Theory of Earth’s Thermostat. It is the poles. In winter.

Some time ago I noticed that temperature varied in the arctic regions far more than in the equatorial regions.  This lead to my theory that the poles act like a thermostat regulating the earth’s temperature.  The post just prior to this one looked at the absorption spectrum of the atmosphere versus the emission spectrum of the earth and atmosphere to show that CO2 is largely irrelevant as it is either overwhelmed by water vapour or in a temperature zone where its effects are largely reduced.  In examining two articles on Watts Up With That by Willis Eschenbach, I realized that his theories and evidence, meshed with mine, provide for an enhanced theory for which there is considerable evidence.  The four relevant articles can be seen at

None of these completely explains why the earth’s temperature has been so stable long term, but each is part of the puzzle, and taken in combination, they result in an even stronger theory that is demonstrable using even limited current data.

My original theory was that the earth’s convection systems cause heat to be pumped from the equatorial regions to the poles.  As the earth heats up, the convection system accelerates causing a disproportionate temperature change at the poles.  The amount of energy radiated by the poles goes up exponentially, limiting how high the temperature can rise.  Willis’s theory starts with the same concept, and he shows the following picture to describe what happens:

Earth's heat pump. Taking it to the poles.

Willis’ article also quotes the research of paper of Bejan

 which studied this effect and concluded that the amount of energy being moved is in fact proportional to the temperature differential between equatorial and arctic regions.  At first glance, this would suggest a negative impact on the regulation system I proposed.  The heating up of the poles would diminish the very heat pump I was proposing to limit temperature rise.  Further analysis also showed that while the average arctic temperatures responded almost in concert with equatorial fluctuations, the Antarctic was far more stable.  To explain these things, we need to combine the concept of the poles as temperature regulators with the same concepts we looked at regarding absorption and emission spectrum.  The final thing we have to do is put aside the notion that the poles limit temperature increases when they are warm (the summer).  They actually do the bulk of their work in the winter and this can be demonstrated.

To demonstrate this we first have to demonstrate that the atmospheric window, that portion of the emission spectrum where greenhouse gases don’t operate, exists.  We can do this by combining a couple of pieces of data that are easily verified.  The first one is to understand how much solar radiance (insolation) the arctic regions are exposed to, and more importantly, when.  In his second article, Willis posts the following graphic:

Insolation by latitude. Arctic zones are 0 for many days each year.

The vertical scale shows average daily insolation in watts per meter squared at various latitudes, the horizontal scale is the course of the year.  In summer, the arctic zones get huge amounts of insolation daily because the sun shines right through the night.  In winter, they get long periods of 24 hour darkness.  It is the dark periods that allow us to demonstrate the existence of the atmospheric window which dramatically limits how cold the arctic zones can get.  The following is from the Centre for Ocean and Ice in Denmark and shows the average temperatures within 80 degrees north latitude daily over the last forty years:

Average daily temperature at 80 N since 1958

Now let’s consider these two graphs combined.  If we look at the first graph, we can see that the arctic zone within 80 degrees north latitude is exposed to zero or near zero sunlight for about 140 days per year.  This would be 70 days on either side of the winter solstice, and would correspond roughly to day 300 on the temperature graph on the right hand side going through to the end of the year, and then starting the following year comprised of the first 75 days or so at the left hand side of the graph.  If we look closely, we can see the bottom end of the atmospheric window emerging.


From day 200 to 300, the temperature in the zone drops by 25 degrees.  We would expect the temperature to drop because the amount of insolation is dropping to zero over that time.  We would expect that when it hits zero, the temperature decline would accelerate.  Instead, what we see is that the temperature decline actually decelerates.  In fact, once the temperature hits 245 K (-28 degrees C)at about day 300,  it flattens out completely and stays steady between 245 K and 243 K for the next 90 days even though it is getting zero energy from the Sun during this time.  In 100 days, with the Sun shining, it dropped 25 degrees.  In 90 days with no sunshine at all, it dropped 2.  Since there is no sunshine, something else must be preventing further cooling.

We’ve already eliminated insolation.  Ocean current can similarly be eliminated as we are talking about the depths of winter when the arctic is covered with ice and snow.  Any warm currents under the ice would be very muted in terms of driving surface temperature flcutuations.  The only two possibilities are re-radiance from cloud cover and heat being moved by atmospheric processes.  Cloud cover being intermittent, the driving factor is more likely the atmospheric window closing.  Earth surface is allowed to get so cold… and no colder.  While fluctuations in any given year certainly go lower than that temporarily, rarely does it get colder than 40 degrees below zero, anywhere.  In fact  winter cities like Winnipeg at 50 degrees north latitude experience far fewer days at -40 than does the arctic, but the extreme lows are very nearly the same, just around -40.  The atmospheric window just won’t allow earth surface temperature to decline much more because it intercepts upward longwave and prevents it from escaping into space, even though earth surface is trying to shed about 200 watts/m2 via longwave radiance.

If we accept that this demonstrates the existence of the atmospheric window (at least at the low end) we can take the next step and show how that limits any rise in the Earth’s temperature over all.  Consider the average temperature in the arctic, but this time let’s take a look at a single year.  I’ve chosen 2006 because it has some very interesting temperature fluctuations that serve to make the point.  Consider the first 75 days of the year where there is a very unusual temperature spike upward:


For about 30 days the surface temperatures ranged from 10 to 15 degrees higher than normal.  That is huge.  Having eliminated insolation and ocean currents, the possible sources are changes in cloud cover and large atmospheric disturbances.  Cloud cover seems unlikely as it would have had to stay stable for weeks, and both the rise in temperature at the beginning and the fall in temperature at the end are very sudden while cloud cover would more likely show up as a gentle rise and fall.  That aside, consider the effects on surface radiance to space during that time.  On a strict radiance to space as proposed in my original theory, this would boost the output of the arctic from about 200 watts per square meter to 250.  That’s a big jump.  But the real jump is much larger.  Since we’ve already demonstrated the existence of the atmospheric window that closes, preventing any significant temperature drop below 243 K, we can also assume that the window is wide open at 253 to 258 K.  In brief, that atmospheric disturbance resulted in an increased radiance to space of not 50 watts per meter squared, but 50 watts/m2 PLUS what the atmospheric window now lets escape.  Another 50?  100?  200?.  Once the disturbance plays itself out and stops feeding heat into the arctic, the temperature falls like a stone back to about 245 K.

While the area within 80 degrees north latitude is small in comparison to the planet, it was used only to analyse the data available and validate the theory.  In practice, this effect would occur anywhere that temperatures drop to the 245 K level which includes large parts of the temperate zone in lower latitudes, and in the opposite season, the Antarctic and south temperate zones.  With variations in radiance in the range of hundreds of watts per square meter inside of the atmospheric window, there is more than enough surface area to compensate for massive increases in greenhouse gas forcing with only a small change in temperature.

In fact, this brings a number of pieces of the puzzle together.  For starters, it makes it clear that the rate at which heat is transferred to the poles is in fact governed by the temperature of the planet on average, but not as much by the temperature of the poles.  During the time period when the arctic and north temperate zones are below the atmospheric window, they hover within a very narrow temperature range of plus or minus only a degree or so.   That being the case, the amount of heat being transferred from the equator to the polls, which is proportional to the temperature difference between them, is proportional primarily to the equatorial temperature during the same time period. 

Going back to the 80 degree north latitude example, that area would be in darkness for 140 days, but stable in temperature with the exception of heat transferred by atmospheric processes.  These in turn would be governed by the average temperature differential between equator and poles.  The poles being stable for that period of time, the temperature differential is dominated by the equatorial temperature.  A small rise in equatorial temperature would pump heat to the poles.  Based on Bejan et al, a 2 degree rise in temperature at the equator would cause an increase in heat being pumped to the pole of 3.6% (using 300 K +2 at the equator and 244 stable at the pole), a massive change in the amount of heat being pumped versus a very small change in temperature.  Could the poles handle it?

If all we were talking about was the poles increasing radiance according to their temperature, this would be fairly dramatic.  If the putative 3.7 watts/m2 from CO2 doubling were all gathered from the entire earth surface and forced out to space at one of the poles, how much would the polar region have to heat up to balance it out?  Since we can include the arctic and some pretty big chunks of the north temperate zone, let’s call it 15% of the earth surface.  The area would have to pump an extra 25 watts/m2 out to space to maintain equilibrium, which would translate to a temperature increase of (again using 244 as a starting point) about 16 degrees.  But the 25 watts/m2 would be accomplished at a much lower temperature increase, because the atmospheric window would go from “closed tight” to partly open with only a very small temperature change.  How much additional radiance would there be from the window opening?  100 watts?  50?  Experimentation would be required to arrive at a decent estimate, but given that the atmospheric window can clearly contain about 200 watts/m2 of radiance from the arctic region at 244 K, it has enough available to cancel the global energy being retained from CO2  forcing with only a small change in temperature.

The southern hemisphere doesn’t always oscillate in tandem with temperature changed in the northern hemisphere.  It plays to its own beat, and this theory explains why.  The Antarctic is colder than the arctic for a few reasons.  The two biggest ones are the earth’s orbit, and land mass.  The earth’s orbit is elliptical which works with summer warming in the north, but against it in the south.  The Antarctic also sits atop a land mass in the middle of an ocean, while the arctic sits on top of an ocean in the middle of several large land masses.

The consequence of this is that the lows during the Antarctic sunless period are a few degrees colder than the arctic.  When the seasons are reversed, and it is the antarctic’s turn to regulate temperature, a lot more heat is required to drive it up to the point where the atmospheric window begins to open.  It gets little help from the immediate area because it is mostly ocean, which resists temperature fluctuations while land masses amplify them.   As a consequence, the planet as a whole must heat up considerably more before the Antarctic starts playing a major role in regulating temperature.    Our expectation would then be that in a warming cycle, the northern hemisphere warms less before the arctic draws the line and says no more.  The Antarctic would have to continue warming for what ever period those extra few degrees would requires before the atmospheric window opens… and then no more.  We’ve seen this in the actual temperature record where warming “stalls” in the northern hemisphere, but continues in the southern.  For a while.

This entry was posted in Uncategorized. Bookmark the permalink.

One Response to Theory of Earth’s Thermostat. It is the poles. In winter.

  1. Gene Zeien says:

    The only two possibilities are re-radiance from cloud cover and heat being moved by atmospheric processes.

    Another heat source, especially at -40C would be IR radiation from the stratosphere & mesosphere. Although, given the chronic tropospheric downdraft around the north pole, I would favor atmospheric processes as a stronger mechanism.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s