This technical note shows how the ocean surface temperature responds to the solar radiation at the top of the atmosphere. It highlights that the ocean surface is temperature constrained to the range -1.8C to 30C over a yearly cycle apart from less than 1% of the ocean surface near land masses in the tropics where the deep convection cycle is disrupted.
The atmosphere over oceans is dullest at 26C but the amount of reflective cloud increases above and below this temperature. At 30C, the reflected EMR achieves the level where just enough sunlight reaches the surface to evaporate the water required to power the deep convection engine driving global-scale air circulation. This precise temperature limit is similar to a speed governor on a reciprocating engine where the fuel supply is regulated to maintain engine speed. In the case of the oceans, the surface sunlight is the energy source that is being restricted by reflection from ever-brightening cloud with rising temperature from 26C until the 30C limit is reached.
The operation of the atmospheric heat engine is briefly explained with reference to the atmospheric temperature profile; referring to well-known meteorological phenomena such as Level of Free Convection and Convective Available Potential Energy. Those familiar with heat engines will draw analogies to the well-known Carnot cycle.
The factors causing climate change are discussed with reference to how the solar radiation is gradually changing over any point on Earth’s surface due to Earth’s orbital changes. The observed changes in temperature are the result of changing solar radiation at the top of the atmosphere except the Equator where the temperature has remained with no trend since satellites have been recording surface temperature over the past 40 years.
The actual temperature measurement in the Nino34 region in the tropical Pacific is compared with the output from climate models that all predicted warming where there has been none. No climate model has predicted the observed temperature trend in the important Nino34 region of the tropical Pacific. All have a warming trend where there has been a slight cooling trend observed.
Ocean Surface Temperature
Global oceans have a near linear temperature response to solar electro-magnetic radiation (EMR) at the top of the atmosphere (ToA) averaged over an annual cycle with the exception of two clear inflection points and a third less distinct inflection point.
The distinct inflection points are at 4C and 30C. The less distinct point is at 26C where there is an increase in the slope from the near linear range between 4C and 26C.
The following image indicates the temperature distribution across the global oceans averaged over a yearly cycle.
From the image, the bright red areas are the only locations warmer than 30.5C. These are the locations where mid-level air divergence to land disrupts the normal deep convection over the oceans. The green regions are where sea ice forms and melts. Perennial sea ice is shown as grey as it responds to solar EMR in a similar way to land. The yellow hatched area in the central Pacific is known as the Nino34 region, which is used to forecast the different phases of the Pacific Ocean that influence global weather.
The inflection point at 30C is distinct and immutable with the present atmospheric mass. At 30C, the ocean heat engine literally runs out of steam due to persistent cloud formation. This can be observed globally by the reflection of the atmosphere in response to the surface temperature.
The most reflective white zones in the above image align with the bright red (>30.5C) zones in the temperature image. The inter-tropical convergence zone is also distinctive in the high reflective power near the Equator.
It is noticeable that the Southern Ocean is also highly reflective. This aligns with the surface temperature in the range -1.8C to 15C where convective instability cannot occur and some of the 15C to 22C where convective instability is not cyclic. The Southern Ocean and small regions in the North Pacific and North Atlantic have bright condensing/solidifying cloud due to air in circulation from low latitudes being cooled at higher latitudes.
The darkest, muddy yellow region of the reflection image broadly aligns with a temperature range from 24C to 27C where the ocean atmosphere is least reflective.
Deep Convection – The Heat Engine
The atmosphere over the oceans behaves as a self-governing heat engine. The Level of Free Convection (LFC) acts as the expansion nozzle between the high pressure and low pressure sides. The ocean surface is the evaporator and the atmosphere above the LFC is the condenser/solidifier. The formation of reflective cloud as the water vapour solidifies to highly reflective ice particles, limits heat input to the ocean by reducing surface insolation.
The following image examines the atmospheric isenthalpic lines for a surface temperature of 22C (295K), the lower limit of cyclic deep convection. It shows the moist isenthalpic line and the corresponding dry isenthalpic line through the LFC.
In this situation the LFC is at 2000m. The water vapour condensing zone below the 273K level will have 14.4mm equivalent of water vapour and the solidifying zone, where reflective ice cloud forms, will have 7.9mm. A complete cycle, under saturated conditions after cloudburst at 295K surface temperature, will have 40 hours of cloud forming and 35 hours clear sky or clear sky for 47% of the cycle. Normally instability occurs before the air above the LFC is completely dehumidified so cycles are shorter than 75 hours.
Here it is noted that moist air has lower density than dry air but there is a level of moisture where dry air will remain buoyant on moist air below. This is the altitude where the LFC forms. As the air above the LFC radiates energy to space, the water vapour above the LFC condenses or solidifies to cause the drying air to compress. The moist air below the LFC is more compressed than what it would be if free convection was not inhibited. The potential energy builds until the column becomes unstable giving rise to cloudburst. Cloudburst is the result of water vapour bursting into the dry zone due to the increasing buoyancy of the air as it becomes saturated. An analogy is that of a balloon bursting where compressed air suddenly expands once the balloon is punctured. The LFC is like the thin membrane of a balloon.
The profile of the atmosphere changes significantly when the surface temperature reaches its 30C (303K) limit.
The LFC in this condition is now much higher at 5600m. The moisture between the LFC and 273K is now only 7.4mm and the moisture above is 8.9mm. Under fully saturated conditions above the LFC, a full convective cycle will have 45 hours of cloud forming and just 8 hours clear sky. Without advection, the clear sky will persist for only 16% of the cycle time.
From observation at tropical moored buoys, the heat engine requires surface insolation averaging 200W/m^2 to keep the engine fully stoked. The drop in surface insolation with temperature is observed to be 20W/m^2/C. Pacific moored buoy 8N, 137E exhibited a trend of -21/2W/m^2/C for the six year period 2015 to 2020.
Atlantic moored buoy 4N, 23W exhibited -19.3W/m^2/C over the same period. The two charts use daily averages of one day for the Atlantic buoy and 5 day averages for the Pacific buoy. The thermal lag in the Pacific column averages around 25 days while the Atlantic, which is cooler on average, has much shorter response time of just 5 days.
From an energy balance perspective, it was observed that the convection column absorbs approximately 190W/m^2 of thermalised solar EMR in the column to drive the convection; most being in the form of sensible heat transfer to cooler regions. It only takes 11W/m^2 to extinguish the Convective Available Potential Energy (CAPE). The 30C column over open oceans always experience mid-level convergence and high level divergence. They are as powerful as the ocean atmosphere heat engine can get. CAPE in these systems is limited to about 3300J/kg; corresponding to peak updraft of 81m/s.
The Inter-Tropical Convergence Zone is dominated by these powerful convective towers that pump up the tropical atmosphere to drive global scale air and ocean circulations. They are responsible for distributing heat from the tropical oceans to the rest of the globe.
The Sun and Climate Change
Earth’s orbit around the sun is constantly changing. It is possible to determine the solar EMR available at the top of the atmosphere using orbital parameters and somewhat complex geometry.
The minimum average over the year 2021 was 170W/m^2 and the maximum average just over 417W/m^2 occurred immediately south of the Equator. Note that locations where sea ice exists can have lower annual average solar input than 170W/m^2 because the ocean water does not absorb sunlight when sea ice is present. This means the solar input to sea ice prone locations gets as low as 90W/m^2 over annual average.
The changes over time are relatively small and the best way of observing these changes is to consider the difference over a period. The following image indicates the change from 1982 to 2020 with 2020 having generally higher solar intensity.
If the sun was the sole driver of the ocean surface temperature then it would be expected to see reducing temperature in the high southern latitudes and increasing temperature elsewhere. Accordingly the NCEP Reynolds Optimally Interpolated SST data was used as basis for the following three charts.
So the Southern Ocean has cooled over the period as expected.
The mid-latitudes in the Northern Hemisphere have warmed as expected.
The equatorial zone, specifically the Nino34 region, has a slight cooling trend and does not respond to the increase in solar EMR intensity. This highlights the immutable limit of 30C on open ocean temperature.
It is also interesting to observe Jupiter’s beat (11.8 years) in this temperature record. Jupiter is so massive that it moves the sun slightly relative to other planets and that alters Earth’s orbit relative to the sun.
The performance of all climate models in the Nino34 region is poor. They all show warming trends in the region between 1980 and 2020 so do not even get the trend in the right direction. The following chart shows the output of the CSIRO Climate Model using the CMIP3 input data of early 2000s vintage.
Besides having the trend wrong, it also forecasts the physically impossible of the open ocean surface temperature sustaining 30C for months on this chart and permanently by 2100 in the model output. Climate models bear no relationship to reliable measured data.
The following chart shows the USA GISS (CMIP3) model forecast.
It also has an upward trend and forecast an average of 28.2C in 2022 where the observed average has been 27C.
The European MIROC (CMIP3) model also has a warming trend.
The average for 2022 is 25C, some 2C below the measured.
All the models have similar upward trends where there has been no trend while the predictions vary over a wide range, well outside any recorded temperature. The following chart compares the CMIP3 forecasts for the period 2020 to 2100.
The models all have an upward trend where none has yet been observed but the averages in 2020 vary from 24.5C for the MIROC model to 28.2C for the GISS model; almost 4C difference. The CSIRO is predicting sustained temperatures above 30C in this region but that cannot happen.
Simple Questions the Need Answers
Reliable temperature records show the global oceans are not undergoing universal warming. The ocean water in the mid northern latitudes is warming. The Equatorial oceans are showing no cooling or warning trend. The Southern Ocean is showing a sustained cooling trend. How can CO2 be selective in how it warms, cools or neither cools nor warms different locations on the globe?
Climate models are based on constant solar input. However the maximum solar intensity peaks when Earth is closest to the sun, currently in early January – 4th January in 2022. The timing of perihelion moves later, on average, by 13 days per thousand years. In 2104, perihelion will occur on the 7th of January for the first time this cycle. The distance from the sun at perihelion and aphelion also changes each year. Earth’s distance from the sun in 2020 averaged 18,000km less than it will in 2027. These small orbital changes influence the intensity of solar EMR. Why aren’t climate models using orbital changes to show both warming and cooling in response to the changing solar intensity?
Deep convection is a crucial atmospheric process that limits open ocean surface to 30C. How can climate models hope to reflect the real world when clouds are parameterised with no sensitivity to surface temperature?
The image below is from the USA National Storm Prevention website. This shows the atmospheric profile in a skew-T plot from radiosonde sounding over Fort Worth/Dallas in June 2022. It includes values for CAPE and LFC among a number of other parameters used for storm prediction. It is similar to the simplified plots displayed above in this paper but the isotherms in the simple plots are shown vertically rather than skewed.
The rapid reduction in humidity above the LFC is the result of water vapour condensing/solidifying.
The data used to determine top of atmosphere reflected EMR were downloaded from NASA’s Earth Observation web site.
The data for climate models and NCEP measured SST were downloaded from the Climate Explorer website.
The data for the moored buoys was downloaded from the NOAA PMEL website.
The orbital parameters for determining the changes in ToA solar EMR came from the Astropicxels site:
Planetary Ephemeris Data (astropixels.com)
The image showing the Skew-T plot was downloaded from the USA National Storm Prevention website
The single column model (used to produce the atmospheric temperature profiles)is in the form of an Excel file with macro and is available by contacting the author using contact detail found at this link.
Richard Willoughby is a retired electrical engineer having worked in the Australian mining and mineral processing industry for 30 years with roles in large scale operations, corporate R&D and mine development. A further ten years was spent in the global insurance industry as an engineering risk consultant where he developed an enduring interest in natural catastrophes and changing climate.
Failure to model clouds properly is a major failing of climate models.
Huge. Wrote about it in essay ‘Cloudy Clouds’ in ebook Blowing Smoke. And it cannot be fixed mathematically due to CFL constraint on numerically solved PDEs. So clouds are parameterized, which drags in the attribution problem. A major reason climate models run hot.
None of the key processes of the water cycle are understood in a rigorous mathematical way which would permit modelling.
Climate models do NOT model the water cycle from “basic physics” , it is all parameterisations which allow modellers to insert their own biases, political or otherwise to retain the results which they find the most “reasonable”. ie it is bias confirmation. They effectively rig the model to produce what they expected ( or wanted ) it to produce.
Then pretend this must be right because it was produce by a computer.
All that they seem to demonstrate has been GIGO
They are worse than useless unless they can link clouds to the surface conditions.
They are driving policy in the wrong direction. Humans need to recognise the importance of atmospheric water. Anything that removes tress from the land surface are bad for the local climate.
The problem here is the clouds don’t stay put either, so while they do have effects on surface conditions the effects are NEVER predictable: two clouds of the exact same coverage moving at different speeds can have literally the opposite effect on surface conditions even ignoring wind factors
If you’ve ever stood in a forest on an 70% overcast day with no wind and then stood the same on a 70% overcast day due to hurricane feeder bands moving in waves through the sunlight you’ll be able to experience the difference. Life responds to slow and fast intermittency in different ways.
The major failure of climate models is they are programmed by people who were hired because they expected a coming climate crisis. The models are programmed to predict the climate crisis they believe in. There is no attempt to provide an accurate climate forecast, and no attention paid to the Russian INM model that least overpredicts the rate of global warming.
The same scary computer game forecasts would exist if clouds were modeled perfectly. You are confusing real science, where accurate predictions are the goal, with political junk science, where scary predictions are the goal.
I challenge anyone here to provide evidence that the average model has a goal of accurate climate predictions, or that 40+ years of model revisions have made the average model more accurate.
Science + Politics = Politics
tho not as bad as homogenizing temperature over distance resulting in 95% of the input being anecdotal and wrong
Nice essay Rick. Sometimes I have been testy with you about your 30C limit, as an oversimplification of SST evaporation, cloud condensation, and Clausius-Clapeyron but your work here is so good you have my 10/10 vote.
Especially the Skew T references, most clisci’s don’t seem to understand that meteorologists know quite a lot about cloud formation, while neither meteorologists nor clisci’s understand that engineers know quite a lot about heat transfer and condensation.
A couple of typos I noticed:
“Simple questions that (not “the”) need answers”, and 3 sentences later “warming” not “warning”
Most people aren’t cognizant of how much energy does NOT reach the ground as clouds obscure the sun, other than feeling shaded from the sun. Here’s my location yesterday aft. Units are watts/sq.M., Latitude 54, sunlight reaching the surface being part of your “30 C max” rule….
In a practical, empirical sense everyone can observe this by standing in the open on a partially cloudy day and observe the change in radiation as clouds drift across the sun. If the wind velocity is slight the shaded periods feel distinctly cooler.
Photovoltaic banks provide very good data on how much sunlight is reaching the ground. They are quite sensitive to clouds.
Latitude 54? Are you sure that shouldn’t be 34 (South)?
John, “So” of 1361 times cos of zenith angle at lat 51 is about 1120 July 23, so the atmosphere was letting most of the SW through to the ground at the graph peak. Other recent clear days it has also maxed around 850. Much more variable and in a short time period (due to clouds) than one would infer from the usual Trenberth diagram.
Zenith angle calculator:
How were these data measured?
Ecowitt 2320 weather station is what this is from, with PV cell sensor. You need Kipp and Zonen, Hukseflux, or Apogee pyranometers and pyrgeometers to do it more accurately, and they are 10% accuracy too…
Thank you – I appreciate the score. As you point out it has a couple of typos that I will correct.
This is a much shorter version of a more detailed paper that goes into much more detail on convection.
A recognition that ocean temperature is set by atmospheric mass at any given level of insolation
Also that global convective overturning provides the means to keep the system stable.
Ties in well with the earlier work of myself and Philip Mulholland.
It does tie in. It is the precision of the 30C limit that I really wanted others to grasp and why.
In a more detailed analysis, I determined that the average power output of the convective engine to raise the atmosphere, driving the overturning, averages only 11W/m^2 at the 30C limit. However the 30C column transfers an order of magnitude higher power to the adjacent overturning air. So most of the power goes into sensible heat transfer from the 30C towers to cooler zones apart from over water near land where the land convective towers can be more powerful. The 30C zones actually absorb latent heat from cooler zones at the mid-level during overturning because they make up to twice as much precipitation as they evaporate and condense in the heat engine.
Another subtlety is that it takes almost 10mm of atmospheric water to increase the ocean surface from 29C to 30C.
I am making my single column model available and have provided a link to my contact details in the references.
The next step is to appreciate that convection and not radiation creates the surface temperature enhancement above that predicted from the S-B equation but that is a bit off topic at this stage.
I would think that the fact that any increase in surface temperature decreases volumetric mass thus maintaining a fairly equal specific heat per measured volume due to gravity being the regulating factor would be important to modeling.
Sea ice plays a significant role in heat retention. Some water surface exists where average ToA solar EMR is only 90W/m^2. When there is no sun in the high latitudes, the sea ice forms to prevent heat loss but then gets melted quite quickly once the sun returns. So water surface at 271K can exist for part of the year where average ToA solar EMR is only 90W/m^2. Sea ice covers a significant area of the oceans so is a large factor in keeping the globe warm.
So it is not just the atmospheric processes. But, I agree that deep convection is the sole reason Earth is not a snow ball.
The ability of the atmosphere to form a free convection zone below an LFC and a dehumidifying zone above an LFC create the basis of the heat engine that gives life to the atmosphere.
If the atmosphere did not partition then the atmospheric column would reach 100% RH; clouds would persist indefinitely and the ocean surface would never see sunlight so would just cool until it was all ice – this is the basis of persistent cloud above ocean surface cooler than 15C where there is no deep convection. . Deep convection is the reason we can exist on Earth.
A key understanding of this exercise is that the whole atmospheric column cannot be fully saturated once an LFC forms. The cloudburst draws moisture from low levels to ultimately reduce surface RH after the cloudburst but saturates above the LFC immediately after cloudburst. Moisture is counter-cyclic between the two zones. At a 30C warm pool, surface RH ranges between 80 and 90% while the high level RH ranges from about 40% to 100%.
Note that the surface temperature enhancement beneath a convectively overturning atmosphere would be present even if there were no water, ice or water vapour available.
The phase changes of water just make the stabilisation process easier which is why Earth is less windy than Mars despite having a much more dense atmosphere.
Mars has planetwide dust storms but there is nothing comparable on Earth.
I see that my comment at 8.12 pm has attracted a negative comment score. There must still be a number of natural climate change denialists lurking in the background here.
Deep convection cannot be avoided on a sphere illuminated from a nearby sun because surface temperature and density differentials in the horizontal plane are inevitable.
Only by eliminating all temperature and density differentials in the horizontal plane can one eliminate convection and achieve an isothermal atmosphere. That is impossible.
great paper Willis is all over this as well
That Jupiter response is a story on its own.
It is a huge assumption. The solar cycle is 11.4 years and is synchronized to the Niño 3.4 region graph. I fail to see why it should be Jupiter.
Thank you for the support.
I have looked at just the orbital influence on the the Nino34 response and it has a reasonable correlation but does not give high enough correlation to be the sole contributor to the variation. I am certain ENSO is an independent contributor but there could also be sunspot.
It was nice to see the 11.8 year cycle so clearly in the measured data. The NOAA/NCEP Reynolds OI temperature record at an outstanding data set for ocean surface temperature.
Who also has a very good understanding of meteorology!
I would not claim that. I understand the heat engine that drives global scale convection and grasp enough of the convergence and divergence in the vicinity of warm pools to determine the heat flows. I would not try to predict the the weather. I am gradually understanding climate change though.
The heat engine is the basis of meteorology in that it involves conduction and convection rather than radiation.
The models rely on radiation and so omit the thermal effects of that heat engine completely.
That is why they must be scrapped.
A fresh start with new models is required.
We can certainly agree on this.
Temperature is not the only impact on the biosphere. Even if the models were to get cloud correct and starting using conduction and convection they would still be ignoring the impacts on the biosphere as a whole from climate change. You simply can’t make informed judgements on actions required if you don’t understand the entire biosphere on a holistic basis. That was Freeman Dyson’s largest criticism of the so-called “climate” models. They are really not “climate” models at all, just temperature models (and poor ones at that).
not radiation so much as “conduction” like somehow the heat MUST go through every molecule in the column.
Outside of continental drift, climate will not change without alterations to the environment the climate resides within. Forests and grasslands have an amazing tendency to remain forests and grasslands in spite of politicians’ declarations. Throw in a thousand farmers and some city planning tho and you get things like Optima Lake.
This is a third way to show the same basic hypothesis about adaptive thermoregulation over oceans. The others were Lindzen’s 2000 ‘adaptive iris’ paper and WE’s many posts here on TStorm thermoregulation.
I like triangulation. When three folks use different underlying observations to reach basically the same conclusion, your confidence level in the basic result (not the details) increases enormously.
And, as posted over at Judith’s years ago on Lindzen’s adaptive iris, when that mechanism was added to a CMIP5? EU climate model it dropped the ECS by about 40% using just the medium (low, medium, high) effect scenario. It dropped it into near observational energy budget range with the ‘high effect’ scenario.
This effect is totally missing from all CMIP5 and CMIP6 models, except possibly the INM CM 4, CM4.8, and 5.1 via the indirect addition of more realistic ocean rainfall so more albedo enhancing rainy clouds and thus less water vapor positive feedback.
Rud – I am now making my single column model available and provide a link to an email address. The model is in Excel. It does contain a macro , which is not particularly efficient.
I cannot wait to study it in detail. I am an Excel pro. But what you already posted was sufficiently convincing to me, having studied this for now 11 years.
Good article. To help see the results of some of these concepts in near-real time, I encourage the reader to view the NOAA visualizations for the GOES-East and GOES-West geostationary satellites. These links are for Band 16, the “CO2” band of wavelengths centered at 13.3 microns. There are also selections for the other visible and infrared bands and the GeoColor and other composite image versions. You can select longer or shorter animations.
GOES-East band 16
GOES-West band 16
Watch from space. One readily appreciates the heat engine-driven motion and the resulting formation and dissipation of clouds, especially in the tropics. It is highly self-regulating, and therefore not conducive to the accumulation of heat energy in the oceans or land to harmful effect from what GHGs do.
WE will really like those links. Direct support for his tropical Tstorm thermoregulation hypothesis.
The links show the brightest cloud is over the tropical land masses over rain forests. I eliminated the land portions to highlight the variation over oceans.
Forgive my skepticism. Milankovitch forcing is very powerful but very slow. How much ocean cooling shold be expected from a change of 0.08 W/m2 at the top of the atmosphere (30% reflected), over 40 years? That is what, 0.002 W/m2 per year? I just don’t buy that the cooling trend is due to that despite your “as expected.”
Agreed. The cooling in the Southern Ocean has been in process for 500 years; the last time perihelion occurred before the austral summer solstice. It has 9,500 years to go. Likewise the warming in the Northern Hemisphere has been in process for 500 years. I just took a 40 year period because that is when reliable records have been available.
The word “expected” was that it should be cooling, which it is. The NCEP Reynolds data is very good because it is already showing these subtle changes. Climate models have the Southern Ocean warming.
There is not much thermal lag in the response of ocean surface temperature but any region with sea ice amplifies the change because the exposed area is changing. The sea ice is the reason that the temperature response to solar forcing flattens below 220W/m^2 and water can still exist with average solar EMR of just 90W/m^2.
“The atmosphere over oceans is dullest at 26C”. From the point of view of someone from a high vantage point looking at reflections from the sea’s surface it may look dull, but sitting in my kayak the atmosphere appears to be at its brightest.
Exactly. On average, ocean surface at 26C is going to experience the least cloud.
Nice description of how water vapor in its various forms regulate the oceans temperature. The climate models concentrate on CO2 which is a minor, minor part of the earth’s thermodynamic processes. Water vapor is 1000 times greater and soaks up a ton of heat comparted to CO2. Someday, the models will be trashed and we can start over again.
That day will likely require all the existing modelers to retire.
It is a serious situation that policy is being driven by a misunderstanding of atmospheric processes. Anything that removes available atmospheric water from land is going to adversely impact local climate. Removing trees to install wind turbines or solar panels is a climatic disaster in the making.
The Northern Hemisphere is just 500 years into a 9.500 year summer warming cycle. Places like Europe need every tree they have and more to moderate the climate impacts until the ice mountains build again.
I guess we should be pushing back the demand from the Green energy lobby then. They think feeding millions of tons of freshly cut trees into DRAX power stations in the UK. is a good idea? Having shipped them across the Atlantic from forests in Virginia is considered really sensible??
The UK government agencies think it is such a good idea they paid DRAX grant subsidies for doing so £980 million last year.
Any group of people that champions deforestation to power a developed nations energy needs may be beyond sensible/logical energy ideas.
As Mark Twain said, It is easier to con people than to persuade them they have been conned.
Using trees only makes good sense when they are being replaced as fast as they are being removed. Managed forestry is likely better than letting forests build litter that makes them prone to intense wild fires.
The old argument about the precautionary principle with CAGW is sorely misplaced if it results in reducing forests to replace with wind turbines and solar panels. It is the exact opposite of what is best for climate.
The precautionary principle applies the other way as well. Lets assume that orbital changes are going to increase land surface temperature during boreal summers for the next few thousand years no matter what happens with CO2. All the resources wasted on reducing CO2 are wasted for no beneficial outcome.
Please excuse my boundless ignorance of Climate science for just a moment (I’m an English teacher): But just out of curiosity, what do you think about Dr. Ferenc Miskolczi’s findings? Would you say that they are compatible with your own? Thanking you advance.
I am looking solely at the balance and the determinants of that. There are two temperature limits for ocean surface water. -1.8C and 30C. The lower one has a powerful mechanism for heat loss through the insulating effect of sea ice. The upper one has a similarly powerful and precise control on heat take up by reflective cloud persistence.
Sea ice is not considered in the greenhouse effect so Miskolczi misses this powerful component by only considering atmospheric processes. His treatment of the atmosphere may eventually get to deep convection but it would be an indirect approach.
Deep convection is the reason Earth does not become a snow ball. It is literally a heat engine that pumps up the atmosphere and distributes tropical heat intake to higher latitudes. It is the most important atmospheric process and essential for most life on Earth. It is a process worthy of significantly more study than radiative heat transfer. Radiative transfer drives the engine but the the operation of the engine is vital to the climate.
Thank you for taking the time to produce this extremely helpful, instructive and amazingly concise reply.
‘There are two temperature limits for ocean surface water. -1.8C and 30C.’
I get the -1.8C limit (freezing point of seawater), but had a couple of questions about the 30C limit. First, is this limit unique to our particular water world / heat engine, and second, does this limit vary over time? My reason for asking is that the Earth was much warmer in the past, say 65 mya, when continental drift allowed for greater equatorial / lesser polar ocean circulation than what occurs today.
Btw, very nice work – it’s always good to learn about water’s role in providing negative feedback within the climate system.
Frank -The 30C limit is due to the buoyancy of water in the atmosphere and the fact that it solidifies above 273K. The ability to form a Level of Free Convection is essential for most life on Earth. Earth would be a snowball if there was no partition into the free convection surface zone and the high altitude solidifying/condensing zone.
The temperature limit has slight dependence on atmospheric mass. At 1100mbar surface pressure the limit increases to 33C.
Thanks, Rick. I find it very intuitive that our heat engine pushes back against perturbations to the system, which is something the alarmists ignore in their consideration that CO2, a ‘non-condensing GHG’, is the ‘control knob’ of climate. On the other hand, I wouldn’t want to fall into a similar trap that says climate change can’t occur because of convective feedback. In fact it has changed a lot over geological time due to continental drift, so one might consider that your 30C limit was operating even under ‘hot house’ conditions, but the planet was warmer due to the much larger extent of warm water at that time. Hopefully, this makes sense.
“CAPE in these systems is limited to about 3300J/kg; corresponding to peak updraft of 81m/s”
81 m/s = ~180mph/~290km/hr. That’s a lot of updraft!
The really powerful storms over land can produce very large hailstones.
Above oceans around 81m/s may not occur over much of the rising column or even at all because there is air entrainment from cooler zones below the LFC.
I believe South Africa has recorded the largest hailstones;up to about 200mm in agglomerated form. That requires seriously high updraft. The peak updraft velocity can be guaged by the size of the hailstones produced.
Speaking of convective velocity, the entire water cycle, not just clouds makes it very much easier for convective overturning to maintain system stability.
Without it there would need to be far more variability in such velocities than we see on Earth.
The dry planet Mars has much greater variability in wind speeds to the extent that periodically they become strong enough to produce planet wide dust storms.
It is the changes in the rate of convective overturning that keep the system stable by altering the rate at which kinetic energy returns to the surface from atmospheric CAPE (convective available potential energy) for radiation to space.
That variability neutralises any thermal effect from radiative imbalances however caused.
Rick’s work provides substantial backup for such a proposition.
All planets move the sun relative to other planets All planets influence each other’s orbits. The only difference is in the amount.
Correlation is not causation. Unless you can find a mechanism by which Jupiter influences the sun then this correlation is nothing more than a coincidence.
Gravity can’t be it, because Jupiter’s gravity affects all portions of the sun equally.
Earth’s orbit around the sun is significantly altered by Jupiter. I can replicate some of the changes in the recorded Nino34 temperature just using the orbital mechanics as determined by others. I use the orbital data from the Astropixels site to calculate the solar EMR.
I am making the point that there is a definite beat to the Nino34 region temperature record that ties in with Earth’s orbital changes due to Jupiter’s orbital period..
The sun is not a point source, therefore one side of the sun is nearer Jupiter than the other side, in the amount of the Sun’s diameter. How does that not produce gravitational differences on the near side vs the far side?
The mass of protons, free neutrons, gas, molecular elements, dust and particulate in the solar system outweigh Jupiter by a factor of about 40x not including other planets. The solar wind itself, just protons, weights upwards of 8.5×10^22 kg alone (Jupiter being 1.9×10^27)
The sun-side impact of average solar wind on Earth is more than 15w per square meter of cross section, the specific heat of it is negligible but its all moving at around 360 KILOMETERS per second.
Since the sun is a fluid mass and the center region in which the only real effect Jupiter’s gravity could have is under mechanical compression it is unlikely that Jupiter actually has any functional influence over the sun’s existence compared to the inner 3 planets. Remember that a 737 at 30,000 feet has more gravitational effect on you than Jupiter.
We could go into linear granularity of “point source” forces but then I’d have to use the faulty mathematics that astrophysicist undergrads use to prove that everything they see is a black hole.
Nope. Jupiter causes the largest tides on the sun. Venus is close to Jupiter, Earth is 2.2 times less than either of them. Tidal acceleration on the sun is 2*(G m Rsun)/r^3. You also got it wrong that the center of the sun is the only region that Jupiter’s gravity could influence – the tidal acceleration, although only one part in 10^7 of the centrifugal force of the sun’s rotation, does (very slightly) influence the surface too. Fluids experience forces too – think of the tides that the moon and sun cause on earth’s oceans.
Since you got relatively simple things wrong, I have to question (without checking) all of what you wrote.
I would only note that it is not unusual to assume the forces act on the center of gravity when analyzing distant forces. Otherwise you have to assume distortions in the shape of the body. E.g. tidal forces acting on the ocean don’t act everywhere in the same amount. In reality the forces “smear” things and that is difficult to handle, at least for use in a simple explanation.
The “real effect” is not just from the center, that’s just an easy simplification that isn’t reality.
If you go to the high-altitude lakes around the world where the atmospheric pressure is lower, you would expect to have a lower “threshold maximum temperature” do any get to this state?
Lakes get warm fast but do they hit the wall so as to say.
It’s a possible validation of this effect.
The best validation of the process is the Persian Gulf but as an exception. Its surface reaches 35C in August. The dry air from the north prevents the formation of an LFC.
The surface humidity can be quite high but the relative humidity drops off with altitude and is low where the LFC would form if there was no dry air convergence
Proximity to land is a factor in the convective cycle. Tropical rain forests will tend to have dominant convective towers because the surface warms quicker than adjacent ocean and the base of the column is at higher elevation. For example, Amazon pulls mid level moisture from the tropical Atlantic and the land towers dominate the convection.
All of this is way over my head can someone provide a short concise explanation in non scientific terms to explain what he is saying?
Open ocean surface cannot get warmer than 30C. All climate models show open ocean surfaces warmer than 30C so are wrong.
The observed climate change is explained by the changes in Earth’s orbit.
Bob, this is a somewhat hard to follow explanation of cloud formation. Better labels on the graphs would help. Essentially, warm, humid air would rise and form clouds freely. However, the atmosphere forms layers and the less dense air sometimes is stopped at the layer of dry air above it, even though it is either the same density or lower density. This is a non-equilibrium process. However, like most processes, once enough warm, humid air builds up, it breaks through the dry air above it and starts to form a cloud. However, enough clouds form to cool the surface and reduce the supply of warm, humid air. So the entire system is self-regulating because as soon as it speeds up, it starves itself.
Noticeable by their absence from this thread are those (we all know who they are) who would have us believe that back radiation from that nasty CO2 will make the oceans boil. Well done Rick!
Bravo! Good point!
I appreciate the acknowledgement.
Nick is not here because he knows the truth already, he’s just not brave enough to give up his paycheck to voice it.
This seems a compelling account of the negative cloud feedbacks that stabilise global surface temperatures against forcing effects. Is there a good journal paper or preprint on the subject?
The prior work on the 30C limit has been quietly buried. Ramanathan is a noted author but would not like his work on the 30C to be dug out now.
I believe me and a few others commenting at the top of this thread acknowledge the role of deep convection in the thermal energy balance. I believe I am the first to actually quantify the process through the persistence of reflective cloud.
In the mechanical world, most systems that are familiar to me have been designed around only one feedback control. There can be other devices that turn the feedback on off at at some limits, but it is hard to comprehend something like a steam engine with two or more spinning regulators.
So I am regarding this work by Rick as a direct competitor to the validity of the GHG control knob. In my book, one cannot contemplate 2 competitors designed by Nature and my money is on CAPE and co.
If the mechanism did not exist we would not have our watery Earth. Geoff S
The steam engines I have seen at various farm shows have at least two regulators, one mechanical and one biological (the guy controlling the fuel input with a coal shovel!).
Earth’s oceans have a thermostat with upper and lower set limits. The lower limit is -1.8C where sea ice forms to limit heat loss. The upper limit is 30C where cloud persistence limits the heat input to the warm end of the heat engine pumping up the armosphere and driving the global air circulations.
The importance of deep convection cannot be overstated. Without it, Earth would become a snowball.
It is interesting to observe the transition from the depressing condensing/solidifying cloud when the surface is less than 15C to the lively, vibrant clouds that are observed when the surface is in the high teens and above. Then to the more ominous clouds when the temperature gets above 26C..
One of the best climate essay’s I have read for some while. Although it is above my area of expertise, you just know what you have read makes sense and everything fits neatly together, follows known atmospheric processes, and importantly fits the data. The climate system can work on a timescales that dwarf our individual human existence, but the clues are there. Congratulations on a wonderful article Richard
Thank you Kevin. It took some time to reduce what started out as a complicated technical paper on deep convection to a note that was easy for those with an interest to digest.
The simple takeaways are that Earth’s oceans exist within tight temperature limits. Climate models do not simulate the physics of the upper temperature limit so produce useless output. Earth’s orbital changes are driving the observed climate change when reliable temperature records are the basis of observation.
Thanks for the reply Richard. I very much hope this article gets the exposure it deserves and others come on board, so far it has been very well received by those who regularly post here, and rightly so. There is a deafening silence from those who visit here and believe there is a different reason for climate change. I hope they learn something, and they think carefully about what this may mean for climate science as a whole.
I agree the models produce useless output. But I wouldn’t go so far as to say that Earth’s orbital changes are the sole factor driving observed climate change.
I do appreciate your article. I haven’t digested it all yet but it *is* a new, fresh approach that is one more proof that the model outputs *are* garbage.
The big climate changes have been due to shifting of the land masses. The millennial scale are mostly driven by precession- glaciation follows the precession cycle and multiples of it over the past 2Myr. The shorter term changes are many and varied. ENSO appears linked to changes in salinity.
Biological processes are important to climate change. Trees are vital for stabilising the temperature range on land as they retain water and make it available to the atmosphere.
I agree there are many factors but the changes in solar intensity due to orbital changes are already observable in good OST records like NCEP/Reynolds.
I am concerned that attacking CO2 is wasting time and resources on the real climate change that will occur over the coming decades and centuries. The worst possible outcome for Europe would be to mow down forests and replace them with wind turbines and solar panels. Reducing CO2 is not going to prevent the summer surface temperature rising in the Northern Hemisphere and the warming cycle is only 500 years into the 9,500 year warming phase. Essentially just turned the corner.
You mention TOA but never tell us what you use for TOA. There are several heights used on this site. Yours?
Not mine – the level where the CERES reflected SWR is measured.
The distance of perihelion and aphelion from the Astropixel site – no correction for Earth’s dimensions as I am looking at changes and Earth’s dimensions are insignificant in that calculations.
The 30C limit is insensitive to the solar EMR providing there is enough to start the heat engine. The -1.8C is also insensitive to solar EMR providing it is enough to melt the sea ice at some point through the annual cycle. Not all the sea ice melts.
E-M radiation and flux are used a lot in this paper, without specifying how the numbers are determined or measured:
So 136x W/m2 (= total irradiance at the top of the atmosphere) divided by something gives what? Numbers from 0 to 420?
Irradiance is an instantaneous substance, averaging can’t describe what is actually happening.
Here is an example of what I’m referring to, sunlight at a single location in Colorado, notice the difference between direct, diffuse, etc. (this displays a single day, more info will be visible in the afternoon):
The sources are referenced.
SWR reflected is from CERES.
Solar EMR at ToA is calculated based on orbital data from Astropixel site. There is no adjustment for Earth’s distances in perihelion and aphelion for EArth’s radius and the solar constant is assumed constant – I chose 1364W/m^2. But I am interested in relative changes rather than absolute as far as ToA EMR is concerned. Once the surface it at 30C, any more is simply not thermalised..
The surface insolation is taken at the moored buoys.
Deep convection is a robust process. It has a very sharp cut-off in response to surface temperature. It works exactly like an engine governor where the fuel source gets regulated to maintain a steady speed. In this case the fuel is the SWR being thermalised at the surface and the steady speed is the 30C surface temperature.
There is a thermal lag of about 25 days between the surface temperature reaching the 30C and the associated cloud persistence that provides the regulation – It does overshoot a little. The difference in atmospheric water between 29C and 30C is 10mm. The change in surface SWR from 29C to 30C is minus 20W/m^2. So it takes a long time to pump up the extra 10mm as the surplus thermalised EMR is ever diminishing as it approaches the limit.
Convergence at mid levels and divergence above the LFC complicate the energy balance in the column. 30C warm pools are always net precipitation regions, getting more precipitation than they evaporate. So they are getting latent heat from adjacent zones but transferring very high sensible heat in the upper level divergence.
The formation of an LFC is not accidental. It is inevitable. The 30C limit is immutable under the present atmospheric mass. It is far from some delicate radiation balance.
Getting into accuracy of measurements and errors in measurements of EMR is a trivial pursuit in comparison to the enormous power of the 30C convective towers. and the formation of persistent cloud.
If you want to understand anything about the global energy balance you need to understand why, how and where an LFC forms. Or you could simply accept there is a 30C limit on open ocean temperature without understanding the physics.
Your graph shows ocean surface temperature as a function of “solar electro-magnetic radiation flux W/m2”, numerated from 80-420 W/m2. My point is that these numbers make no sense to anyone who understands solar radiation, yet the graph states that the ocean temperature is a function of this quantity, and in fact implies that it controls temperature.
“Solar EMR at ToA is calculated based on orbital data from Astropixel site.” — I assume this is the air mass zero total (direct beam) irradiance adjusted for the earth-sun distance, which is 1360 W/m2 ± 10-12% depending on the Julian day of the year.
1360 W/m2 >> 420 W/m2, so I still don’t understand where these numbers come from.
Solar irradiance in the atmosphere is highly variable over the course of day due to solar position and cloud cover, but ocean surface temperatures certainly are not. What happens at night to this graph?
All the averages are based on monthly time scale. So average for each month for the declination and ToA solar EMR for determining the ToA solar input at 1 degree resolution in latitude. The OST and reflected EMR are taken over each month at 1X1 degree resolution and then averaged over twelves months.
I am mostly considering the EMR at the top of the atmosphere. My main interest is what is NOT thermalised – the energy rejected by cloud formation. The cloud formation above 273K is providing the reflection. Its persistence is the key feature of the temperature regulation.
The SWR that makes it to the surface to drive the engine does not care the path it takes to the surface. It is radiation heat input that goes solely into evaporating water to work the engine. There is little or no net radiation flux contributing to sensible heat in the water when it is at constant 30C. The only heat flux is cooling from the rainwater and that is offset by cooling of the mixed layer. The measured surface SWR happened to be at 200W/m^2 for the conditions I looked at the moored buoy at 0N 156E. The full heat balance for the 30C column over a 17 day period is attached.
Note that a significant portion of the heat goes into heat transfer through high altitude divergence. – I determined 190W/m^2.
OK, this makes more sense now, if there was an explanation that these are one-month averages, I missed it. From the diagram here, this is hemispherical irradiance (also called global horizontal irradiance), which is the direct beam irradiance plus the diffuse sky irradiance.
I have never seen the irradiance on a horizontal surface at 100km calculated as a function of time and latitude, likely tedious but otherwise straightforward. At the surface it is a much different story as aerosols and cloud cover have tremendous effects on the instantaneous irradiance.
At this point I’m wondering if the x-axis on your first plot could instead be translated to latitude. (Insolation is such a generic word that it is almost meaningless).
Very good comment and illustrious of why average insolation on a “flat” earth is only good for “back of the envelope’ stuff.
Thanks, Jim. I hope it is only tangentially related to his main thesis that ocean temperatures are pinned by water vapor processes. But whatever the processes are, they stop or reverse once per day when the sun goes down. Obviously there are thermal intertias involved, but does an average really capture what happens?
It also bothers me is these energy balance calculations always seen to use irradiance, i.e. power, but something tells me they should instead be done with time-integrals to get energy, i.e. Joules.
This is not the case for open ocean. The surface temperature is more affected by rainfall than the sunlight.
The cool end is always losing heat to space.
The rainfall occurs any time of the day, indicating that the cloudburst is not dependent on sunlight.
This is not what I meant—at night there is no energy being added from sunlight.
So you meant something other than what you wrote. Clearly the process of deep convection continues. But it definitely needs regular fuel input to the surface to keep cycling.
Planck, Stephan-Boltzmann, etc. are only applicable under very specific conditions. One is “equilibrium”, in other words, at a single point in time. You need to integrate those over time to get the total energy. But that also means you need extra variables that describe the gradients involved.
It is tedious but that is what computers do well once you work out the geometry. I did state “solar EMR available at the top of the atmosphere using orbital parameters and somewhat complex geometry”
This post is a precis of a much more detailed analysis. That is why it is described as a technical note.
My aim is to reduce the climate drivers to their simplest form for clear understanding. Deep convection is the most important feature of Earth’s atmosphere. Without that, Earth would be a snowball. Deep convection is not featured in climate models. It is reasonably understood for storm prediction mostly over land but there is limited work on its operation over open oceans.
The key insight, that is not covered in the note, is that deep convection prevents the atmosphere reaching 100% humidity throughout the column. If that happened then there would never be clear sky so the whole atmosphere would deflate as the water solidified. That condition occurs once the surface is below 15C but advection of both air and water still produces clear sky conditions. The convective engine in the tropics drive the whole heat transfer process powering advection.
The attached charts complete the picture for the moored buoy.
The atmospheric cloud response is about 25 days behind the surface temperature at this location in the Pacific. That means there is overshoot in temperature and a a period before it settles down.
The delay is really noticeable in the Arabian Sea and Bay of Bengal. The temperature gets above 30C almost a month before the monsoon sets in.
The delay is the cloud response at altitude is the reason the UAH temperature lags the surface measurement. The lag is a function of the surface temperature and the size of the regulating pool. The lag is shorter mid Atlantic.
Way over my head (I’m also old ) but also relatively clear enough ( dumbed down? ) for those of us struggling to convince the general population that the planet isn’t burning up /flooding any time soon. If only we could get the politicos to give it a read.
Glad to see this discussion on the part played by water, particularly via the Hydrological Cycle, in controlling the global temperature.
IMO there is a severe dearth of good information available on the internet leading me to believe that there some form of “Platforming” going on here; mainly as the science seriously challenges the IPCC claim that water provides a POSITIVE feedback to the GHE.
I look forward to the subsequent parts being published.
I already have some interesting questions to raise; but will leave those ‘til later; these mainly to tie in with my own work on the subject.
I have provided a link to my email address in the references and I am making my single column model available. It is in Excel and not complicated but has a macro that takes about 40 seconds to execute.
Well done, Richard, and very helpful. Your posts always make sense and are clarifying.
Thank you Andy. I really appreciate your comment.
No, OST is not a function of SEMR at ToA. It’s a function of all energy inputs, at the surface. In the WPWP for instance, <200 W/m^2 solar and >400 LW of which >100 is release of latent heat transported from higher latitudes.
The leading chart shows the OST response has clear break points in response to the solar EMR. The rest explains why.
Once OST reaches 30C, the solar EMR being thermalsed is regulated to keep the engine running at that level – no more, no less. The heat engine does real work averaging 11W/m^2 in elevating the atmospheric mass. The 30C column absorbs latent heat from converging moist air below the LFC and distributes sensible heat to the cooler surrounding air through upper level divergence.
Surface solar heat flux is 200W/m^2, all going to evaporation to power the engine and rejected through mostly water solidifaction at low temperature end high in the atmosphere.
“Once OST reaches 30C, the solar EMR being thermalsed is regulated…”
That would make more sense, OST limiting the available solar in the warm pools, like Willis’ fig. 3 here, https://shortfall.blog/how-clouds-think-about-climate-change-d6f4728b174c. Available solar a function of OST, above 27 C.
Wonder why you find a break point at 30 C while Willis finds one at 27 C.
There is a break point is at 26C. That is where the reflected EMR bottoms. It is clearly stated in the post. It is the darkest region in the image of reflected EMR (muddy yellow).
The 30C is a hard limit for open ocean surface temperature. That is where the engine runs out of steam because the heater gets turned off by reflective cloud. It is the point of energy balance between thermalised EMR, latent heat input from converging mid level air, OLR rejected at the top of the column and the sensible heat transfer to the adjacent air through high level divergence. See diagram that shows the conditions for a moored buoy at 30C surface temperature.
By ‘break point’ I mean a point where the 1. derivative changes sign, but it’s more interesting to discuss your diagram.
For 30 C to be a hard limit you need to assume the latent heat in of 73 W/m^2 is max achievable. Based on what?
You also assume no advection into top 25 m water, which is a faulty assumtion. There is a lot of heat transported into the warm pool.
oops! Retract last paragraph.
The latent heat into the column varies for different warm pools and different times. The 73W/m^2 in the case I looked at is based on the extra precipitation over what the column above the buoy was condensing/solidifying. I can determine the rate of solidifying/condensing based on the OLR.
The only unknown is the net sensible heat transfer in the advection. so it is the remainder. In this case it averaged 190W/m^2 for the moored buoy. The average vertical velocity to achieve that rate of transfer to convergence at the base and divergence above the LFC is very small.
The 30C regulation simply depends on the LFC approaching the the 273K altitude. That occurs at 32C but would prevent direct solar reaching the surface and there would be insufficient input to keep the convection engine going.
The top of atmosphere is no refuge for CO2 warming IR pinball. At the TOA CO2 repels incident solar IR, reducing total solar irradiation (TSI).
this is a great paper
we need skeptic GCMs with these kinds of feedbacks… it’s too bad we can’t just fork a clone in an IPCC model, plop in some cloud feedbacks and publish some low ECS hind/forecasts here on WUWT
suspect it’s indecipherable unreleased spaghetti anyway but you never know
Great to see that the work is appreciated.
I disagree on this. The world needs useful models to drive decisions. Does not matter who produces them as long as they reflect real physics of the atmosphere that produces results consistent with observation.
The current models are driving decisions that are detrimental to local climate and wasting precious resources.
I think this supports what Willis has been saying for years – it is a self-regulating system.
Awesome article. The sea surface temperature actual versus model predictions graph is worth 1000000 words. What an indictment. Can someone do this for CMIP4, 5, 6?
Have I missed the visit by a special person telling why this post by Richard Willoughby is wrong.
The dog hasn’t barked. Wow.