heat of solidification or liquification at play? It seems akin to that thing that makes ice cream makers work. How would one go about estimating the kind of energy it takes to maintain the extreme cold of the abyss of the ocean?

Reply to  ristvan
January 8, 2016 5:30 pm

HH, I had an interesting day with Lindzen himself at MIT, since I asked him to critique the climate chapter of The Arts of Truth pre publication, at my expense. He asked me your very question (among many others–resulting in the lengthly Svalbard footnote in the continental drift example).
There is poleward heat transport both in the atmosphere and the oceans. Skip all the specific known mechanisms. The simplest answer is heat flows from hot to cold.
So the equators heat (subject to the Lindzen/ Eschenbach regulation theories), and what happens at the poles says what eventually happens to that heat.

January 8, 2016 2:24 pm

An anecdote:
Going north in the Malacca Strait at around 20:00 local time I ran into a real humdinger of a thunderstorm, loaded with a full cargo of petrol (gasoline) and rain so hard that the radar was totally blank. Before running into the rain I had several very large vessels steaming in the same direction in the vicinity and good reason to believe that there were similar vessels coming the other way.
The quartermaster on the wheel had a spinal deformation and could not stand up straight.
So we steamed along in almost zero visibility with the E/R on standby and hearts in mouths.
Then the wind picked up from ahead, and at the same time the E/R phoned to say they had a problem which was causing sparks to come out of the funnel The fumes of gasoline from the masthead vents from the cargo tanks were therefor going straight back into the sparks.
What to do?
Well I turned round to go the other way so that the wind was aft.
And just after I turned round a lightning strike hit the water about 100 yards away. I swear that the quartermaster stood up straight as a ramrod. For the next few minutes we had St. Elmo’s Fire on the mast rigging. Then the rain stopped, all was well and we proceeded on our way.
But during the actual thunderstorm the temperature dropped noticably, and rose again once we were clear.
I think I aged 10 years in that half hour.

January 8, 2016 2:32 pm

Is this controversial? Are there “climate scientists” disputing this theory … a theory that Willis has been writing about, with data, for years? Are they ignoring it? It’s not like this blog is unknown and unread by “climate scientists.”

Gloateus Maximus
Reply to  rovingbroker
January 8, 2016 2:47 pm

It’s not a theory. At most it’s an hypothesis, but really more of an observation. It’s not in the least bit original. Here is a skeptical climate scientist’s take on the phenomenon of thunderclouds and evaporative cooling and of Willis’ views about the same:
Clouds and evaporative cooling are indeed however an important part of climatology ignored by the models of so-called “climate scientists”, ie GIGO computer gamers.
The extraordinary heat of the mid-Cretaceous I mentioned above has been attributed to lower biological productivity from the very hot seawater of that epoch, hence fewer cloud condensation nuclei.

Reply to  Gloateus Maximus
January 9, 2016 3:27 pm

“Clouds and evaporative cooling are indeed however an important part of climatology ignored by the models of so-called “climate scientists”, ie GIGO computer gamers.”
“Moist convection releases latent heat and is important to the Earth’s energy budget. Convection occurs on too small a scale to be resolved by climate models, and hence it must be handled via parameters. This has been done since the 1950s. Akio Arakawa did much of the early work, and variants of his scheme are still used, although a variety of different schemes are now in use.[20][21][22] Clouds are also typically handled with a parameter, for a similar lack of scale. Limited understanding of clouds has limited the success of this strategy, but not due to some inherent shortcoming of the method.”
“Moist convection causes the release of latent heat and is important to the Earth’s energy budget. Convection occurs on too small a scale to be resolved by climate models, and hence must be parameterized.”
To list but 2 references.

Gloateus Maximus
Reply to  Willis Eschenbach
January 8, 2016 4:09 pm

Apparently you don’t know the difference between a theory and an hypothesis. Some theories are those of universal gravitation, atomic matter, germ-caused diseases, evolution, relativity and quantum mechanics. Do you seriously expect anyone to place your unfounded conjecture in the same category as these well-supported, general scientific theories (actually laws in some cases)?
Sorry, but I’m going with Roy on this one, who has produced far more than enough evidence in the link provided.
What exactly do you imagine to be original in your hypothesis, if your assertion may be so dignified? The roles of evaporative cooling and moist convection have been well understood in meteorology for decades, at least. Is it your unsupported claim that tropical thunderstorms in particular limit global temperature, to the exclusion of other factors? Here’s a discussion of moist convection from over ten years ago:
If there’s any nonsense here, it’s all yours.
I’m also still interested in your response to my comments on the role of the sun and volcanoes in SSTs, which have been much higher in the past than permitted by your “theoretical” limit. I don’t see how anyone can argue that solar activity currently plays no role in ENSO and climate. The sun heats the ocean. When it shines more brightly for periods than in other periods, El Ninos are more frequent and stronger. Also, when its magnetic field is stronger, fewer clouds form, leading to more surface heating.

Gloateus Maximus
Reply to  Willis Eschenbach
January 10, 2016 11:07 am

If there’s any lip-flapping and hand-waving going on here, it’s by you. “Continental drift” via seafloor spreading of tectonic plates is a theory. Your unoriginal observation of the well-known effects of tropical thunderstorms is not even an hypothesis, let alone a theory.
Clearly, you didn’t do enough of a literature search before claiming to have made a discovery and developed a unique “theory”, which your unoriginal observation isn’t. No surprise that you can’t state what you imagine to be original about your observation.
That earth’s temperature is self-regulating is about as far from an original observation as possible. That tropical clouds (cirrus in this case rather than cumulo-nimbus) play a role in this homeostatic process is also no surprise. Lindzen, et al, wrote a paper on the topic in 2001 in the AMS Bulletin:
Commenters here discuss the phenomenon which you imagine you have discovered. The observation is a commonplace, so no wonder Roy took exception to your false claim of originality:

Crispin in Waterloo
Reply to  Willis Eschenbach
January 10, 2016 12:09 pm

You are digging over a field already ploughed. Willis already presented here months ago an excellent defence as to the novel nature of his theory. It may be worth your going to read that so he doesn’t have to repost everything here. You are guessing and hoping. The readership here is waiting for you to catch up. Please don’t waste more of our time.
It is novel, it is published, it is a theory now. It is also correct, IMV.

Ian W
January 8, 2016 2:34 pm

Willis a nice article.
You may be interested in this NOAA FAQ site. This part looks at energy released by hurricanes per day. A hurricane is after all a bunch of thunderstorms working together and hurricanes only occur when the SSTs are sufficiently high (85F or more) so they fit very well with your hypothesis.
This FAQ states interalia:
It turns out that the vast majority of the heat released in the condensation process is used to cause rising motions in the thunderstorms and only a small portion drives the storm’s horizontal winds.
Method 1) – Total energy released through cloud/rain formation:
An average hurricane produces 1.5 cm/day (0.6 inches/day) of rain inside a circle of radius 665 km (360 n.mi) (Gray 1981). (More rain falls in the inner portion of hurricane around the eyewall, less in the outer rainbands.) Converting this to a volume of rain gives 2.1 x 1016 cm3/day. A cubic cm of rain weighs 1 gm. Using the latent heat of condensation, this amount of rain produced gives
5.2 x 1019 Joules/day or
6.0 x 1014 Watts.
This is equivalent to 200 times the world-wide electrical generating capacity – an incredible amount of energy produced!”
Note: that is 200 times the world wide electrical generating capacity per hurricane day
I often am amazed by the hubris of humanity thinking that they are all powerful. We are dwarfed by nature. Ants and Termites produce far more CO2 than humans. Not that CO2 matters.

Gloateus Maximus
Reply to  Ian W
January 8, 2016 2:59 pm

Interesting, to me anyway, is that, while thunderstorns can and do form on the equator, hurricanes do not, due to lack of so-called Coriolis forces. No tropical cyclones have been observed to form within five degrees latitude of the equator, but once formed they can move into this zone. I might be wrong, but last I read, none had been seen to cross the equator however, at least not in the Atlantic or Pacific. Not sure about the Indian Ocean.

January 8, 2016 2:53 pm

But what about all that most powerful of GHGs, water vapor, being released into the atmosphere at low altitudes via evaporation following the storm?

Berényi Péter
January 8, 2016 3:03 pm

During the day, on the other hand, evaporation is quite large, hundreds of watts per square metre, because the strong tropical sunshine evaporates the water directly, plus the thunderstorms are largely a daytime phenomenon.

Rate of evaporation is not only a function of fluid temperature, but it is also proportional to surface area of the interface. As long as there is no storm, this area is pretty small (it is the sea surface). However, once a storm is started, high winds produce a prodigious amount of sea spray, which increases interface area by several orders of magnitude, so the entire thing becomes a self propelling engine. Until sea surface cools down sufficiently, of course. Anyway, there is this hysteresis, which should also be taken into account.

Ian L. McQueen
January 8, 2016 4:17 pm

Another of your excellent postings. Continuing thanks.
There is one thing not touched on here, which I have never seen anywhere else either. That is the amount of energy moved upward worldwide compared with the same for radiation. Do you have such a comparison?
Ian M

January 8, 2016 4:41 pm

Here’s something else to ponder. Take a surface of sea, say a square foot by an inch deep. The surface area of that water is 1 ft^2. Now evaporate it and condense it at 50,000 ft. What is the surface area of the water now? It is several orders of magnitude higher. Therefore you have much more surface area to radiate heat to outer space. This is the piece that the climate alarmists don’t understand.

January 8, 2016 5:14 pm

The graph reminds me of a fat hockey stick. 🙂

January 8, 2016 5:32 pm

I worked outside for like 20 years in the Chicago area, and as a bit of a weather geek, I noticed that if the cumulus started getting puffy by say 10 or 11 in the morning, there was a good chance of thunderstorms later in the day.

January 8, 2016 6:07 pm

“plus the thunderstorms are largely a daytime phenomenon. ”
Not necessarily.
“Over the tropical and southern oceans, showery precipitation tends to peak from midnight to 0400 LST. Maritime thunderstorms occur most frequently around midnight.”
Title (if link doesn’t work): Global Precipitation and Thunderstorm Frequencies. Part II: Diurnal Variations

January 8, 2016 6:54 pm

An opportune time for a reprise, an encore post.
First off a discussion of units.
A watt is a metric unit of power, energy over time, not energy per se. The metric energy unit is the joule, English energy unit is the Btu. A watt is 3.412 Btu per English hour or 3.600 kilojoule per metric hour.
In 24 hours (sun shining, what happens at night?) ToA power of 340 W/m^2 will deliver 1.43 E19 Btu to a spherical surface with a radius of 6,386 km. The CO2 RF of 2 W/m^2 will deliver 8.39 E16 Btu, 0.59% of the ToA.
At 950 Btu/lb of energy, evaporating 0.74 inches of the ocean’s surface would absorb the entire ToA, evaporating 0.0044 inches of the ocean’s surface would absorb the evil unbalancing CO2 RF.
More clouds. Big deal.
ToA spherical surface area, m^2……………5.125.E+14
W = 3.412 Btu/h……………………………………3.412.E+00
ToA, 340 W/m^2, Btu/24 h……………………1.43E+19
CO2 RF, 2 W/m^2, Btu/24 h…………………..8.39E+16
Ocean surface , m^2………………………………3.619E+14
m^2 = 10.764 ft^2………………………………….1.076E+01
Ocean surface, ft^2………………………………..3.895E+15
Water density, lb/ft^3………………………….62.4
Lb of water in 1 foot of ocean………………..2.431E+17
Evaporation, Btu/lb……………………………950.0
Amount of ocean evaporation
Feet needed to absorb ToA…………………..0.062
Inches needed to absorb ToA………………..0.74
Feet needed to absorb CO2 RF………………0.0004
Inches needed to absorb CO2 RF…………..0.0044
Don’t like my work? Think it’s wrong or irrelevant? At least I took a shot at doing it. Where’s your work?

January 8, 2016 7:02 pm

I’m glad someone finally started talking about the verticle transport of latent heat.
The point that everybody is missing is that the AGW models don’t factor this in. They only look at convection of sensible heat. This is a fatal error.
Also, ponder this: According to AGW enthusiasts, if there is more CO2 in the lower troposphere then more infrared radiation is reflected back to the surface. That means more evaporation. More important, it means less infrared being absorbed by water vapor above and therefore a cooler upper troposphere. That means more clouds. More clouds means higher albedo and less sunlight reaching the surface. That means cooling. Therefore CO2 causes global cooling.
If that isn’t enough, the increased CO2 below actually creates a reverse greenhouse effect by reflecting the infrared radiation coming down from above back out towards space. Oops. Nobody thought about the fact that the “greenhouse effect” is a two edged sword.
I sent Anthony a paper on this a few days ago. Ask him about it. Somebody needs to take the ball and run with it.

Reply to  Willis Eschenbach
January 9, 2016 11:26 am

Thanks for the thoughtful reply Willis.
You misunderstand my claim about the models. The models try to account for convection of sensible heat, but they don’t factor in the verticle movement of latent heat – from evaporation at the surface to condensation at higher altitudes. I’ve found no place in the models that accounts for gas-liquid phase transition or convection of latent heat. Unless they started doing it recently I don’t believe it’s there.
This ties in with your second critique about reflection. The conventional wisdom says that infrared-absorbtive gas in the atmosphere absorbs and re-emits infrared. Half goes up and out and half goes back to the surface. True or not, that’s the fundamental basis of the so-called greenhouse effect. I used “reflect” as a shortcut for the sake of brevity because reflection versus absorbtion-reemission is irrelevent to my argument.
Think of the gas as the “roof” of the greenhouse, If you move latent heat (and sensible heat, for that matter) from the surface to higher elevations then you have partially or completely taken the heat above the roof. The conventional wisdom still applies – the gas layer forms a barrier in the downward direction in the same manner as it does in the upward direction. Presumably half goes one way and half goes the other way. It makes no difference which way is up or which way is down. Insulation generally operates in both directions. That’s a fact.
There is a more subtle mechanism at work here that is independent of convection. There is only a certain quantity of infrared being emitted at the surface. If a molecule of CO2 absorbs a fraction of that radiation then a molecule of water vapor is not absorbing it. Ignoring all other mechanisms, more CO2 at lower levels will mean less infrared reaching the water vapor at higher levels. Think of it as lowering the roof or adding more insulation. That means more condensation and more cloud formation above.
Think about the purported cooling effect of increased cosmic radiation causing cloud formation in the upper atmosphere. We’re told it has an overall cooling effect.
Consider this in simple terms. There is a tremendous amount of latent heat being moved from below the insulation (“greenhouse gases”) to above the insulation. That heat is now deterred from returning to the surface. Additionally, the cooling above the insulation causes more cloud formation.
Let’s try it another way. CO2 is a nifty cloud maker. It traps heat over the oceans around the equator and pumps clouds out to the upper latitudes. It might (maybe) cause some surface warming at the equator but it might cause cooling everywhere else.
The claim that more CO2 will actually cool the planet is deliberately off the rails. The point is this: It’s a lot more credible than the idea that a tiny bit more CO2 amongst a lot of water vapor can cause warming. It considers important physical mechanisms that have not been considered and it is more consistent with the known facts. Whether these mechanisms merely reduce warming or cause outright cooling is beside the point.
If you let go of the presumption that CO2 must cause warming then this doesn’t seem so crazy. I say the two-phase nature of water is a much more powerful mechanism that a little CO2 (I think you do too). If a little cosmic radiation can cause cooling then why can’t a little CO2? We can’t let ourselves get caught up in their groupthink.
The AGW conjecture is dying anyway but wouldn’t it be nice to have a better counterargument? Let’s at least make them account for these mechanisms. Since you’re already going down this road you might want to take a closer look.

Werner Brozek
Reply to  CM
January 8, 2016 9:09 pm


Well there is this article at Judith Curry’s blog:
“How increasing CO2 leads to an increased negative greenhouse effect in Antarctica”
Of course Antarctica is not the globe, but you may find it interesting.

Reply to  CM
January 9, 2016 3:33 pm

“The point that everybody is missing is that the AGW models don’t factor this in. They only look at convection of sensible heat. This is a fatal error.”
Of course they do!
It’s a MASSIVE transport of heat from the surface aloft.
Fundamental to the working of the climate system.
See my response to Gloateus.
And Google it there are pages of references to it being so.

Reply to  Toneb
January 9, 2016 4:32 pm

Last time I looked, the literature wasn’t talking about vertical transport of latent heat. Fiddling with parameters is just guessing.
Regardless, convection and latent heat transport are two different things. You can’t assume that you can Supersize the convection parameter to account for latent heat. One really good reason off the top of my head is that there is massive poleward transport of latent heat. That has nothing to do with convection. Sorry.
Also, the change in latent heat transport with increased CO2 cannot be assumed to be the same as the change in convection. There may be a different level of feedback.
Therefore, I’m quite certain that they aren’t modeling latent heat. Shortcuts don’t count. You might want to look for some primary references on this if you still disagree.
Either way, the models are pathetic. I looked at them a couple years back and was shocked by the crudeness and downright vulgarity of it all. They are little toys for little boys. Those models will never, ever, ever be right because they don’t accurately model the thing they are trying to model. They need to be reconceptualized. I used to be a programmer and I’ve done a lot of systems analysis so I know that of which I speak. Those guys should be hoisted on their parameters.

January 8, 2016 8:09 pm

IPCC AR5 gave clouds a -20 W/m^2 RF.

January 9, 2016 1:24 am

Are you going to update your cloud thermostat paper, to include all of these developments?

January 9, 2016 2:27 am

OK, the twit is erased, we can rid of that now !!

January 9, 2016 3:54 am

From Held:
A warming atmosphere typically results in larger horizontal moisture transports. In addition to the implications for the hydrological cycle and oceanic salinity discussed in previous posts, this increased moisture transport also has implications for energy transport. If energy is used to evaporate water at point A and the vapor is transported to point B where it condenses, releasing the heat of condensation, energy has been transported from A to B. This latent heat transport is a large component of the total atmospheric energy transport. Outside of the tropics, eddies are mixing water vapor downgradient, resulting in a poleward transport. Close to the equator, the Hadley circulation dominates, with its equatorward flow near the surface that carries water vapor from the subtropics to the tropical rain belts (the compensating poleward flow near the tropopause carries very little water vapor in comparison).
Held speaks about the ratio between horisontal an vertical energy trensport. The vertial is very little compared to the horisontal. As I understand some of the comments thunderstorms modify this by increasing heat upwards. And that this can go to latitudes that give more radiation to space. I think it is very ineresting to see the “thermostat” as a result of both vertical and horisontal motion of air and moisture.

Reply to  nobodysknowledge
January 9, 2016 3:56 am

Correction: And that this can go to altitudes that give more radiation to space.

richard verney
January 9, 2016 4:25 am

I have a couple of questions for you arising out of various statements quoted below:

“Here is the distribution and strength of rainfall, and thus evaporation, around the middle of the planet”

We live on a water world, and understanding the water cycle is the key to understanding the planet’s climate, but I do not understand the above assumption. As I have previously mentioned to you, energy absorbed in one part of the system re-emerges in a different part of the system because of oceanic and atmospheric currents, and of course, not all energy is absorbed at the surface, and some of the energy absorbed say in the 20cm to 5 m depth of the ocean is mixed vertically downwards.
Fisrt, If you look at Fig1 (and if I understand matters correctly), are you really saying that there is no evaporation in the dark blue areas, say off the East Coast of Africa, or off the West Coast of Australia? The Figure portrays that much of the Indian Ocean (vast areas of which are particularly warm) has no evaporation!
You also state:

“It takes about 80 watt-years of energy to evaporate a cubic metre of water…”

I do not know how that approximation/rule of thumb has been calculated, but (if my maths is correct), given that the specific heat of water (which is 4.186 Joule/gram °C), that means that it requires some 293.02 Joules of energy to heat 1 gram of water by 70 °C (ie., from about 30 °C to 100 °C).
Now then due to the omnidirectional nature of DWLWIR approximately some 60 to 70% of all DWLWIR is fully absorbed in just 3 microns of the ocean. This is a volume of just 0.000003 cubic metres and is a mass of some 3 grams. According to K & T, the global average DWLWIR is some 324 W/m^2, and ignoring the fact that in the tropical region DWLWIR will be considerably higher than the global average, there is approximately some 226 W/m^2 (ie., 324 x 70%) or 226 Joule Seconds being fully absorbed in just 3 microns. Thus, it would take just 1 second to entirely evaporate the 3 micron layer (in which about 60 to 70% of all DWLWIR is fully absorbed), and this means that if DWLWIR possesses sensible energy and is capable of performing sensible work in the environ in which it finds itself, the oceans would boil off from the top down, and would no longer remain given that they have received this energy for the best part of some 4.5 billion years, unless the energy that is absorbed in the top 3 microns can be sequestered to depth and hence diluted and dissipated by volume at sufficient speed to prevent this energy from boiling off the top microns of the oceans. This leads on to my next question.
Second, the question is what physical processes operate to mix this energy and sequester it to depth at sufficient rate to prevent this boil off?
In the past, you have mentioned processes such as ocean overturning, the action of the wind, waves and swell. My response to that is that these are all slow mechanical processes, and do not operate 24/7 365 days of the year.
For example, ocean overturning is a diurnal event, and therefore can at best only operate for half the day. Further, there may be no equivalence in some large inland seas, lakes such as the Great lakes, the Dead Sea, the Sea of Azov etc.
The action of wind, waves and swell cannot operate effectively when weather conditions are in the order of BF2 when there is simply insufficient wind to break surface tension, and to cause waves etc to carry out any mixing.
When we last discussed this (before Christmas), I mentioned that I was overlooking the Mediterranean and it was as still as a millpond with not a ripple in sight. The local weather station suggested that the prevailing wind speed was 1 mph, so one can see why there was no mixing by this slow physical process. I asked you why the Med was not boiling off from the top down. Again, today, similar conditions are being encountered although it is said to be a little windier at 5 mph. This is midway in BF2, ie., a light breeze and according to the BF scale, there should be “small wavelets, crests of glassy appearance, not breaking”. Where I am the trees are not rustling at all, and one cannot feel any breeze on the skin even if licked, and I would estimate that the local conditions are more like BF1 (I have 40 years plus of sailing experience), so I would again enquire why is the Med not boiling off from the top down as we correspond?
All I want to know is what physical process can on a 356 24/7 basis sequester the DWLWIR being absorbed in the top few microns of the oceans down to depth (and hence dilute and dissipate that energy by volume) with sufficient rate so as to prevent that energy from simply boiling off the top microns?
I do hope this time to receive an answer to the questions posed.

Smokey (can't do much about wildfires)
Reply to  richard verney
January 9, 2016 6:29 am

“If you look at Fig1 (and if I understand matters correctly), are you really saying that there is no evaporation in the dark blue areas, say off the East Coast of Africa, or off the West Coast of Australia? The Figure portrays that much of the Indian Ocean (vast areas of which are particularly warm) has no evaporation!”
Not so. Those are area where the evaporation results in zero effective cooling, as expressed in W/m^2. It does NOT say there is no evaporation. Having been to Bahrain, Kuwait, etc., I can tell you this jibes with what I’ve personally experienced; nighttime fog at 30+C air temp is a freaky thing.
“Thus, it would take just 1 second to entirely evaporate the 3 micron layer (in which about 60 to 70% of all DWLWIR is fully absorbed), and this means that if DWLWIR possesses sensible energy and is capable of performing sensible work in the environ in which it finds itself, the oceans would boil off from the top down, and would no longer remain…”
If I’m understanding you correctly, you estimate that the 3 micron layer amasses about 3 grams worldwide. Given that, you then assume that the evaporation of 3g/sec from the oceans would be enough to evaporate them entirely, given geologic timescales.
This seems a reasonable conclusion until we remember that there are a number of 24/7/365 processes which add water BACK to the oceans at the same time this evaporation is taking place. If it were not so, the Mediterranean (at least) certainly would not exist, given that it seems to lose a lot more water to evaporation that it gets back via rainfall alone — as you have eloquently pointed out via personal experience & observation.
Thus, just as it makes sense to you (as it does me) that there would be much more energy available for evaporation at the tropics than at the poles, so too it makes sense to me that is there quite a bit more rainfall/precipitation in general in the tropics than at the poles. I’d be willing to bet the two phenomena nearly cancel, with other water cycle processes making up what would otherwise be a measurable oceanic deficit.

richard verney
Reply to  Smokey (can't do much about wildfires)
January 9, 2016 10:29 am

Thanks your clarification of fig !. That makes more sense.
As regards my 3 gram figure, this is 1m x 1 m x 3 microns, ie., it is the 3 micron wafer of a 1 metre by 1 metre square area of the ocean which square receives the 324 W/m^2 in the K&T energy budget cartoon. I use 3 micron wafer since about 60 to 70% of DWLWIR is fully absorbed in that wafer.
As a matter of vertical penetration 50% of LWIR is absorbed in 3 microns, but since DWLWIR is omni-directional with a large component having a grazing angle of say 25%or less, it follows that somewhere between 60 to 70% of DWLWIR is fully absorbed in a vertical slice 3 microns thick.

Crispin in Waterloo
Reply to  richard verney
January 10, 2016 12:28 pm

The energy needed to evaporate a g of water from a 30 Deg C ocean is 2501 J minus (30 degrees x 2.44 J). This subtraction accounts for the extra energy it takes to directly evaporate the water without heating it up to 100 C first (which is what happens).
So it is 2,428 J/g or 2.4 GJ per cubic metre which is 77 Watt-years.
324 Watts average DWLWIR is enough to evaporate 324/2428 g per second = 0.1334 cc or 11.53 litres per day. That is a depth of 11.53 mm per day, presumably in the tropics.
The calculation is not much help because the annual average evaporation from the entire ocean is nothing like that much. But that is the calculation you wanted.

January 9, 2016 6:31 am

Richard Verney “…DWLWIR…”
IMHO (BSME, PE, 35+ years in power gen) whether conduction, convection or radiation heat flows from hot to cold, not from a colder LT to a warmer surface. S-B would be negative in this delta T direction. This popular upwelling/downwelling GHG radiation loop is a basic perpetual motion violation of thermodynamic principles.
Plus because of Einstein’s Nobel winning photo electric effect any re-emitted energy has to be less than the incident energy, difference being the work function and in the microwave range.
Which is all academic because:
1) Mankind’s net 4 GT/y CO2 contribution to the globe’s 45,000 GT of stores and 100’s of Gt/y natural fluxes is trivial.
2) CO2’s 2 W/m^2 RF is trivial.
3) The GCMs are trash.

Hot Air
Reply to  Nicholas Schroeder
January 9, 2016 7:40 am

“IMHO (BSME, PE, 35+ years in power gen) whether conduction, convection or radiation heat flows from hot to cold, not from a colder LT to a warmer surface. S-B would be negative in this delta T direction. This popular upwelling/downwelling GHG radiation loop is a basic perpetual motion violation of thermodynamic principles.”
Nicholas, as a fellow ME I couldn’t agree more. The first time I saw that picture showing radiative heat transfer from the cooler atmosphere to the warmer planet I knew it was designed to obfuscate. Anybody that actually has to make things work in the real world, as opposed to between their ears, would NEVER attempt to calculate this for the reasons you state.
The other is it is clear they have never seen nor used a psychrometric chart, nor understand why they should.

Reply to  Hot Air
January 10, 2016 9:53 am

Oh dear.
The usual misinterpretation of the GHE that is found often on these pages.
It is the NET flow of energy that is of concern.
Of course a colder object cannot “heat” a warmer one but it DOES slow it’s cooling.
All objects emit/absorb EM energy (above 0K).
An object that receives (absorbs) a photon cannot know (or care) where it came from, and somehow magically reject it whether from a hotter or colder object.
It just absorbs it.
Think of the hotter object (Earth) as a tank of water that is leaking, say 10gals of water per hour.
Then link a smaller (colder) tank of water that drips in 1gal an hour whilst the gig one still leaks the 10 gals.
Result – The big tank is still leaking (cooling) but at a slower rate of 9 gal/hr. IE the NET leakage of water (energy) is still AWAY from the big tank (Earth) but it’s SLOWER whilst the small tank (cold atmosphere) continues to feed in that small water supply.
The 2nd L of Thermodynamics is not broken – it applies only to NET flow.

Reply to  Toneb
January 10, 2016 9:57 am

Toneb says:
It is the NET flow of energy that is of concern.
Wrong, the major concern is the fact that global warming stopped many years ago. But instead of trying to underrstand why your crowd has been flat wrong about everything, you still try to ‘explain’ what isn’t happening.

Reply to  Hot Air
January 10, 2016 12:49 pm

I was addressing the science of GHE (yes I know it’s badly named).
Go to Roy Spencer’s site for an explanation and experiment to demonstrate.
PS: No one is saying there are not natural climate cycles that can mask it’s effect for some years.
Which is what you diverted onto.

Smokey (can't do much about wildfires)
January 9, 2016 6:36 am

Reading richard verney’s comments got me to thinking along a bit of a tangent to this topic, but I’d love to hear your thoughts on it anyway:
I wonder what the net effect of continental run-off is on ocean temperatures? It seems like water running off the land would tend to be more easily warmed prior to entering the sea, thus providing a net source of warming to the oceans. This would vary, of course, as regional/planetary climate cycles alternate between warm and cold, seasonally & by epoch. But I wonder if part of the reason the ocean warms at all in large scale is due to a warming of continental run-off compared on average? The “run-off anomaly,” if you will?
Just thinking of the possible plus sign(s) which might run counter to the minus sign you got going with the ITCZ. Thanks for your work!

January 9, 2016 8:23 am

I found an atmospheric heat/power flux balance diagram among Bing images that is labeled “Fig 10 Trenberth et. al. 2011.” Some of you might recognize the paper. What is interesting is that there are eight values and an average displayed for each of the major state points. From what I can tell there are eight different studies/data bases/calcs and in some cases with quite different values. What happened to consensus? A couple of the variation ranges/uncertainties are an order of magnitude greater than anthropogenic CO2’s 2 W/m^2 RF. And there is the 333 W/m^2 perpetual (GHG?) power flux loop between earth’s surface and sky, i.e. lower troposphere.
A summary table:
………………………..ave W/m^2…..+/- %…+/- W/m^2
ToA net…………………342.1……….1.5……..…..5.13
Albedo, %…………………29.8……..…..3.5……..…..1.04
Surface Radiation……..398.0……….2.5…….……9.95
Back Radiation……..….338.3……….8.5…………28.76
Evapo transpiration…….85.1…….….9.0……….…7.66
Reflected Surface………27.4…………5.7…………1.56
Reflected Solar………..101.4…….….11.5….…….11.66
Latent Heat…….……….88.1………….8.0…..….…7.05
Per IPCC AR5 the cumulative CO2 RF between 1750 and 2011 is 2 W/m^2. OLR uncertainty is +/- 22 !!!!!!! Reflected solar +/- 12!!! Even RCP 8.5 gets lost in uncertainties this large. How can anybody claim significant confidence in the present or future global temperatures with such huge uncertainties?

January 9, 2016 9:32 am

Thanks, Willis, for a very good article.
Yes, I can see thunderstorms moving heat up to be radiated out, and how this conforms a feed-back control of local weather in the tropical and sub-tropical regions.
Your “scatterplot of sea surface temperature versus thunderstorm evaporative cooling” (Figure 2) is very convincing evidence.
Happy New Year!

glen martin
January 9, 2016 10:38 am

Willis, if the thunderstorms are shifting heat upwards:
“For thunderstorms, the working fluid is water. When it evaporates at the surface, it cools the local area, and the heat is moved from the surface to the clouds and on upwards.”
Shouldn’t this show up in the satellite data as warming in the upper troposphere like the global warming fingerprint that hasn’t been found?

Craig Loehle
January 9, 2016 11:25 am

This I believe is the big flaw in the climate models which use a constant lapse rate to view water vapor and ignore heat transfer by thunderstorms. The model grid cells are much too large for thunderstorms anyway. If thunderstorms were not dissipating heat, they would not exist. They transfer massive amounts of heat to the upper atmosphere where it is more easily lost to space. It is like boiling a pot of water. As long as bubbles can form and rise, the water stays at the boiling point. Only in a pressure cooker can the water get hotter than this.
I think this is also why the ice ages were very dry–not much ocean warm enough to create convective storms.

Reply to  Craig Loehle
January 9, 2016 12:18 pm

Yes. The models consider convection in some crude fashion but I don’t believe they consider the verticle movement of latent heat in water vapor. I raised this in my post but Willis didn’t give a clear response. I’m hoping he will clarify.
Latent heat movement is many times higher than sensible heat movement in water vapor. By the way, it’s not just true of thunderstorms. It applies to the entire hydrological cycle.
In my post I made the additional point that if CO2 has an insulating effect then it should start to deter the return of infrared back to the surface as convection carries the water vapor above the CO2 (and other greenhouse gases). Insulation works in both directions, after all. A blanket will keep your beer cold just as well as it keeps your body warm. More insulation, less heat coming in, colder beer.

Reply to  CM
January 10, 2016 10:10 am

“Yes. The models consider convection in some crude fashion but I don’t believe they consider the verticle movement of latent heat in water vapor. I raised this in my post but Willis didn’t give a clear response. I’m hoping he will clarify.”
They consider LH release via convective uplift (and also by frontal/baroclinic/orographic uplift).
LH is far greater that sensible heat transport by clouds – even small Cumulus cloud has WV condensing and warming the Trop at lower levels.
And BTW: Unless and until we can get a massive supercomputer to model each individual cloud – then all we can do IS to parametarize the process. Even in NWP (weather models) it cannot be done explicitly.
“Some of the datasets in Table 3 are estimated; for instance latent-heat fluxes and the other components of the surface energy balance are estimated, based on the extent of empirical formulas
and energy-conservation principles.”

January 9, 2016 11:48 am

In the HVAC industry a common term for cooling is the ton. It’s the amount of energy stored in a short ton, 2,000 lb of ice, 12,000 Btu or 3,517 Wh. That’s the energy needed to freeze a ton of water into ice or the energy released from melting that ton of ice and cooling the air circulating in the building.
Antarctica covers 14 E6 km^2 (14 E12 m^2) more or less. Snow and ice a meter deep would be 14 E12 m^3. A m^3 holds a tonne of water (slightly more ice), 2,204 lb, and would represent 13,224 Btu.
Precipitate a meter deep by square meter (i.e. cubic meter) of snow/ice over 24 hours, 13,224/24 h = 551.3 Btu/h or 161.6 W/m^2. That cubic meter of snow/ice precipitated over Antarctica would remove 5.43 E16 W from the atmosphere. Wow! Good thing it doesn’t snow much in Antarctica.
In 24 hours ToA of 340 W/m^2 delivers 17.4E16 W.
So it would take almost undetectable amounts of snow/ice at the ice caps and sheets to suck up mankind’s pitiful 2 W/m^2 of CO2 RF.
Subject to peer R&C.

January 9, 2016 1:36 pm

It’s time to acknowledge that the atmosphere is a three-phase system.
Willis, along with some of the commentators, has opened the door on something that goes way beyond thunderstorms. Where this must ultimately lead is the realization that all current models are 100% useless – because they don’t properly consider the transport of latent heat as water transitions through gas, liquiid and solid phases.
A thunderstorm is the biggest example of latent heat transport but every cloud is doing the same thing to some degree. You have to look at the entire hydrological cycle. Every bit of evaporation at the surface moves up in the atmosphere and eventually returns to the surface. It must – otherwise the atmosphere would load up with water.
Latent heat is captured at the point of melting/evaporation/sublimation and released at the point of freezing/condensation/desublimation. Most of it is captured at the surface and released somewhere in a cloud, but that’s not the only way.
There is also a massive amount of capture happening up in the atmosphere. Virga (raindrops that evaporate before they hit the ground) are just one example. All clouds are really continuous heat pumps that are capturing latent heat below and releasing it above. You can see this if you watch fast-motion video of cumulus clouds. The roiling of the cloud is caused by droplets falling from the top and re-evaporating somewhere down below.
Furthermore, a lot of evaporation & condensation is happening invisibly at the microscopic level. Just because you don’t see a cloud doesn’t mean it isn’t happening. When you look up and see sky that isn’t cloud but isn’t as blue as it should be, or when you see haze, you’re getting a peek at this phenomenon.
So, the entire atmosphere is a continuous, dynamic phase-transition process. That’s important because latent heat released up in the atmosphere will presumably escape to space more readily – and because latent heat is a significant portion of the heat in the atmosphere.
Good luck modeling that. Heck, good luck even measuring that.
Now here’s the real mind blower: Once you acknowledge a three-phase system you have to start distinguishing between temperature and enthalpy. Delta T does not equal delta H so you can’t just stick a thermometer out the window any more. Poof (that’s the sound of the entire climate paradigm going up in smoke).
I’m glad I’m not a climatologist. You’ve got your work cut out for you. I’m out of here.

January 9, 2016 1:43 pm

“Delta T does not equal delta H”
For liquid water it’s close enough, beyond that, good luck! e.g. 50/50 glycol & vapor,

old construction worker.
January 9, 2016 6:46 pm

Thanks Willis Good job
“At night the evaporation is small, some tens of watts per square metre. During the day, on the other hand, evaporation is quite large, hundreds of watts per square metre, because the strong tropical sunshine evaporates the water directly, plus the thunderstorms are largely a daytime phenomenon.”
To really nail down evaporation, one needs to keep record of wind speeds.

Reply to  old construction worker.
January 9, 2016 8:40 pm

old construction worker, replying to Willis E

“At night the evaporation is small, some tens of watts per square metre. During the day, on the other hand, evaporation is quite large, hundreds of watts per square metre, because the strong tropical sunshine evaporates the water directly, plus the thunderstorms are largely a daytime phenomenon.”
To really nail down evaporation, one needs to keep record of wind speeds.

Well, sort of. Even if yoiu assume the water temperature and air temperature are in equilibrium with each other – which ABSOLUTELY is NOT the case in real life, the evaporation heat losses from open, wind-swept water are only one of four “other losses” goin on simultaneously:
LW radiation down to the water (a heat energy gain)
LW radiation up from the water (to – or through! – the near surface atmosphere & relative humidity, into the high altitude air mass and then into space’s darkness)
SR radiation down into the water (from the sun, proportional to day-of-year, time-of-day (solar elevation angle), wind speed, wave height, and air cloudiness, and air clarity)
Evaporation (a heat energy loss, and a very minor mass loss as well.)
Convection (a heat energy gain or loss, depending on surface air temperature)
Conduction (replacement of heat energy into the depths, if no ice is present.)
If you know, or assign, surface water temperature and near-surface (2 meter) air temperature, then the last three become
Energy in (+LW in + SW in) = Energy LW Out + Energy Evap + Energy Convection + Energy Conduction
Three of the four Energy “losses” are proportional directly to DeltaT, or ( T Water – T Air)
Two (long wave radiation in and out) are proportional to Twater^4, Tair^4, and the relative humidity near the ground and Tair (altitude) and a few other factors (such as air clarity and cloud cover reflections.)
So you have to solve the four equations simultaneously to get an approximation for the approximation of a single hour’s equilibrium losses and gains.
In the absence of ice covering the surface, you have to keep track of (or estimate) each hour’s surface air temperature, each hour’s surface air wet bulb temperature, (and from that) each hour’s relative humidity, each hour’s actual air pressure, each hour’s wind speed, each hour’s surface water temperature, and some approximation of each hour’s deeper water temperature. That let’s you begin calculating approximate heat transfer coefficient for the other heat losses ….
It’s a “fun” problem. Not. Given data for one place at one hour, I’d like to run these “other heat losses” for even one day at some particular latitude and day-of-year.
the total “other losses” of each square

January 9, 2016 7:20 pm

You wrote:
“I’m sorry, but your uncited and unsupported claim that volcanoes “have a strong global cooling effect and the effect lasts until the dust settles” does NOT constitute “evidence that the thermostat effect of thunderstorms, while real, is not sufficient to overcome the cooling effect of aerosol forcing”. Evidence is verifiable facts and observations, Thomas, not your uncited claims.”
But I did cite the evidence to support my claim. I told you the global temperature data are from UAH and the Optical Thickness data are from the following GISS sites.
GISS Global Stratospheric Aerosol Optical Thickness site:
GISS Global Stratospheric Aerosol Optical Thickness DATA
I also emailed a chart to Anthony. (I don’t know how to post an image here.)
You wrote:
“Nope. If they had, the lack of volcanoes since 1992 would have led to continual warming, but in fact the world hasn’t warmed in almost two decades.”
I disagree. There were major volcanos that increase aerosol loading, which decreased sunlight (see the GISS data) in the 80’s and 90’s but none since Pinatubo in 1992. Those eruptions clearly caused cooling, which lasted until the aerosol loading settled or rained out. We would not expect “continuous warming” after recovery from volcanic cooling. All else being equal, we would expect temperatures to recover back to their previous levels not to continue warming
As it turned out, all things were not equal. A major El Neño caused “warming” at the end of the 1990’s. I put warming in quotation marks because an El Neño doesn’t introduce more energy into the system, it just spreads around heat that had been concentrated in the wester pacific.
Nevertheless, the dramatic increase in stratospheric aerosol loading caused by El Chichon and Pinatubo are clearly evident in the GISS aerosol data and the UAH temperature data shows concurrent cooling. The mechanism is obvious, the aerosols cause less sunlight to reach the surface.
Your idea that thunderstorms act as a governor on surface temperature seems correct to me. If the surface is hot, convection causes more surface winds, the heat and wind cause more water to evaporate, causing more or larger storms, which cool the surface. Conversely, if the surface is cool, fewer or smaller storms form so the surface gets hotter.
Additional evaporation during the cumulative stage of a thunderstorm, which is caused by both increased surface wind and heat, does lower the temperature but the total heat content of the air is constant. Evaporation is an adiabatic process, meaning there is no change in total heat (called enthalpy). The cooling from evaporation is returned as sensible heat when the water vapor condenses to form cloud particles. Likewise, when air rises in a thunderstorm it cools due to expansion, but this is also an adiabatic process and the heat is returned to the surface when the air descends.
It seems to me that the only process in your thunderstorm hypothesis that actually removes heat from the atmospheric system is the fact that clouds reflect sunlight. But if clouds can cool by reflecting sunlight why wouldn’t volcanic aerosols also cool?
Anyway, it does seem to be true that your thunderstorm-governor effect is not large enough to keep the surface at a near-constant temperature when the system is forced by, for example, volcanic aerosols. Even though they lower the global temperature by only a few tenths of a degree C. The GISS data show large increases in stratospheric aerosol loading accompanied by precipitous declines in global lower troposphere temperature and those declines persist for extended periods of time (years) while thunderstorms operate on daily time scales.

Reply to  Thomas
January 10, 2016 4:33 am

It seems to me that the only process in your thunderstorm hypothesis that actually removes heat from the atmospheric system is the fact that clouds reflect sunlight. But if clouds can cool by reflecting sunlight why wouldn’t volcanic aerosols also cool?
I think there is a far more important issue here. I am probably going to get the exact physics wrong, but let me have a go.
The CO2 thesis is that GHG in the atmosphere gets itself warmed up by absorbing radiation, and then re-emits it. So in a sense CO2 warms you at night by ‘reflecting’ radiation back down. Park that for a paragraph or two..
All heat loss and gain from the earth is ultimately by radiation, so as you have surmised, things like albedo are really crucial. Albedo stops heat getting in..
However that’s not the only game in town. The earth as a complete system including its atmosphere will radiate heat to space depending on the average temperature of its ‘surface’. What is important is to realise that the ‘surface’ that does this radiation is not the Earth’s actual sea/land surface, but something like the cloud tops or a large section of the (upper?) atmosphere.
The GHG thesis says the earth is wrapped in an insulator: The CO2 and water vapour loaded atmosphere. HOWEVER a big woolly sweater will keep you warm until the wind blows, when you need a windproof jacket as well.
What does this mean? This means that any vertical circulation that carries heat from the actual surface to high in the atmosphere is actually piercing the ‘insulation’ of the atmosphere, and allowing more heat to radiate to space, unimpeded by a GHG laden atmosphere. Which makes any CO2 variation practically irrelevant. GHG will have an effect over dry deserts where there is insufficient water to do much in the way of taking surface heat to a great height, and in particular GHG will reduce night-time temperature FALLS in the desert.
But over the oceans the water cycle will dominate. Both in terms of GHG and in terms of albedo and extra radiation from cloud tops that are partially or substantially beyond the bulk of atmospheric CO2 and other GHGs.
So that ties in with your question of ‘how can the water vapour be losing heat’ And the answer is ‘by direct radiation to space’. Albedo stops the energy coming in, but cloud top radiation is how it gets out again, and with a 4th law, its a pretty non linear curve there. And that radiation is happening high up in the atmosphere, beyond a large fraction of the CO2 which allegedly acts to prevent it reaching space. We know that energy has to be being lost, because warm wet air goes up and far colder rain comes down. And at the same altitude (surface level) too, so adiabatic nonsense has been neutralised…
And this is where we see, I believe, the ultimate flaws in the AGW model, it doesn’t really take account of the water cycle as an active system, it thinks in terms of radiation from the actual earth’s surface through an insulating layer that contains a constant amount of water and CO2 and GHG, where ‘constant’ means ‘only varying on quite long times scales’ – not on a minute by minute day by day cubic kilometre by cubic kilometre basis – because the computers are not powerful enough to model that, so its parametrised and left out of the actual models except as a lump sum term.
And that is why Ellis is, in my opinion, poking into a very sensitive and useful area. If we can show that the parametrisation of water is simply inadequate in terms of the models, and that there is even a moderate chance of water and the water cycle being such a major player that it dwarfs CO2, then there is an end if AGW.
Its not necessary to come up with a competing theory that predicts the climate better, in order to invalidate AGW – all that is necessary is to show that the AGW model is based on such a simplified picture of the atmosphere that it’s useless as a predictive tool.
Another corollary of water cycle rather than CO2 being a major driver of climate, is that it allows us to construct a lot of models with very long time delay feedback paths, like ocean currents that push warm water to polar regions and generalised ocean circulations, and these sort of huge time delays that are seen as the PDO and the NAO and all that other stuff, renders us a model with many time delayed negative feedback loops, and with a bit of non-linearity thrown in, we have all we need to define a chaotic system with no ‘average’ temperature at all, just a chaotic attractor that is ‘more or less’ where it is now, and around which we orbit, model wise, with quite considerable variation on all time scales, before flipping to a new attractor called an ice age…and the flipping may well be triggered by e.g. a Milankovitch cycle, but that is not the cause, that’s just the trigger…we are seeing this behaviour with all the ocean cycles including the El Niño, so its real all right, but no one has made a holistic model incorporating it to predict – if not the actual temperature fluctuations – the general order of the fluctuations – i.e. on what time-scales and what amplitude might we expect, and what peak values?
I am sorry if this is a bit incoherent. I haven’t really thought through the ramifications of all this, BUT the point is to start a ball rolling in terms of people with younger brains than mine thinking about how the water cycle actually might work and what this means in the context of atmospheric CO2., because I have a gut feeling that this is the elephant in the AGW room, and we should try and sketch it out.

Reply to  Leo Smith
January 10, 2016 4:34 am

damn. I hate this system where on missing brace can cause a whole section of text to go bold, and you cant correct your own post.

Reply to  Leo Smith
January 10, 2016 8:18 am

” Which makes any CO2 variation practically irrelevant. GHG will have an effect over dry deserts where there is insufficient water to do much in the way of taking surface heat to a great height, and in particular GHG will reduce night-time temperature FALLS in the desert.”
Except there is no trace of this in the surface record, none, zip, nada.
What hold the surface temp at night is the slow cooling rate of the surface of the land, asphalt, concrete, dirt, sand, grass and trees cool quickly, asphalt doesn’t, but the air, where the ground cools quickly, falls like a rock.

January 10, 2016 6:46 am

Hmm. After I had posted my long post I went back and looked at Gymnopserms post about IR radiation equivalent temperatures of cloud versus open skies.
The notches in the clear sky spectra show exactly the effect of CO2 when its NOT cloudy, and exactly the effect I mentioned when it was cloudy, that in essence the cloud radiation is almost independent of CO2 concentration.
What I couldn’t understand, and still can’t fully is that his contention that clouds radiate a lot less than hot ground, is supported by the evidence. Well some of it. It fully explains warm nights with cloud cover versus cold ones without.
And that makes me suspect I may have been wring to criticise Thomas when he said that albedo would be the only dominant effect for daytime clouds, yes Thomas perhaps that is in fact the case. My bad.
Looking at Gymnopserms data and doing a rough calculation reveals that yes, the difference between ground temps and cloud top radiation terms is the same as a T^4 rule would allow for ( (290k/215k) ^4 ~ = 3.3 or so) which is of course precisely the sort of difference one sees at night.,.,. but that doesn’t gybe with the fact that overall thunderstorms cool the surface, and so do clouds, by day.
So where is all the heat going? According to gymnosperm its not radiating to space in the infra red… Thomas must be right, something greater than a factor of 3.3 must be reflecting visible and UV light (not covered by IR thermometers) back into space before it even gets near the ground. And that’s albedo. So to get a net reduction in temperature that means net outflowing radiation must be greater than net inflowing..and that means that the reduction in radiation in terms of reflected visible light must be greater than the 3.3 figure gain that the IR radiation shows.
Well as a photographer who grew up in the days of manual cameras and hand held exposure meters, for sure full cloud is two, three or four stops, depending how thick it is. Which is in incoming (visible) radiation reduction of 4-16 times! And the bulk of the sun’s radiation is at or near the visible spectrum as are some of the absorption spectra dues to GHG etc.
What does this mean – well it means that when we stop looking at local transport, we have a figure – albedo modulation – more than big enough to account for the loss of radiation due to cloud cover. By day the net energy balance of cloud cover will be to cool – often dramatically, just as by night the reverse is true.
Ergo I may have been half right, because yes, clouds do carry energy way to space to radiate there, but not as much as direct ground and a clear sky does, BUT the important thing is, that the albedo of total cloud cover is massive and more than large enough to provide overall net cooling by day.
What this leads to is a picture of thunderstorm energy balance that goes like this. Warm wet air rises, and due to adiabatic lapse, cools condenses and transfers heat to the upper atmosphere, but if it were to fall as rain, would warm up again and become warm wet air again with only a little overall cooling from the 215k cloud top radiation.
BUT what must be happening is that albedo increase absolutely stamps on the incoming radiation. Not where the IR satellites are measuring it, but in the visible and UV spectrum. That’s where the net energy balance is adjusted, and because it’s cloud tops above the GHG effect, and radiation in a band where CO2 holds no sway, CO2 et al are completely irrelevant. So with clear skies CO2 rules, but when its cloudy, it’s almost irrelevant.
What this boils down to is this: ultimately we know that the energy balance of the earth has to be – for a steady temperature – a net zero. Negative feedback has to be in place to maintain it that way or we would go into thermal runaway. Radiative effects on an assumed steady level of incoming radiation are what AGW theories deal with – the T^4 radiation means that night time radiation loss is hugely proportional to temperatures, so the AGW alarmist – perhaps we should call them scientists of the Dark Side – of the planet are correct in terms of analysing energy loss (by night) but they have completely failed to account for the energy gain modulation of incoming radiation by albedo variation by day..
If there were no albedo variation then AGW would probably be almost correct. Gymnosperms graph showing deep water vapour and CO2 notches in the IR would rule the day (or rather the night!!!) .
But what we are seeing perhaps is that where it’s really at, by day, on the Light Side, is albedo. Anything that modifies albedo will have a massive effect on surface temperature. If a volcano erupts, that’s extra albedo, If a thunderstorm happens, that is, too. And consider, if land masses concentrate in polar regions, that’s a massive increase in albedo, as not only will the tropics be open sea and therefore more albedo than land, but the polar land will be snow covered and have a higher albedo as well. Land all at the poles might very well be a high albedo snowball earth.
So whilst its true that the effect of a thunderstorm is to suck heat from the surface and chuck it up high which cools the surface, it also must be warming the upper atmosphere somewhat, and the energy hasn’t been lost, just moved. But the albedo gain will reflect massive amounts of extra sunlight away from the surface so that such radiation as there is in the 215k cloud top radiation is enough to overall cool the system because the incoming radiation is now far lower. That’s the only way it makes sense to me. And of course if that results in rain and clear night time skies, you will get a very nice extra cooling effect. Albedo is no use at night – it doesn’t cool you – just stops you getting too hot by day.
This brings a much clearer picture to me than I had earlier today, about the net effect of cloud cover..a subject I couldn’t get my brain around, which was this. Is the net effect of cloud cover overall warming, or cooling?
The IR graphs absolutely show that clouds massively reduce IR radiation which is the dominant form of heat loss by night, and of course by winter too where there is ‘more night than day’. Cloudy polar regions ought to be much warmer in winter than cloudless ones. So here where heat input is transported heat from the tropics, and the cooling is IR dominated, maybe we expect to see overall warming if its cloudy
But in the tropics, where the heat input is direct solar radiation, extra cloud cover by day will be more than enough to compensate for extra cloud cover by night (and that may go if there is a thunderstorm to ‘clear the air’) , and clouds will have a net cooling effect overall.
And that confirms my suspicion. CO2 and the GHG effect is real and it affects night time radiative LOSSES a bit, but its almost completely irrelevant by day, where albedo variations from cloud cover will dominate the energy balance, especially in the tropical regions.
And the argument from scientists of the Dark side, that water vapour will amplify temperature variations, is true. By night. And somewhat in the Polar regions. By Day its exactly the opposite effect, extra water vapour means extra cloud, and that means a lot cooler overall, in the tropics.
And as there is a lag in transporting heat from tropics to poles, that means that positive fluctuations in cloud will first cause high latitude warming, where heat loss by night dominates, before cooler ocean currents driven by cooler tropical temperatures will lead to colder winters, and then snow ice and less cloud, and a nasty big freeze up.
And so on ad infinitum Colder subsea currents will then cool the tropics, reduce cloud cover and the tropics will start to warm up again…
And it may explain why the Antarctic is somewhat different from the arctic – the heat transport there is weird. It is a land mass surrounded by sea and the circumpolar current more or less isolates that land mass from tropical heat. It would be interesting to compare polar cloud cover over the last few guess is there is less cloud in the Antarctic.
[Long analysis. Thank you. .mod]

Reply to  Leo Smith
January 10, 2016 9:38 am

Thank you for taking the time to put that into words.

January 10, 2016 7:11 am

The GHE & blanket analogies are both inadequate since they don’t account for the water cycle. If I chop wood on a cold day in a heavy coat, sweat will cool me off. If I blanket my house & don’t turn down the heater it’s going to get warm. The sun doesn’t have a thermostat, although there are both short and long term fluctuations in output. The earth does have a powerful thermostat in the clouds and water vapor cycle. However: 1) they are difficult to model and 2) not due to man.

Peter Foster
January 10, 2016 10:22 am

Your world map of the TRMM data shows the main band of thunderstorm activity (in the Pacific and Atlantic oceans) to be north of the equator.
Given that the Earth is some 5 million km closer to the sun in the southern hemisphere summer (Dec) one would expect the global average activity to be slightly south of the equator.
Is there any known reason for it to be North of the equator ?
Here in NZ the intensity of sunlight mid summer is far greater than in the Northern Hemisphere (in terms of how quickly one gets burnt by the Sun) A lot of that will of course be due to significantly lower air pollution but not all of it can be so attributed.

Reply to  Willis Eschenbach
January 12, 2016 9:05 am

“Energy is neither created nor destroyed, so where are the >400 W/m2 of DLR going if not into the ocean?”
First of all:
Several of the power flux balance graphics I see in Bing images including Fig 10 Trenberth et al 2011 look something like this: 340 W/m^2 ToA, about 30 % reflected, 102 W/m^2, about 78 absorbed and 161 to the surface. There is a frequently, but not always, +/- 330 W/m^2 GHG perpetual heat loop.
So where y’all getting DLR of 400+ W/m^2? Some kind of magical breeder?
Second of all:
As noted elsewhere the formation of snow/ice on the polar & Greenland ice sheets can suck up yuuugge amounts of energy Btu’s, i.e. Wh.

Reply to  Willis Eschenbach
January 10, 2016 1:41 pm


Next, density. Again from the MIT paper, seawater at 30°C and 35 g/kg salinity is about 2% heavier than freshwater. So it takes 2% more energy to evaporate the extra weight, or a total of 2350 * 1.02 ≈ 2400 MJ.

Good summary using 24/7/365 average values.
I would likely challenge the additional 2% density correction, as the evaporation energy reference you quoted is already using nominal (salt-solution density) seawater.

Reply to  Willis Eschenbach
January 10, 2016 2:44 pm

” at the TAO buoy at 165°E, what physical process is providing over 400 W/m2 of energy 24/7 to the ocean surface if the energy is not from DLR?”
The majority of the heat at the surface is the surface, then the heat stored in all the water vapor, best I can tell it’s got to be a lot.
You point an IR thermometer straight up there isn’t near 400W/m2. At 41N (what I know the best) clear sky it 80F to over 100F colder than the surface. A 50F day it was about -40F. This was at noon.
And you’re wrong about the thunderstorm line, the vast majority of the water vapor is still there, it ends up raining out over the continents, and that tropical air makes a 10 to 20F difference in temps. I have the same 20 or so F swing in daily temp, but I can have clear days a couple days apart that are 20F different max temp, separated by a line of Thunderstorms.
I m not complaining about your ability to detect the thunderstorm line in the tropics, just that I know the water that evaporates in the tropics has a big impact in extratropic and subpolar thunderstorms and air temps.

Reply to  Willis Eschenbach
January 10, 2016 7:44 pm

” Set the sensor altitude to 0 km, set it to looking upwards, and choose “Cirrus clouds”. You get 418 W/m2 of DLR.”
Oh, I didn’t notice the cloudy conditions. Clouds warm the sky, as compared to clear skies. What I’m measuring is the IR window from 8u- 14u to space, or whatever the optical column terminates with, but you can convert the temp to a flux rate, and add the flux rate for co2. I have seen the entire Co2 flux listed as 22 W/m2, I think that’s about 12F at -40F, but the largest measurable DLR is from cloud bottoms, I can measure a 70F or 80F difference from clear skies off cloud bottoms. So the 418W/m2 seems to be from a lot more than Co2. Even 95F humid days, the sky is still quite cold.
If you can borrow an 8-14u IR thermometer even with the limited view, I think it’s enlightening.

January 10, 2016 1:39 pm

Rain can evaporate as it falls but moisture from the surrounding air can also condense on cold rain droplets. This is why relative humidity at the surface often falls during a rain shower. Rain droplets are born relatively high in the troposphere, where the air is cooler due to the lapse rate. Therefore, rain is usually cooler than the air it is falling through when it reaches close to the surface. Water is incompressible so, unlike air, rain does not heat as it falls—except for some minor heating due to friction with the air it is falling through.
However, it seem to me that no amount of evaporative cooling can remove heat from the global troposphere because evaporation is an adiabatic process. Evaporation consumes heat, which cools the surrounding air, but when the evaporated moisture condenses back to liquid, the heat is returned through the latent heat of condensation. Heat is moving around in the system but no heat leaves the system.
Therefore, it seems to me that no amount of thunderstorm evaporation could overcome even a small amount of greenhouse gas heating. Greenhouse gasses “trap” radiant heat in the atmosphere, which causes the atmosphere to equilibrate at a new, higher, temperature. Evaporation and condensation are internal processes and they do not change the exchange of heat between the troposphere and deep space.
Thunderstorms can have a damping effect on global temperature but the effect would be mostly, or only, due to the fact that clouds reflect sunlight. This effect does not seem to be sufficient to overcome the cooling effect of a those rare volcanic eruption that are large enough to spew large amounts of reflective particles into the stratosphere. Those particles reflect sunlight and that causes global cooling because less of the radiant energy from the sun reaches the surface.
Since the damping effect of thunderstorms does not seem to be sufficient to counter act the cooling effect of stratospheric aerosols, I doubt that the effect could counteract greenhouse warming. I suppose it’s possible that the effect might require years to counter the rapid cooling that occurs with increased aerosol loading but I doubt that too because global temperature falls when stratosphere aerosol levels increase, then rises again when the aerosols settle out.
The two coldest years in the UAH temperature recored occurred after the eruptions of El Chichon and Pinatubo, when aerosol loading caused stratospheric optical depth to increase by as much as a factor of 24.

Reply to  Thomas
January 10, 2016 2:54 pm

If nothing else it carries billions of gallons(?) of 80F water to someplace that’s 20, 30F colder.
How much heat is that compared to a dry atm column?

January 10, 2016 1:48 pm

At 12:56 PM you wrote:
“Next, density. Again from the MIT paper, seawater at 30°C and 35 g/kg salinity is about 2% heavier than freshwater. So it takes 2% more energy to evaporate the extra weight, or a total of 2350 * 1.02 ≈ 2400 MJ.”
The salt doesn’t evaporate. It stays in the ocean. Only water vapor leaves the surface.

January 10, 2016 4:16 pm

We know that heat loss to space can only occur via radiation
We know GHGs absorb most of the radiation that hits them of the correct wavelengths
We know this absorbed radiation is mainly lost to other molecules via collisions. (At altitudes of 20km the MFP is 9.139e-07 metres and the Collision frequency is 4.354e+08 / sec)
we know that heat loss to space can only occur where the mean free path is sufficiently great and radiation finally leaves the earth
we know that only radiation from the ground which is not absorbed by ghgs reaches space
it must therefore be true for radiation from cloud tops – radiation that would have escaped from the ground will also escape from the cloud tops (clouds will emit a near BB radiation pattern). BUT radiation that excites ghg molecules will still excite ghg molecules and will be indistinguishable from the IR from the ground.
Radiation from cloud tops is therefore exactly the same as radiation from the ground from where the heat was transported.
If this is the case then I do not see where thunderstorms change the overall energy balance. Do they simply transport the heat to cooler areas. Where is the mechanism for reducing the overall energy content of the globe?

Reply to  sergeiMK
January 10, 2016 4:47 pm

” Where is the mechanism for reducing the overall energy content of the globe?”
The energy released as water vapor become liquid water is lost to space at cloud top, if that alone doesn’t explain it, add that cloud top is on the other side of a large percentage of the atm GHGs when this state change takes place.
Consider the mechanical energy of a thunderstorm over a 20 or 39F weather front. Then account for the water carried in and condensed.

Reply to  sergeiMK
January 10, 2016 5:33 pm

“Where is the mechanism for reducing the overall energy content of the globe?”
How about the snow and ice accumulating at the poles and Greenland? Making a ton of ice sucks up 288,000 Btu, 84,407 Wh.

January 10, 2016 4:49 pm

I think the answer is that clouds reflect visible light, which has too short a wavelength to be absorbed by greenhouse gasses. Look at a photo of the earth from space. Clouds are bright.

January 10, 2016 7:04 pm

Converting a ton of water into a ton of ice requires 288,000 Btu/ton, aka 317,376 Btu/tonne.
One cubic meter holds 1,000 kg, aka 1 tonne of water. Yes, ice is less dense, but does it really matter?
The area of Antarctica is about 1.4E7 km^2, 1.4E13m^2. A meter deep layer of ice would be, duh, 1.4E13m^3, aka 1.4E13 tonne. Converting that tonne of water to ice absorbs 4.44E18 Btu.
ToA spherical surface area is 5.12E14 m^2. ToA solar input is 340 Btu/m^2. Over 24 hours the complete ToA spherical surface area would receive 1.74 E17 Watts, 5.95E17 Btu/h, total over 24 hours 1.43E19 Btu.
It would take 3.21 meters of ice to absorb the entire ToA input of 340 Btu/m^2. The additional 2.0 W/m^2 of anthropogenic RF between 1750 and 2011 could be absorbed by 0.02 m, 0.74 inch. RCP 8.5 could be absorbed by 0.08 m, 3.16 inch.
Water’s latent heat of evaporation/condensation is about 1,000 Btu/lb, 293 Wh/lb. Sensible heats: water, 1, air, 0.24 Btu/lb-F.
This all makes it pretty clear why water vapor is the 1,000 pound/454 kilo gorilla of all the GHGs.
A similar analysis could be applied to the ocean surface.
The numbers are subject to review. Everybody makes mistakes. Want it all metric, help yourself.

January 10, 2016 7:24 pm

@ Willis.
“And at the altitude of the cloud tops, there is very little of any of the GHGs. This means that whatever heat is left in the air at the top of the cloud towers is free to radiate to space.”
are you sure?
That statement says we have a real problem because above the cloud tops there is little GHG so no saturation. Increase the GHGs and you have a linear not log relationship to the GH effect.
for discussion below see 2009 Schlatter_Atmospheric Composition and Vertical Structure_eae319MS-1.pdf
1st From wiki
Genus Cumulonimbus (heap, storm/rain)
Altitude 2,000–22,900 m (6500-75000 ft)
so upper limit is about 23km
at this altitude co2 is still well mixed
And collisions occur every 45ns so a photon would only be able to travel 13.5 metres so plenty of collisions of photons between GHGs before top of clouds and space. and since the mfp is only a couple of micrometres there will still be a lot of heat transfer molecule to molecule by collisions. heat transferred to non ghg molecule would of course have to be retransferred to a ghg molecule in order to escape to space.
The only ghg in short supply at these altitudes would be water vapour so that would allow some heat to escape.

Reply to  sergeiMK
January 10, 2016 7:57 pm

” are you sure?”
Air density drops pretty quickly, ghg is a fraction of that.
Beyond that, no.

January 10, 2016 7:35 pm

Thanks. You wrote:
“Remember that the metric at issue is not the total amount of heat in the system. It is the surface temperature.”
That’s part of the problem. GHGs should increase total heat in the system but temperature is not a measure of total heat. Enthalpy is a measure of total heat. It’s expressed in kJ/kg of dry air. At a given temperature the enthalpy will be higher when the humidity level is higher. So knowing only temperature doesn’t tell us anything about the energy in the earth’s system. The total heat on a 110 °F day in Phoenix could be the same as in Orlando when the temperature is only 80 °F. Air temperature is at best a rough proxy for total heat.
The air at the top of a large cumulus cloud might be hotter than it would be without the convective activity, even though it is still very cold. That might mean that it radiates more heat to space then it otherwise would but that depends on the distance it takes, at those rarified altitudes, for a LW IR photo to strike an air molecule. If that distance is short, there is no net removal of heat from the system.
It seems like the the only way thunderstorms could remove significant heat from the troposphere is by reflecting sunlight.
In the article you wrote:
“This means that [the thunderstorm] effect will be to maintain the same temperature, regardless of reasonable-sized fluctuations in the amount of forcing. Clouds don’t know about forcing, they form and disappear based on local conditions.”
Both El Chichon and Pinatubo ejected sufficient aerosols into the stratosphere to cause global cooling that lasted for several years. The two lowest annual temperatures in the UAH record correspond with those two eruptions. El Chichon occurred when the ENSO 3.4 index was low and it took several years for temperature to recover. Pinatubo occurred when the ENSO index was high, which might explain why the cooling was less and the recovery was faster, but it still took several years to recover to the mulit-decadal average.
(ENSO data from NOAA.)
So your posited thunderstorm effect seems to not be able to quickly overcome the forcing from stratospheric aerosols. However, it would be interesting to take a look at what happened to tropical storms directly after those eruptions. If they really did decrease, that might support your theory.
It’s been interesting discussing this with you. Thanks for taking the time.

Reply to  Willis Eschenbach
January 11, 2016 8:20 pm

You wrote:
“It is clear from your comments that you have not read [your articles on volcanoes], or else you wouldn’t persist with your claims about the strong effects of volcanoes.”
There have been only two volcanoes in the past 35 years that produced significant amounts of stratospheric aerosols, El Chichon and Mount Pinatubo. (see Your “Spot the Volcano” articles do not seem to address either of these volcanoes.
After El Chichon and Pinatubo, we see cooling in the UAH lower troposphere record and the GISS surface record. Both eruptions occurred when large El Niño events were forming. El Chichon happened just after ENSO peaked and temperatures soon fell to the lowest point in the entire record. Pinatubo, which was much larger, occurred when ENSO was still building and resulted in the second coolest period in the record.
I think you’re going down the wrong path looking at surface cooling due to evaporation because evaporation doesn’t result in heat leaving the system, it just moves it around. Moving heat to the poles doesn’t reduce the energy content of the system because there is much less radiant heat loss from the poles (because they’re cold).
I think it is only reflection from clouds that causes significant heat to leave the troposphere when thunderstorms form. If clouds increase, then more of the suns radiation will be reflected and that can cause overall cooling.
Actually, the “proof” of your theory may be visible in a graph of ENSO and the first few decades of the UAH temperature record. ENSOs come fairly regularly, every three to six years. Global temperature tending to rise to a peak a year after the ENSO index peaks, then temperature falls back to a low value after the ENSO passes. The temperature varies by about +/- 0.5 °C from peak to valley. This could be your thunderstorm “governor” overshooting, then undershooting, but still holding the global temperature in a fairly tight band.
When the trade winds are blowing strong, they push warm surface water into the western Pacific so it get’s warm. But that causes a lot of convective activity, which causes cooling, which slows the trade winds, so the warm water sloshes back (or water stays where it was and gets warmer due to a lack of wind?), which causes the trade winds to speed up, which pushes warm surface water to the west, which causes a lot of convective activity, which … etc.
It’s an oscillating governor. I recall that happening to an old lawnmower engine that I had “fixed” when I was younger (much). It just sat there and revved up, then revved down, then up again, then down.
If you graph it out, it looks like Pinatubo disrupted the cycle and things didn’t recover until the super El Niño of 1997.
I suggest you set aside localized evaporative cooling as a counter force to forcings and instead look at reflection and ENSO as a manifestation of your theory.

Reply to  Willis Eschenbach
January 12, 2016 8:22 pm

Thanks for the great post.
Sure the surface is where we live but if CO2-caused warming is going to be counteracted by thunderstorms there needs to be actual increases in radiant heat loss. More clouds reflect more visible light, so that’s one source of increased radiant heat loss. I agree with your excellent point that Hadley Cells and Polar Cells move can heat to drier areas where heat can be lost by radiation to deep space. Deserts are probably the primary source of any additional radiation loss because they are much hotter than the poles and radiation is to the fourth power of temperature.
I keep harping on about El Chichon and Pinatubo because, following these volcanoes, there were clear and significant increases in upper atmosphere opacity due to aerosols, followed by cooler periods lasting several years and both those cool periods were the coolest in the 35 year record.
Volcanic cooling is pronounced enough that one wonders if there would be an upwards trend in the satellite temperature record at all if there had been significant volcanoes in the last half of the record, or no volcanoes in the first half.
Try putting UAH temp, GISS stratospheric optical depth and the ENSO index all on one chart. It’s pretty fascinating.
After Pinatubo, the ENSO index was higher than normal for four or five years. Since a high ENSO is normally followed by a high global temp, that might be evidence of your thunderstorm effect working to counteract cooling. It might be worth your while to look into that.

Reply to  Willis Eschenbach
January 13, 2016 5:03 am

willis you interesting say
“The issue is and has always been the temperature here at the surface. For example, we don’t care if the thermosphere goes up or down a bit. And we don’t care if the total heat in the ocean goes up or down a bit. In general neither of those affect those of us living here on Earth’s surface.”
Which must really upset the likes of satellite measurements quoters. Sattellite temperatures do not measure surface temperatures but at TLT levels.
Are you saying that these have no relevance to surface temps?

Reply to  Willis Eschenbach
January 13, 2016 6:52 pm

Spaceba, Sergei. It’s a good point and one that I had intended to make in my last post. It’s not possible for the surface to heat without the lower troposphere also heating. Willis can’t logically argue both that convection carries heat from the surface and that convection does not cause heating of the lower troposphere.

January 11, 2016 4:50 am

Sergei, the difference regarding CO2 is not in the ratios of the gases. The difference is in the density. At an altitude of 10 km, for example, the pressure is only a quarter that of the surface. So it is correspondingly easier for a photon to make it to space unhindered.
please see new plot
As I said and you agreed water vapour is not going to have much effect above thunderclouds. However CO2 is well mixed even at 100km so increasing CO2 will widen absorption band and reduce the transfer of heat to space. from the plot the current transmittance at 16um is still only 10% at 15km
However, the main change is that as your lovely graphic above neatly illustrates, there is very little water vapor up high.
agreed. The IR blocked by H2O vap will be the only change by which thunderclouds will be able to cool the planet
But you must then agree that increasing CO2 will actually cause a reduction in IR transmitted from top of cloud to space reducing the cooling effect.
As a result, between the reduced pressure and the reduced water vapor, any heat at altitude has a much easier path to space than does heat at the surface.