I’m sure WUWT readers will recall this excellent guest post at WUWT just over one year ago:
Now published in E&E Volume 21, Number 4 / August 2010
The thunderstorm thermostat hypothesis: How clouds and thunderstorms control the Earth’s temperature
Authors
Willis Eschenbach
Abstract
The Thunderstorm Thermostat Hypothesis is the hypothesis that tropical clouds and thunderstorms actively regulate the temperature of the earth. This keeps the earth at an equilibrium temperature regardless of changes in the forcings. Several kinds of evidence are presented to establish and elucidate the Thermostat Hypothesis-historical temperature stability of the Earth, theoretical considerations, satellite photos, and a description of the equilibrium mechanism.
See it here, PDF is available (£18.00 worthwhile to support E&E in my opinion). Or, read the WUWT version here:
Thanks Willis, very helpful.
I think that the best indication we currently have as to whether any such effect is large or small is the ‘coincidence’ of albedo trends changing just when the jets started moving equatorward again around 2000.
Interestingly the earlier stratospheric cooling stopped and there may now be slight warmig, ozone levels started to recover and the sun became less active.
All around the same time and all the reverse of what had been going on for the previous 30 years or so.
That still leaves plenty of room for your hypothesis though because the tropics are an integral part of the mechanism.
“One is that oftentimes thunderstorms are nearly as tall as they are wide (10 km or 6 miles). This means that their albedo doesn’t change much with changing solar angles.”
They cast longer shadows though.
Re: vukcevic
Agree that there is also a direct (oceanic) Pacific connection. Where we seem to disagree is on the direction of causation. My instinct is that geomagnetic variation is a byproduct of other processes. I do not suspect that magnetism drives weather/climate. However, it appears to be a useful indicator of patterns & processes (e.g. geography, lunisolar tides, hydrology).
Willis, Congratulations!
For another comment I did a quicky look at afternoon rain on the east coast (USA)
Florida had about 20 -25 days of rain per month. Georgia,South Carolina and southern North Carolina had 20 days of rain per month. In the middle of North Carolina – between Fayetteville and Sanford (or Rocky Mount) a distance of fifty miles further north, the number of rain days changed from 20/month to 10/month.
The Sanford area is also the present “snowline” It snows regularly further north and rarely further south.
I think this hypothesis has much explanatory power.
I haven’t been able to read the paper but I have some data which I think very much supports it.
Outgoing long-wave radiation is a method to measure cloudiness. OLR falls when there is more clouds and rises when there is less. [There is also less solar energy getting in as a result of more clouds].
There are cloud cover estimates which semi-match the OLR numbers but I don’t think we have good measures of cloud cover and with my work on the ENSO, I know that OLR and cloudiness are closely related.
http://a.imageshack.us/img130/8107/eqauolrclouds.png
Here is the OLR for the wide-tropics 20S-20N versus the ENSO going back to 1974. The OLR values don’t change a great deal (+/- 12 Watts/m2) but this area represents a huge percentage of the Earth.
It represents 34% of the total surface area and 41% of the total solar energy received by the Earth so it would have a big enough impact.
Well, the OLR (which is a close indicator of cloudiness) does track the ENSO very closely and it does vary by large enough amounts to support Willis’ hypothesis. [One would also have to build-in the resulting reduction of solar energy from the clouds to complete the picture].
http://a.imageshack.us/img718/7858/ensovsequatorialolrjune.png
Thanks Willis! Seems “Climate” is a function of the size/strength of Hadley, Temporate, and Polar Cells, and the strength/location of Jet Streams (as well as much more;-) Has anyone produced a graphic of the globe similiar to the one you show at the deepest point of the Glacial periods? Wouldn’t the cells be different? Thanks again!
kadaka (KD Knoebel) says:
July 26, 2010 at 1:59 pm
Thanks, again. those were a couple of very clear explanations. I appreciate your taking the time to answer thos equestions.
Re: Bill Illis
Thanks for the notes & graphs. That inspired a look at how cumulative OLR relates to OHC — see some of Bob Tisdale’s OHC graphs here:
1) http://bobtisdale.blogspot.com/2010/02/ohc-linear-trends-and-recent-update-of.html
2) http://bobtisdale.blogspot.com/2009/09/enso-dominates-nodc-ocean-heat-content.html
Mr. Willis Eschenbach,
I read your article a year ago and liked it very much.
I have just republished it in my Website, Observatorio ARVAL, at
http://www.oarval.org/Thermostat.htm (English) and
http://www.oarval.org/Termostato.htm (Spanish).
I corrected a couple of obvious typos in the original and did a careful Spanish translation.
I hope this pleases you, and will make any and all changes you might suggest.
Thank you very much,
Andrés Valencia
Observatorio ARVAL
http://www.oarval.org
Andres Valencia says:
July 28, 2010 at 4:59 pm
Qdo. Andres;
Mil gracias pa’ traducirla en Español. Es muy interesante ver mis palabras en otra lengua.
Lo mejor a Vd,
w.
Dear Willis,
In “But where is that mechanism?”, you write “And curiously, in this thought experiment called “A Day In the Tropics”, there is such a timeless point of view, where not only is there no day and night, but where it’s always summer.”
How could this GOES-West view be “always summer”?
Best regards
Dear Willis,
In the next paragraph down, you write “And it’s always summer under the sun.”
It is always noon, but not summer, I think.
Best regards
Andres Valencia says:
July 29, 2010 at 8:34 pm (Edit)
Andres, thanks for your question. The sun is north of the Equator during the NH summer, and south of the Equator during the SH summer. When the sun is north of the equator, it is not summer in the south, and vice versa. So yes, summer follows the sun …
You ask:
It’s not, because it is only an approximation of the view from the sun. It is in geostationary orbit directly over the equator, so the sun moves north and south in the GOES-West view.
Thanks,
w.
Thanks Willis,
I had mixed up your thought experiment with the GOES view.
Best regards
I have added at the end:
ARVAL Note:
The global images from the meteorological satellites in the visible, infrared and water vapor wavelengths show the Earth’s recent cloud cover and storm activity.
See NRL Monterey Global Imagery (GOES 11/13, METEO 7/9, MTSAT2).
http://www.nrlmry.navy.mil/sat-bin/global.cgi
Hi Willis,
Hope you noticed this in a buried thread.
We’ve had a some scorching heat lately here in south central Texas but I noticed something that begs for quantization. We get afternoon thunderstorms here quite often on hot days with 40+ mph winds and hail. When one of those boomers rolls through, often right at the peak temperature, it will lower the air temperature 10-20 degrees F in a matter of minutes for miles in every direction around the center of the cell. The rest of the day stays that much cooler. This cooling won’t register on max/min thermometers as they record the low at night and have already recorded the high, or very close to it, in the afternoon before the storm rolls through out of a clear blue sky. As far as total degree-hours for the day a thunderstorm can knock it down considerably – say if instead of max/min the record was hourly and you added up 24 records for the day and divided by 24.
My question is whether there’s any database or proxy for thunderstorm frequency such that we could estimate whether the frequency changing on a global basis from year to year. A second question would be how these are treated in GCMs – I’m guessing it’s just a constant like they use for albedo.
One final thing – there must be a plethora of hourly records from automated stations that also report min/max used in GISS etc. and also metadata about thunderstorms coming through, particularly I’d think at airports. It would be interesting to see the difference, if any, in station records averaged by daily min/max and averaged by hourly temperatures.
As I wrote this a second cell just passed over me. The first knocked the temperature down from 103F to 93F at 2:30 and this second an hour later dropped it another 10 degrees to 83F at 3:30PM. That removed probably 100 degree-hours or more from our day which is quite a bit of missing energy at the surface.
Dave, thanks for the comment. I don’t know of any database or proxy for thunderstorm numbers, it is a moving target that would be very difficult to define.
In GCMs, thunderstorms are way sub gridcell size, which is typically 200 km on a side. So they are parameterized, which is how climate scientists describe unsupported assumptions …
There definitely are differences between the average of the min/max and the hourly temps. However, many stations don’t have hourly temps, so we’re stuck with (min + max) / 2 for the time being.
And yes, as your experience with local thunderstorm cells highlights, thunderstorms cool the surface greatly.
Thanks,
w.
Willis Eschenbach – August 24, 2010 at 11:33 pm
However, many stations don’t have hourly temps
Couldn’t we use the stations that have hourly temps as proxies for those that don’t? Just thinking that 12 hours of proxy using the same unit of measurement is more likely to be skillful than tree rings at 1000 years.
Nice work. In Ohio, plenty of convection in August. We tend to catch the tail end of hurricanes. Difficult part is quantifying thunderstorm energy on a historical basis. Going forward, the satellite data for albedo should be good. Is that data kept from 1979 forward?
A repeat of my post at other site…
Mr Eschenbach — Wow!
As an engineer whose Dad was a meteorologist i applaud your essay.
I have been asking “Where’s the control system theory in all this?” It’s an equilibrium seeking system subject to the math of feedback controls.
I see a couple prior comments mentioned it, the ones on waveforms and negative feedback.
My opinion is the climate models are just in the appetizer stage, the main course of straight thinking on this matter is still out there in the kitchen .
You are more on track than the rest i’ve seen.
And your points about water vapor are right on target.
Please see my tweak of the ocean heat guys a couple years ago,
comment #96 on this ocean heat content blog,
http://www.realclimate.org/index.php/archives/2006/08/ocean-heat-content-latest-numbers/comment-page-2/#comments
I believe the regulating mechanism you seek lies in the slope of the saturation pressure curve for water.
Given that our atmosphere has a particular weight hence a particular pressure, there exists a temperature around which the vapor pressure of water will affect the thermodynamics of air with maximum effect. That is, it will make density of air change more than it would from temperature alone, and even more significantly will optimally affect its specific heat as a working fluid in your heat engine. Mother nature loves a balance!
Were i forty years younger i’d try to calculate it for you. Partial derivatives of air’s specific heat and density(hence convection propensity) wrt H2O and CO2 content….
I spent a lifetime fixing feedback control based regulating systems and am fascinated by the math involved. See any text on Modern Control Systems.
That math was discovered by Descartes but set aside as an interesting curiosity. Well, that is until WW2 when the German scientists found it’d make their rockets work. The German texts were brought back as a war prize along with Dr Von Braun who explained them to our guys.
But i digress.
Anyhow – if i can find my steam tables and Dad’s old “Climate and Man” textbook i might try to horse out some simple approximations …
meantime, i applaud you as the first climate guy to put Descarte before the hors d’ouvres.
Sincerely, old jim hardy