The Indo-Pacific Warm Pool with Southern Australia

Most of the tropical oceans had small, ~2C variation of the past 200,000 years.  The majority of change in the glacial periods was near the poles.  I have scaled 65N estimated solar variation with the EPICA CO2 for this chart just for reference.  I haven't verified how accurate the estimate is so far, or how well I repaired the spread sheet borrowed from the WUWT post Milankovic Cycles Defended.  The interesting thing in this chart is the Mohtadi et al reconstruction has a longer positive slope in the temperatures and sea level (in light blue) than many reconstructions.  CO2 change clearly lags the Calvo southern Australian SST reconstruction and there is a blip at ~19 k years in the Mohtadi SST data.

The Mohtadi 30 ka slope change aligns with the Pahnke and Sikes Southern Ocean shift with the 19ka also roughly aligning with the exit of those shifts.  Those shifts are likely due to Antarctic Circumpolar Current changes which is what I have been looking for.  The 30ka period does have solar as a possible major factor where we would have reduced SH solar timed with reduced SH temperature possibly cause larger impact.  Note in this chart is the dark blue Greenland ice accumulation which is likely an indication of whether the Arctic was iced in or more open.

This is just a memo post to keep trace of things.

This is kinda messy, but I added the Tierney '10 Lake Tanganyika back to 60 k years BP.  This shows a little better the tropical range of temperatures which are extremely stable.

Unforced Variations in Climate - Not Something to Dismiss

When is a Trend a Trend? When it is too late to do any good.  Nonlinear systems have their internal cycles and wanderings.  The wandering would have been "forced" since there is energy in these system they did not create.  When and how they use or lose energy totally depends on the dynamics of the systems.

The Lake Tanganyika surface temperature reconstruction by Tieney et al. has three interesting similar pseudo cycles.  The first, starting 16,500 years ago has two less data points than the third starting ~7000 years ago, and the data is coarse, but let's see how they compare.

I screwed my numbering up a little, Cycle 2 above actually is the first cycle starting around 16,000 years ago.  I picked the peak value as the start date and have all three on the same time scale.  The average is there just for reference, complex cycles would never have a reliable average, but a range that may be useful.

All three of the pseudo cycles in the reconstruction have a rough period 4300 hundred years with irregular dampened oscillations with periods of roughly 1135 to 1350 years.  The Bond cycles or events are irregular North Atlantic climate cycles of 1470+/- 500 years associated possibly with Dansgaard–Oeschger events or by my estimation, just decay curves for ocean perturbations.  Note that near in the 600 to 1250 year end of the ~4300 year cycles above, there is an average warming of approximately 0.75C degrees.  In the blue Cycle one, the increase was about 1.0 C degrees from ~1200 to ~1000 years in the pseudo cycle or about 0.5C per century due to apparently Unforced Variations.  Nasty term that, Unforced Variations.

Lake Tangayika is not in the North Atlantic, it is just south of the Equator on the Eastern Side of Africa nearer to the India Ocean.  Imagine that, the Indian Ocean appears to have irregular Unforced Variability.  

So then next time someone mentions that Unforced Variations zero out and therefore do not impact climate, laugh your butt off.  

When is a Trend and Trend?

Answer:  When it is too late to matter.

Updated Stuff Below:

I can go on about the need to consider baselines until I am blue in the face and it won't matter, people will "see" what they wish to "see".  The average AGW fan will "see" the increased slope in the data following an unexpetced downturn and "see" a signature of Antropogenic Global Warming.  

With a little curiosity, the AWG fan might look at some other data, like say,

this plot that has a double dip ENSO oscillation, followed by an increasing warm ENSO trend culminating in a OMG what happened! event.  Obviously, the Clinton Administration was asleep at the switch because everything boils down to politics.

The 1S 91W is west of South America near the Galapagos Islands.  Tmin from the Best data set was used since being measured on islands, the minimum temperature would be pretty close to the sea surface temperature.  Near the Galapagos Islands, Tmin has bee increasing with a long term trend of about 0.02C per year, 0.2C per decade, 2.0C per century for about a century.  Probably due to the heavy duty Ecuadorian industrial complex.  Yes, that was a joke.

Then again, it could be because the Tropical Eastern Pacific wanders up and down a few degrees all the time, on occasion with a little spike above or below that range.  Generally, those spikes means something is changing.  About 125 thousand years ago, there was a little off the chart spike.  Might be time for another spike.  1998 might have been that spike.  So what happened after the spike about 125 thousand years ago? Not much.

Even though the time scale of this paleo reconstruction by Hebert et al. is a touch longer than decadal, the settling or decay curve following the 1998 spike will likely be somewhat similar, but never exactly the same.

So a trend is not all that helpful, but a pattern, even though not perfectly repeating, can be useful.

The new stuff:

Someone just could not believe their eyes when I showed them the Galapagos Tmin data stating that it makes a PDG ENSO tracker.  Imagine that?

So here is the BEST Tmin for the Galapagos Islands area with McGregor et al. Unified ENSO Proxy Reconstruction.  Just to make sure:


WDC PALEO CONTRIBUTION SERIES CITATION: 
McGregor, S., et al. 2010. 
350 Year Unified ENSO Proxy Reconstruction. 
IGBP PAGES/World Data Center for Paleoclimatology 
Data Contribution Series # 2010-015. 
NOAA/NCDC Paleoclimatology Program, Boulder CO, USA.
www.clim-past.net/6/1/2010/

Now I did not detrend the BEST Galapagos Island Tmin since the trend or lack there of was the point.  ENSO may be considered a neutral oscillation but what oscillates ENSO doesn't have to be trendless.  The +/- 1C variation is "normal" and the excursions beyond +/- 1C would be the abnormal events likely associated with shifts some nature around that general time.  Had I detrended, the "fit" would have been more impressive. 

Battle of the Hemispheres

This comparison of the Hadley Centre northern and southern hemisphere shows the drifting pseudo-cyclic variations of temperature in the two hemispheres.  The oscillations have apparently random shifts with different impacts on measured climate.

The general cause of the shift is thought to be ENSO related weather "regimes", Thermohaline Current variations, Solar forcing fluctuations.

The cycle of glacial periods is also pseudo-cyclic, changing from roughly 41k year cycles to roughly 100k year cycles in the more recent glaciations.  The opening of the Drake Passage some 40 million years ago is likely the cause of the glacial periods with the closure of the the north and south American continents changing the thermohaline current in the world's oceans adding to the change in climate regimes.

The Drake Passage opened when south America separated for the Antarctic starting the Antarctic Circumpolar Current (CPC) which thermally isolates the Antarctic climate from the global climate.  In the chart above, it appears that changes in the flow of the CPC over millions of years contributed to the shift from ~41k year to ~100k year glacial oscillations.  With the CPC increasing in climate influence, impact of Solar variation will likely be much less that historically unless coupled or synchronized the CPC fluctuations.  Antarctic sea ice formation and changes in the orientation of extent should be a major impact contributing to global climate change on all time scales.

By adding the 65N solar insolation estimate in the background you can get a feel for the different sensitivities of the two Sikes temperature reconstructions to solar forcing.  the mixing of the thermohaline at the source, the CPC, has complex dynamics with likely impacts thousands of year after the forcing.

 Data for Antarctic chart:


Pahnke, K., and J.P. Sachs. 2007.
Southern Ocean Midlatitude 160KYr Alkenone SST Reconstructions.
IGBP PAGES/World Data Center for Paleoclimatology
Data Contribution Series # 2007-019.
NOAA/NCDC Paleoclimatology Program, Boulder CO, USA.

Jouzel, J., et al.  2007.
EPICA Dome C Ice Core 800KYr Deuterium Data and Temperature Estimates.
IGBP PAGES/World Data Center for Paleoclimatology
Data Contribution Series # 2007-091.
NOAA/NCDC Paleoclimatology Program, Boulder CO, USA.

This is just a draft but some may find it interesting.

Update:  I am trying to use a last known value method to simplify averaging of various paleo reconstructions.  Instead of interpolating, the last know value is used to infill gaps where the gaps are not humongous.  

The exit from the last glacial maximum is a little different which is what I am trying to isolate.  The tropical oceans would be one reference with the tropical Eastern Pacific being a good indicator of the ENSO type oscilations.

Webster's Follies

The relationship between atmospheric CO2 and temperature is a touch complicated.  The cause and effect can easily be confused.  The chart above is CO2 from the EPICA Ice Core with Southern Ocean temperature reconstructions collected at two depths compared with the Tropical Eastern Pacific Bottom temperature of the ocean.  The Tropical Eastern Pacific (TEP)is an important region because the Antarctic circumpolar current (CPC) can vary at the Drake Passage between South America and the Antarctic causing huge changes in the deep oceans currents and mixing of the ocean heat at different depths.  The change in the TEP Bottom temperature is small compared to the surface temperatures.  In the chart, the TEPBT is multiplied by 4 and inverted to produce the fit of the curves.

The data here indicates that deep ocean mixing and the CPC are major factors controlling Southern Hemisphere CO2 Concentration.  Since CO2 is a fairly well mixed gas, then the CPC and deep ocean mixing would play a major role in regulating global CO2.

Webster, a denizen on Dr. Judith Curry's Climate Etc. blog has meticulously calculated the CO2 uptake of the oceans and the rate of heat diffusion into the global oceans without considering the mixing that occurs in the southern oceans.  That would make his work useless, but humorous :) 

Because it Gets Amplified

While I am playing with some other stuff I thought I would take a break to illustrate the main thing wrong with a "global" average surface temperature.  The chart above is the Herbert et al. Tropical Eastern Atlantic and Eastern Pacific SST which I have detrended  the section from 20k year to 110k years.  The is the downward slope into the last glacial maximum.  The planet was cooling, big time right?

You can see how the smaller Atlantic ocean amplifies changes in the large pacific ocean temperature changes.  Since the Atlantic provides most of the moist energy for Europe and Africa, land surface temperatures would respond to a fraction of the total heat capacity of the "globe".  The average temperature of the "global" oceans only changed about 2 degrees in this period and Atlantic had a frightening "Global" warming event right dead in the middle of the decline into a major glacial maximum.

The reason I am making this quick post is there are idiots that don't realize that data also will lie to you.  You have to be smarter than what you are messing with or it gets really embarrassing.  Start at the beginning, don't assume anything you can avoid and use frames of reference to verify your work.

  

 Since I have it, here is the same chart with the estimated solar.  By detrending I am just hoping to make it a little easier to figure out the how much solar impact where and when.  The thermal inertia of the oceans matter, both with respect to each other and with respect to "normal" for that state.  As you can see there appears to be a large response to solar around 85K then a smaller response with greater lag around 60K.  The ultimate would be to find a "reference" which has more uniform response to solar and a stronger impact on "global" conditions.  Then use a similar "network" analysis on the available data in the Tsonis manner.  It is not rocket science.

The Southern Oceans

Most of my climate curiosity is focused on the Southern Ocean and North Atlantic Ocean relationships with climate.  So I am playing with some reconstructions.

Pahnke, K., and J.P. Sachs. 2007.
Southern Ocean Midlatitude 160KYr Alkenone SST Reconstructions.
IGBP PAGES/World Data Center for Paleoclimatology
Data Contribution Series # 2007-019.
NOAA/NCDC Paleoclimatology Program, Boulder CO, USA.

ABSTRACT:
Late Quaternary climate fluctuated between extremes of glaciations, lasting ~90 kyr on average, and interglacial episodes, lasting ~10 kyr. Still largely unknown are the spatial and temporal evolution of these global climate states, with vigorous debate still underway on the mechanisms responsible for glacial inceptions and terminations. Though it is widely believed that the Southern Hemisphere oceans play a central role in global climate changes, few paleoclimate records exist from there. Here we present three new alkenone-derived SST records from the midlatitude Southern Hemisphere spanning the last 160 kyr, a full glacial-interglacial cycle. Our subtropical SST records from the last glacial period are characterized by (1) warming 47–23 kyr B.P., when high latitudes in both hemispheres cooled, and (2) milder temperatures during the penultimate glacial period than during the last glacial interval. These SST features are found to be of tropical- to subtropical-wide extent, implying increased thermal gradients at times of high-latitude ice sheet growth. This has implications for the vigor of atmospheric and upper ocean circulation and the transport of heat and moisture to the poles that may have been instrumental in the growth and maintenance of polar ice sheets during glacial periods.

I just included the two newest Sikes SST records.

Here I added the Tierney Lake Tanganyika Lake Surface Temperature 60K year reconstruction that I have used before.  There is a good correlation with the southern ocean reconstructions.

Here is the Martin et al. Tropical Eastern Pacific Bottom Temperature with EPICA CO2 showing the inverted relationship.  The Super La Nina aspect of climate or the modulation of the Antarctic Circumpolar Current (CPC) is pretty hard to discount. 

 For the Cyclomaniacs and the "Sun Done It" crowd.  The internal oscillations of the oceans, when complimentary to an external forcing, have the expected impact.  However, when external forcings are out of phase with internal dynamics, more subtle relationships prevail.

Back to the Simple Model

The Eastern Tropical Atlantic and Pacific bottom water temperatures for the Quaternary period are plotted above with my less than stellar but adequate for now estimate of average tropical ocean sea surface temperature using the Herbert et al. tropical Atlantic, Eastern Pacific, Arabian Sea and South China Sea reconstructions.  I only use the first 200 k years of the tropical SST average since that is enough for now.

The reason I wanted a tropical SST average was to compare with the tropical bottom water temperatures.  My little model indicated that the average range of SST was limited to roughly +/- 2 C degrees and that the abysmal depths temperatures limited to roughly +/- 1.5 C by the freezing point temperatures of salt and fresh water.

That is enough validation for me that the basic model will work, now I just need to Tom Sawyer someone into helping with all the heavy lifting required to put it together.

Update:

For the curious:

Chaotic Elegance - Caution! Chart is just for illustration

One of the commenters on Dr. Curry's blog wanted me to simplify the chaotic patterns of various tropical ocean reconstructions compared with the EPICA CO2 record and the 65N solar insolation model.  It  is not all that easy for me to do, but I gave it a shot.  Because the samples of the the various SST data have different ages determined by C14 carbon dating and the number of samples per time period vary, the "Average" SST is at best a rough estimate.  What I did was average 10k year ranges and then average the averages.  While I think I have the done that without too many mistakes, I am not guaranteeing the accuracy or even the relevance of the reconstruction.  It is just an illustration of the mystical chaotic interactions of widely space ocean heat capacities. Because of the inconsistent ages of the samples, the age of the "average" varies by +/- 3K years roughly, so be careful.

I enlarged the "Average", EPICA and 65N solar leaving the original reconstruction "averages" muted in the background.


UPDATE:  This is a average that uses only data within ~2000 years where all series have valid data for that period.  The plots are steps showing transitions between sample and not smoothed.  There is a little difference between the two, but not that much.

Note the 250 5.35 is the log CO2 relationship showing relative CO2 forcing just for fun.

This is a snapshot of the 110K to 150K period in the chart above without the averaging or smoothing.  You will note that there is about a 3 to 5K year difference in the alignment of the curves, some don't assume much in the way of lead/lag relationships from the simplified chart.  That would require an much better procedure or more uniform data which I may be able to interpolate a little better, but don't count on that.

States of Climate II

Southern South America temperature reconstruction by Neukom et al versus Atlantic Multi-decadal Oscillation recosntruction by Gray et al., simple comparison of hemispherical ocean oscillations caused by perturbations, mainly volcanic in this case, causing cooling in the northern hemisphere followed by cooling in the southern hemisphere as more internal energy from the southern oceans is transferred to the northern hemisphere.  With the separate y-axis orientation, the AMO amplitude on left scale is roughly twice the SSA amplitude on the right scale.  Note the 1816 event rapidly reduced northern Atlantic SST with an more gradual decline in the SSA temperatures.  Since the Southern South American reconstruction is highly influenced by the temperatures of the southern oceans, it make a reasonable proxy for the southern oscillation.  There would be a continuous warming trend due to this perturbation from ~1820 in the North Atlantic and ~1850 in the southern oceans.  That would be ~150 years of natural recovery from a natural cooling event of ~0.5C degrees.

  The same recovery with the same dip, or Super Duper La Nina is seen in the Tmin data for Oceania as published in the BEST surface temperature data.  One would need to be blinded by ideology not to see the obvious.

 Just added the HADSST2 Southern Hemisphere to the tail of the SSA reconstruction for them that would complain.  Yes, it is warmer now.

States of Climate

Hopefully, this simple diagram will explain to many what I am searching for, the states of climate.  Not every possible state, just the more common states that generally cause shifts in climate.  R is the distance from the Thermal equator to the average sea ice extent or equivalent thermal sink potential.  Currently, climate is moving closer to the 1.5R state or where the thermal capacity of the oceans are roughly in balance.

Added: For those that have not been following, the true equator is a natural boundary for atmospheric circulations, but due to the location of the land mass and oceans the Thermal Equator or center of the average thermal mass of the planet is south of the true equator.  That changes the thermal equator to polar heat sink transfer times creating an natural oscillation potential.

The upper state, R versus 2R, would be the most unstable climate state and R versus R would be the more stable glacial period state.  While the 1.5R state would appear to be equivalent to the R state, the 1.5 State allows more uniform atmosphere and upper ocean mixing layer warming while decreasing instability.  This would be the high "normal" of the bi-stable climate regimes.  The R glacial state would be associated with the low "normal" of the bi-stable states.

Since the Earth is spherical, in the R state, separation between the thermal equator and the polar sinks would be at a minimum and the area of the polar sink would be at a maximum.  Polar cooling in winter would have less impact and the thermohaline current would have less impact on climate oscillations.  However, since climate would not impact tidal forcing, the more massive ice expanse would be less stable.  With more northern hemisphere ice "fixed" due to land mass, southern hemisphere ice would be more likely to dislodge and cause climate fluctuations.  With the right timing, this would cause a more rapid shift to the unstable R/2R state.

Interestingly, the average sea surface temperature would be nearly the same in both the high normal and low normal states.  The lower SST would be more associated with transistions as the more unstable atmospheric conditions would improve mixing of the upper ocean mixing layer and the deeper ocean layers.

 The exit from the most recent glacial maximum appears to be a fairly uniform rise from a more depressed or colder than usual state.

The exit from the glacial 135k years ago appears to be drive more by southern hemisphere event that lead the solar forcing increase and the CO2 rise in the EPICA ice core data.  So the last glacial minimum was more influenced by the longer solar reduction that the previous glacial period.  The actual lead would be impossible to determine with such coarse data, but there appears to be an unforced or at least less forced transition from glacial to inter-glacial.

So while everyone quibbles over other details, this is what I am working on.  This may have already been solved or refuted by someone else, but since I am playing with more non-standard methods, that really doesn't matter.

Update:  There is a new post on using network analysis to locate synchronization in climate oscillations over at Roger Pielke Jr.'s blog,

Guest Post “Atlantic Multidecadal Oscillation And Northern Hemisphere’s Climate Variability” By Marcia Glaze Wyatt, Sergey Kravtsov, And Anastasios A. Tsonis

They use a much more rigorous method.  What I am doing is more of a black art, but with a similar goal, to isolate synchronizations.  "Synchronization refers to the matching of rhythms among self-sustained oscillators; although the motions are not exactly simultaneous. If two systems have different intrinsic oscillation periods, when they couple, they adjust their frequencies in such a way that cadences match; yet always with a slight phase shift (lags)."  

Super Duper La Nina?

It is fun playing with not standard methods in non-linear thermodynamics on a global scale, but there are not a lot of people that have a clue WTF I am talking about.  So making some wild ass predictions that will be called "crack pot" ideas, is par for the course.  So above is my latest crack pot theory, the 1940s to 1950s Super Duper La Nina event.  The BEST Tmin data above shows very clearly the 1998 Super El Nino.  In the blue I have highlighted my Super Duper La Nina.

In my post, The Best Place to Start is the Beginning, I am trying to show that understanding the Antarctic Circumpolar Current is the key to understanding global climate.

The normal ENSO cycle is related to the changes in the westerly equatorial winds which increase (decrease) the polar current flow up the western coast of South America.  With only a current 2 to 4 year ENSO cycle mainly in the warmer EL Nino sequence, there is not much change in the Circumpolar Current (CC).  In a more prolonged negative of la nina ENSO mode, the CC could cause significant cooling of the Antarctic peninsular increasing the western sea ice extent.  That increase would cause a more perminate disruption of the CC flow and a more persistent La Nina phase of ENSO.  Voi La' Super Duper La Nina!

As crazy as that may sound it fits with the goofy rapid climate change events in the more distant past.

Here is the Western Caribbean with the Milankovic cycles and Antarctic CO2.  The Western Caribbean temperatures are extremely stable until some event, and those events have a greater impact on Antarctic CO2 than the Milankovic cycles.

All good "crack pot" theories require more detail, so I will start exploring more regional reconstructions to dot the "i"s and mess with the "t"s here pretty some.

Keep it touch.

Milankovic Cycles and Climate

The Earth orbit around the sun is less than perfect.  The variations in orbit are used to explain the massive glacial periods and inter glacial periods.  The chart above has the Milankovic Cycles plotted as solar forcing at latitude 65N in blue compared to various tropical ocean surface temperature reconstructions.

I got the data to plot the Milankovic cycles for the spread sheet in a post at Watts Up With That, "Onset of the Next Glaciation". and the Sea Surface Temperature data form the NOAA paleo website.

In this chart I have just the sea surface temperatures which show that there was a major event of some type around 150,000 year ago.  That event only changed the average SST by about one degree C.  The range of temperatures if averaged is only about +/- 2 C degrees for the past 400,000 years.  The event timing jives with the Vostok ice core temperature and CO2 concentration data.  The next coordinated "event" is around 360,000 years ago in the SST data.

Comparing to the Antarctic data provide by Wikipedia, the period between those two events somewhat compares to the ice cores.  The range of temperature change in the ice cores though is 5 times or more the change in the SST data.  CO2 outgassing would be dependent on the average ocean temperature near the Antarctic.

Courtesy of the NOAA paleo site again, Luthi, D et al. with the funky u put together a nice reconstruction of the Antarctic CO2 concentration which I have added in large red.  I made no attempt to scale, just used the secondary axis a let fly.  As you can see there is a fair correlation of CO2 with SST, but it is not apparently driving climate, at least in the tropical oceans.

Since the sample ages are inconsistent between data sets, it may take me a while to do a proper analysis, but then that is not my yob mans.   I am sure some grad student somewhere would love to take this ball and run with it.  If they figure out the fine details of the Milankovic versus CO2 theories, they may get a decent grade.

Just to give them a good start, the plot uses just the Western Caribbean surface temperature data with the Antarctic CO2 and Milankovic cycles.  The smaller fluctuations in the WC temperature near and above zero appear more related to solar fluctuations and the larger somewhat erratic excursions below zero may be related to internal conditions.  With the exception of the most recent 40k years, the WC temperature returned to "normal" rather quickly.  The correlation between Antarctic CO2 Western Caribbean SST is not impressive.

What is the Average Global Temperature?

The Berkeley Earth Surface Temperature project redid the already done global mean surface temperature.  Unlike the other surface temperature compilers, BEST has a chart of the global temperatures with a number in real degrees not anomalies.

They even have one for the Northern hemisphere,

And the Southern Hemisphere

According to BEST, the global land surface average temperature is about 10 C degrees, the Northern hemisphere land surface average temperature is about 11.25C degrees and the Southern Hemisphere land surface average is about 7.8 C degrees.

The land in the southern hemisphere only amounts to about 31.8% of the total global land area and the northern hemisphere land is about 68.2% of the global land area which totals 145,547,000 of the 510,072,000 kilometers squared of the Earth's total surface area, or 28.5% earth on Earth. Wikipedia has a little more land at 29.2% and most folks just round off to 30/70 as the land to ocean ratio.

If 30% of the surface has an average temperature of about 10 C degrees, the other 70% should have an average temperature also.  That number is generally give to be 16C degrees or 289K degrees.  It used to be 288K degrees, but someone had to change that number a touch.

Now most of the "experts" on the fora say that the average temperature of the oceans in 17 C degrees.  If I remember basic ratios, 0.3*(10) + 0.7*(17) = 14.9 degrees which is the original estimates by the "experts" prior to the warming which started in the 1950s.  According to the AQUA satellite, the average temperature of the oceans is about 21.1 C degrees.  0.3*(10) + 0.7*(21.1) = 17.8 C degrees which is about 1.8 C degrees more than the average temperature estimated by some of the "experts".

When I used the 21.1 C degrees as the average temperature of the oceans, the "experts" said I was wrong, the average was only 16 to 17 C in their "expert" opinion.  Wikipedia also lists the average surface temperature of the Earth as 14 C degrees which would put the oceans at 16C degrees on average.  So the average surface temperature of the Earth is between 14 and 17.8 C and the average temperature of the oceans is between 16 and 21.1 C degrees.

I also mentioned to the "experts" than land typically is higher than sea level or it would not be called land.  Since land is higher than sea level, thanks to physics, the temperature decreases as the altitude increases, so the measured temperature at some altitude greater than sea level would have to be adjusted higher if that were to be compared with temperatures at sea level.  The temperature decrease with altitude because the air cannot hold as much energy due to lower thermal capacity cause by the reduced density.  If the average altitude of the land surface measurements is 1000meters, the temperature would have to be adjusted upwards by about 6.5 C depending on if they used the moist or dry lapse rate.  0.3*6.5= ~2 degrees, so there is up to two degrees that would represent temperature but not thermal capacity using temperature as a proxy for energy at the surface.

BEST also provides a northern hemisphere T minimum and

Southern hemisphere T minimum.  .31*3.1 + .69*5.4 = ~3 C degrees which would be the temperature most likely impacted by greenhouse gases retaining more heat.  So if we use T minimum and the ~17C for the oceans, .3*3 + .7*17 = 12.8 C degrees.  12.8C degrees would equal ~286 K degrees which is 2 to 3 degrees cooler than the estimated average surface temperature used to determine how much the Earth has warmed.  That 286K would have an effective radiant energy of ~379.5 Wm-2 which is 16.6 Wm-2 less than the 396Wm-2 currently estimated as the average radiant energy of the surface of the Earth which would be impacted by the addition of greenhouse gases.

Because of all the different possibilities, I, not an "expert", decided to use the measured AQUA sea surface temperature of 21.1 C degrees and focus only on the surface temperature and energy to avoid all the confusion of adjusting this or that to meet the need of having some "average" number than no one really knows what is or what if it matters.

Using that 21.1 C degrees and an equally simple model, I determined that the average range of temperatures that is allowed based on the estimated solar energy available and the limit imposed by the freezing temperature of water, salt and fresh, to be about +/- 1 degree on average for the global oceans.

If you consider that all the areas of the oceans included in these paleo climate reconstructions, the not the "expert" came pretty damn close.  So the next time you get into a rousing Climate Change discussion at your local tavern just remember to say, "natural variability if about two degrees" and look for a less liberal chick to chat up.

The Best Place to Start is at the Beginning

The Antarctic Circumpolar Current is the main driver of the deep ocean currents.  The chart above was borrowed from Wikipedia.  Sailors know that the winds and seas in the southern latitudes from 40 to 60 south are not for the faint of heart.  Those high winds drive the seas that creates the circumpolar current.  With the higher average surface wind velocity and little land mass to deflect the current, huge amount of energy are exchanged between the cold ocean waters and the atmosphere.  This region is the thermal regulator of Earth's climate.

Update: The Flow of the Antarctic Circumpolar Current through the Drake Passage is estimated at 95 to more than 134 Sverdrup which is 10^6 cubic meters per second.  The Gulf Stream current off Florida is estimated at 35 Sverdrup.  Somebody asked so there is the best answer I could find.  

This graph is 131 month sequential standard deviations of the UAH lower troposphere regional temperature anomaly.  The two curves with the least deviation are the southern extend and the southern extent oceans.  The low standard deviation indicates the stability of the temperatures of each region.

This chart is the sequential standard deviations of the GISS LOTI regional temperature anomaly.  Again the least deviation is in the 64S to 44S latitudes.  Finding a regional so stable in a chaotic climate and weather system is exceptional.  That stability is useful in determining the best fit to other data, especially paleo climate proxy reconstructions.  Unfortunately, most paleo climate reconstructions end in inconvenient times.  To determine where those reconstructions best "fit" with newer instrumental data can be challenging.  By using the most stable instrumental regional data and working back towards the noisier regions, it is posible to have greater confidence in the "fit" or splice of the instrumental to the paleo reconstructions.

This is an example of splicing the 64S-44S GISS LOTI temperature to the Southern South American temperature reconstruction by Neukom et al. 2010.  My apologies for misspelling the name on the chart, the link should soothe any perceived slight.  With this "splice", the instrumental data fits well with the temperature anomaly of the paleo reconstruction.  I have not include any error bars because at this time the degree of uncertainty is not easily determined.  The centered 5 year smoothing applied to each data series is intended to not overly smooth any information that they may contain, just eliminate some of the noise in each.  The object is to determine an appropriate baseline to start rebuilding a better picture of past climate with other regional reconstructions.

Perfectly "slicing" imperfect data is impossible.  Using the 1979 to 1990 inclusive satellite era baseline, p.b.l. to combine with paleo climate reconstructions typically ending before 2000, does appear to allow better combination of the different types of data.  Sea level data and reconstructions would tend to be less variable, allowing not only a simpler "splice" but qan indication of the differences in sensitivity of the differing data sets.  The chart above combines HADSST2 southern hemisphere and the UAH Southern Extra tropics lower troposphere data.  Because of the lower density of the atmosphere where the average UAH temperature is determined, there would be more variability in the temperature.  The thermal mass of the oceans naturally smooth the HADSST2 data and the Neukom reconstructions using various proxies would tend to be more noisy.  In the chart above, the 1900 to present time period is highlighted to show the quality of fit using the p.b.l. base period.

Starting the plot in 1250 and inserting the mean value of the UAH data in green and the mean value of the full SSA reconstruction starting in 900 AD, the mean temperature of the southern extra tropics would be approximately 0.5C greater than the mean temperature of this region.  The absolute value of the SSA reconstruction may be uncertain, but the mean should be useful for combining other longer term reconstructions.

One of the issues with combining paleo reconstructions is how much resolution is useful.  This chart combines Cook et al 2000 Tasmania with the Neukom et al. Southern South America and the GISS LOTI 44-64S instrumental.  Using the same p.b.l with 5 year centered smoothing there is a good deal of noise.  The mean value lines for each series is included showing that the range of means is from about -0.3 to -0.5C.  Despite the noise, that is a remarkably close range of mean for 4000 years of climate.  Of course the reconstructions may have issues.  By increasing the smoothing period, there would be less noise reducing the peak values.  Smoothing them enough, we would have a hockey stick with current temperatures about 0.4 to 0.5C higher than the past mean, but that is already shown.  Ideally, any more smoothing would match the natural smoothing of averaging the surface temperature instrumental.

Expanding the Time Frame:

Extending past climate beyond 900 AD is a bit of a challenge.  Since the Ice Ages would have a much more pronounced impact near the poles and at higher elevations, tropical reconstruction would give a better indication of ocean temperatures but not global temperatures.  The southern high latitudes may have been frozen to some point.  Antarctic sea ice advance could have shut down or greatly reduced the Antarctic Circumpolar Current.  So this next step is a bit of a guess.

The Tierney et al. 2008 Lake Tanganyika surface temperature reconstruction is added with the darker green full reconstruction period mean value and the 10000BC to 695AD section in the lighter green with its mean value.  By subtracting 0.6C from the mean of the overlap period with the Tasmania reconstruction, we get the orientation shown.  It could be higher or lower, this is just a rough fit.

Here is the full reconstruction just for completeness.  There are a few other reconstructions of past temperature that generally agree with this rough orientation.  The Nielsen Southern oceans SST for the Holocene for example has what appears to be a longer term oscillation out of phase with the Lake Tanganyika reconstruction.

The range of temperature swing is larger, consistent with Antarctic Circumpolar Current changes driving climate, but will require some deeper digging to relate to global conditions since according to this, the first part of the Holocene was a severe southern extratropics near ice age following an ice age normal temperature range.  Interesting.

During the modern era, the instrument data provides hints of the different oscillations and dampening constants of the hemispheres.

By selecting different smoothing and comparing regions, like the Tropics and Extra Tropics above, you can get a reasonable picture of the heat transfer between the regions of the oceans.  The satellite series started with a small volcanic event that impacted the northern hemisphere.  By 1991, the Southern na Northern Extratropics appeared to have been equalized only to that the Pinatubo eruption in 1991 drive northern hemisphere temperatures down again.  The extra tropics reached the same capacity again in 1996 setting the stage for the temperature equivalent of a rogue wave in the 1998 Super El Nino.  Since then, the temperatures a falling in a dampened manner with various harmonic synchronization generating smaller El Nino and La Nina events.  Much longer term oscillations are likely which are probably generating the "noise" in the paleo climate reconstructions.

The 1991 Pinatubo eruption provided a nice perturbation to the ocean thermodynamics.  Following Pinatubo, there was a self organizing of the internal oscillations that produced the nifty 1998 Super El nino.  Since the rate of heat transfer is different between the hemispheres, the dampening is a bit difficult to follow, but clear enough for the cyclomanic geeks in the crowd.  Since I don't have the more accurate 24N to 44S main thermal capacity of the oceans, the tropics will do for now.

The UAH tropical oceans are in yellow for this plot.  Comparing the HADSST2 hemisphere data you can follow the somewhat chaotic oscillations back in time.  Remember that the Southern Hemisphere contains only about 1/3 of the land mass of the Northern Hemisphere and the Antarctic continent makes up a substantial portion of the Southern Hemisphere land mass.  For the instrumental period, we have had remarkable stable temperatures well within the "normal" range of an interglacial as far as the oceans are concerned.

With the exceptions of the polar regions, there is no real reason for the average temperature of the oceans to change much during glacial/interglacial climates.  The total energy of the atmosphere would change much more than the total energy of the oceans since sea ice advance insulates the polar oceans.  The main caveat to that is the Antarctic Circumpolar Current which is the main heat sink of the oceans.

Once Antarctic sea ice extent increases to the South American peninsular, the efficiency of the ocean/atmosphere heat transfer most likely to released energy to space would be reduced.  This would likely increase the heat loss at the northern pole increasing high latitude precipitation where the mass snow and ice could accumulate at high elevations, not only in the Northern Hemisphere, but in all higher elevations of the Earth.  Oddly, the accumulation of snow on the more permanent Antarctic sea ice would likely be a significant driver of reduced global sea level where the sea ice was "fixed" or landed.  The building and breaking of these Antarctic fixed ice accumulations could produce some interesting "Red Herrings" in the iconic Antarctic ice core history of global climate.    I'll have to search for a few of those anomalies.

The tropical oceans reconstruction above for 400,000 years compiled by Herbert, T. D. et al. have different calibration periods.  Instead of attempting to average exactly, the lower glacial temperature anomaly is estimated with the blue bar and the interglacial average estimate is the red bar.  That implies a range of about 3 C or +/- 1.5 C which is pretty consistent with the +/-1 C control range likely impsed by the freezing points of salt and fresh water.  The most stable of the reconstruction is the tropical Atlantic which can be directly compared to the Western Caribbean reconstruction of Schmidt, M. W. et al. in the next segment.



Anywho, to be continued.

I wasn't up to the Caribbean yet, but since I got into a discussion on the Milankovic Cycles that don't quite match the ice ages because ice ages are not all the same, I just threw this in for the moment.  I hope it does give away the ending :)  Schmidt, M.W. et al. 2006 is on the NOAA paleo site if you are interested.

The Schmidt et al. Western Caribbean is one of many longer term reconstructions not often mention in climate science.  By combining the Western Caribbean with southern oceans reconstruction and the Lake Tanganyike lake surface temperature reconstruction you can see why.  There are longer term internal ocean oscillations which tend to confuse most folks.  The southern oceans are the main heat sink for the planet, but they don't catch up with global heat capacity changes quickly in all cases.  That produces the internal oscillations on all time scales.

The tropical Atlantic ocean and the western Caribbean have a better correlation to energy input.  This plot is the Milankovic solar cycle model with the Herbert et al. Tropical Atlantic SST reconstruction.  This is a fairly good fit with the typical miscues due to volcanic and other internal impacts which have to be sorted out.  The time scale above is in thousands of years, so techically we would be in what is called a Glacial Episode.  With not unoccupied land mass to accumulate glacial mass, there is not good likelihood of entering a glacial period.  That is an impact of Antropogenic Global Warming caused mainly by land use and our ability to clear snow.