More Land Use versus Climate

In a resent post I showed how BEST "global" surface temperature data made a very good reconstruction of "Global" Sea Surface Temperatures by using a modern satellite era baseline and scaling BEST Tave to the HadSSTv3b data,  For all the BEST Fans.  The I compared the Houghton land use impact on CO2 to the NCDC Extended Reconstructed Sea Surface Temperature ERSSTv3b) with the land only data sets to show how land amplification tends to over enhance more than just CO2 forcing.  A large portion of the land use amplification would be related to changes in the soil moisture/hydrology cycle which is more related to the land Tmin data than Tmax data.

Showing that is not that difficult, just difference the BEST Tmax and Tmin data with a modern era baseline, in this case 2000 to present to avoid some of the volcanic influences.  Not all of that correlation is just land use in the normal sense, but a combination of some natural recovery, black carbon and land use.   All land use does not have the same impact in all regions which also complicates things.

In the United States where both the temperature and land use data appear to be more complete, there is a nice fit with the expected 15 year lag for nominal soil degradation.

In Canada, there is little land use impact per the Houghton data and the Canadian temperature data was reduced and sparse.

In Russia, the is an issue with data quality and more of the actual land use changes was in former USSR member states.

In China, farming practices appear to have an inverse correlation with changes in the diurnal temperature anomaly.

Land use impact on diurnal temperature also does clearly indicate impact of actual temperature range or the impact on the temperature measurements, Suburban Heat Island Effect is likely a good description, since many of the temperature stations are rural but located near smaller community centers and airports.  Unless this puzzle can be solved reasonably, even this quick illustration of land use impact on land surface temperature data should be enough to question how much of the "Global" land amplification should be considered when estimating "Global" warming.

It is also not very comforting that all the models and creative multiple linear regressions used to "show" alarming amounts anthropogenic change didn't notice that diurnal temperature range has not been following CO2 forcing but "Green Revolution" forcing.  Imagine that?

Update: Comparing the Northern higher latitude land use, US, Canada, China and the former USSR to the difference between the NH SST and BEST also shows an interesting correlation.

 As much as a half of a degree C of the land surface temperature change could be related to land use/snow field impact. 

Simply Another Over-Simplification

A new paper is generating a lot of press with the typical "unprecedented" warming not seen in this case nearly in 44,000 years, though the authors did notice a 5,000 year old "precedented" warming.  The temperature proxy in this case was moss of the reindeer snacks variety.  The cause is of course CO2 since that is the only "simple" explanation.

The story goes that some 11,000 years ago the summer sun that shines on Baffin Island near Greenland was 9% warmer than it is today.  That is because the Earth has a little tilt and an elliptical orbit that causes the stronger currently winter sun, being closer this time of year in the Southern Hemisphere to switch to stronger in the northern hemisphere in another ~11,000 years.  In about 5,000 years the Southern Hemisphere Spring will be warmer than Fall while Northern Hemisphere Fall will be warmer than the Northern Hemisphere Spring.  This causes roughly four standard seasonal modes and mankind has only been taking notice of a portion of one so far.  The paper ignores the more subtle Spring/Fall modes and just concentrates on the NH hot, NH not so hot modes.

Unfortunately, the Baffin Island location near Greenland means its climate is inspired by the North Atlantic portion of the ThermoHaline Circulation (THC) which gets much of its flow from the Southern Hemisphere's Antartic Circumpolar Current (ACC).  This causes the North Atlantic SST temperatures to be out of phase with the general glacial mass of the high northern latitudes.

The four precessional "modes" plus the ACC out of phase lag produces fairly regular "pulses" in Global Climate likely stimulating  Bond Events every 1470 +/- 500 years noted in NH ice cores and in the tropics and Southern Hemisphere orbital harmonics of ~4000 years, 5000 years and 5800 years depending on the lead/lag of the Bond Events relative to the precessional four mode pulses.

paleo data leaves a lot to be desired, but these two reconstructions of the sub-tropical North Atlantic and the extra-tropical northern hemisphere fairly well show the Bond Events and one of the precessional "pulses" noted as the 5ka Moss zone.  Notice how the sub-tropical North Atlantic was cooler prior to the 5ka Moss pulse.

The de Menocal et al. goes back to ~21,000 BC with one hiatus making it hard to see the "pulses", but around 7000 BC there was another pulse with a drop following.  So from the 5ka moss pulse back at least another 10,000 years there would not have been any other Baffin Island ice loss likely due to ACC influences.  Higher solar irradiance could, but if the snow/ice is clean, it would tend to reflect most of the solar energy. 

So while I am sure that the mother of the paper's author would be proud of the effort, it is a bit hard for me to swallow the magical leap to CO2 done it without at least addressing why there is a Baffin Island Moss Barrier that tends to agree with more complex "Global" dynamics and Bond Events than his over simplification.

Since the older moss age falls into a range between interglacial cycles, this Tierney et al reconstruction tends to show the "pulses" better and roughly places the "pulse" prior to the old moss around 54,000 years before present.  So we could have a moss barrier around the 5 ka and today North Atlantic SST or pick any number you like.

 This shorter version of the Tierney et al with OpenOffice default smoothing nicely shows the pulses and the weakly damped decay following.  Unfortunately, paleo absolute temperature appear to not be reliable but they do tend to show the "norm" +/- a degree or so quite well. 

The data made available by the authors noted on the charts can be downloaded at NOAA/Paleoclimateology

Land Use - Land Amplification and General Data Confusion

The land and ocean instrumental temperature data is a measure of something with a mean error of about +/- 0.125 C degrees.  If all the data where measured in exactly the same way, that data would be a uniform measure of whatever it is measuring.  When you mix methods you begin to add a different kind of uncertainty that is not included in the standard published error.  Part of the problem with "Global" land and ocean data is that land is measured with a (Tmax +Tmin)/2 average where Tmax and Tmin have different influences.  Oceans are measured as a simple average of spot readings at a location that has less variability or noise.  You have apples and oranges if you can't tease out the "other" factors that are being measured.

With all the issues each of the data sets have it is hard to "convincingly" illustrate that there is an "other" influence that is not properly considered.  This is an example of land use changes and their impact on northern hemisphere land "surface" temperatures.  Land use, mainly mechanized farming and snow removal impact the land hydrological cycle.  CDIAC has estimates of land use impact on the carbon dioxide cycle that are far from perfect but useful to show how land use may have have a larger than anticipated impact on land surface temperature measurement.  There may be less "real" impact on land surface temperature, but ther appears to be a significant impact on the measurement of land surface temperature at least.

So how well can models be performing if they cannot find simple anomalies like land use impact.

Note:  The 15 year lag is an indication that land drying and general deterioration over time has the more significant impact.

Teleconnections - The Indo-Pacific Warm Pool

Teleconnections are defined by the American Meteorological Society as:
“1. A linkage between weather changes occurring in widely separated regions of the globe. 2. A significant positive or negative correlation in the fluctuations of a field at widely separated points. Most commonly applied to variability on monthly and longer timescales, the name refers to the fact that such correlations suggest that information is propagating between the distant points through the atmosphere.”

 That is borrowed from the Dr. Roger Pielke, Sr. blog with weather in my bold.  There is a difference between weather and climate which is time.  In climate science a lot of weather teleconnections or oscillations are used under the assumption they are climate indicators while climate isn't all that well defined.  A repetitive signal of an ~ 60 year period with about four confirmed cycles, that is about 240 years might qualify as a Climate Teleconnection.  The problem though is all those trick other things that impact climate.

If it were not for volcanic activity having an irregular impact on climate, the Indo-Pacific Warm Pool would be a solid candidate for Climate Teleconnection.  There was a major volcanic impact in 1816 which was in phase and another around 1915 that was out of phase.  But the Indo-Pacific is just one tiny spot on the huge planet right?

When I did the For All the BEST Fans post the intent was to show that Best land and the Global ocean SST have a strong correlation and that both have a strong correlation to the Indo-Pacific Warm Pool.  That is not news, though the scaled BEST correlation might have been interesting. 

The correlation with volcanic activity is a little more difficult to see and might be news to some.  Many assume that Solar variation due to records of Sun Spot Number (SSN) counts is THE driver of climate, but because not every point fits just perfectly, the Solar/Climate relationship has been relegated to pseudo-science by many of the die hard minions of the great and powerful CO2.  That doesn't stop the minions from creating their own pseudo-scientific explanations using "canned" weather oscillations to ignore the cause of the climate part of the system.  They assume that "their" oscillations are true climate teleconnections and conveniently ignore the fact they have no clue what causes their oscillations.  Someone else said it is a teleconnection and zeros out some they just take that as gospel and solve all the climate problems for centuries with maybe a 150 years of "proof".

What we may have with the Indo-Pacific Warm Pool is a weakly damped recovery response from the western Pacific depths of the Little Ice Age.  But what drives it?

Because of the huge thermal mass of the oceans the solar and volcanic reconstructions don't "fit" perfectly.  The more energy that has to be transferred the worse the fit, but there is always enough to be interesting. 

When you add a planetary size lag to the impacts with a 252 year cumulative estimate (252 year trailing average), you begin to see a better "fit". 

Just for the current modeling crazy, here is my contribution.

This is simply the Stienhilber dTSI 10Be "TSI" reconstruction smoothed by 21 years with one century scale lag also with 21 year smoothing compared to GISS loti global and hemispheres with 11 year smoothing.  I just use "Anomaly" because the dTSI appears to represent more than just the dTSI in its name, like combined orbital forcing

For all the BEST Fans

 The chart above is the BEST monthly "Global" average temperature with several layers of smoothing plus a volcanic forcing overlay.

The BEST land surface temperature data set is a nice piece of work, but it is still just land surface temperature data with most of the land in the northern hemisphere.  Surface temperature anomaly in general is falling out of favor with Climate Scientist because of all the noise in the data mainly due to land amplification of dozens of possible "effects".  

BEST is still useful though.  You can see all the land areas that have warmed along with about a third of the areas that have cooled.  You can also use SST data to filter out some of the BEST noise.

This is BEST "Global" Tave scaled to fit the Hadley Center Sea Surface Temperature data for both hemispheres.

Here is the BEST data scaled and spliced to a reconstruction of Indo-Pacific Warm Pool Data by Oppo et al. 2009.  The humps, bumps and slope appear to indicate there is a long term trend common to BEST and HadSST3 which might just be something truly "Global".

And that "Global" something appears to have something to do with Solar and Volcanic (Sol y Vol) activity in the past.    Imagine that?

Update:   Since I am using the newer Crowley and Unterman volcanic forcing estimate instead of the BEST volcanic forcing I should show how they differ.

I didn't try to make this a super fit just a close match to BEST on some of the major eruptions.  Circled are some difference that may appear to be insignificant at first, but with the solar component included is a lot larger than first glance.  The 1915 to 1918 influence was much larger in the SST data and likely contributed to the 1941 SST pop and drop. 

Sol y Vol Planetary Scale Smoothing

 This chart uses all of the Steinhilber dTSI and Crowley & Unterman "Global" volcanic sulphates reconstructions with the full Oppo 2009 Indo-Pacific Warm Pool.  In this case the 252 year trailing average of dTSI seems to replicate a reasonable smoothing that a planet with vast oceans might produce.  A combined solar and volcanic forcing change lasting centuries likely takes centuries to work its way through such a massive system.

So using just a reconstruction of the Indo-Pacific Warm Pool Sea Surface Temperatures which has a remarkable fit to the instrumental temperature records back to 1750 AD once you deal with the Northern Hemisphere amplification of temperatures seems to indicate that there was a strong solar and volcanic impact starting throughout the past 2000 years with the coldest period being between 1250 ad and 1850 ad.  Not inconsistent with the MWP and LIA periods that once were a part of all the high school textbooks. 

On shorter time scales you have to consider the hemispheric interplay as each hemisphere has its own dampening constant creating obvious ~20 to 60 year pseudo-oscillations with 60 to 120 year pseudo-oscillations evident in most paleo data depending on the natural smoothing of the proxy used and location of collection.  With all they complex interactions, you can get just about any theory you like to fit for some time frame, but planetary smoothing tends to indicate that the Sun and Volcanoes "done most of it". 

Since there is a lot going on in  producing the Sol y Vol reconstruction I have the simplified chart above as an eye catcher.

This chart shows the different cumulative smoothing that were used to find a fit.  The Crowley and Unterman Volcanic data was just added to the better 252 yr trailing average.  You can browse the Sol y Vol series of post to see various data sets and smoothing choices that lead to this fit for the tropics.  

Sol y Vol Where to Start

The biggest problem with the instrumental period is it is too short.  BEST has currently the longest instrumental record and that has issues with other influences than just Solar Volcanoes and Carbon Dioxide.  Scaling BEST to the Hadley Center Sea surface temperatures adds a lot of confusion, but it helps complete a bridge to paleo reconstructions, provided you remember the uncertainties.

Using the scaled BEST global with the HADSST3 global oceans I have a splice to the 2009 Oppo Indo-Pacific Warm Pool temperature reconstruction.  There is a pretty good fit with at least +/- 0.25 C uncertainty  in the 1850 to 1955 overlap with HadSST3.  With the scaling, the fit with BEST is as good possibly better from 1750 to 1850.  The Indo-Pacific Warm Pool is not the global oceans but it should represent them fairly well.

This allows me to introduce the Sol y Vol Paleo.  Unlike the sun spot number TSI reconstructions the Steinhilber 2009 dTSI is based on isotopes of 10Be and 14C which provides a surface referenced TSI reconstruction.  Volcanic aerosols impact solar radiation reaching the surface and whatever orbital influences that cause sun spot cycles could just as easily effect techtonic plate movement and volcanic activity.  It seems easier to lump solar and volcanic "forcing" together until someone can sort that out.

With this chart, Sol yVol likely started the Little Ice Age around 1240 AD.  It took about 500 years for the IPWP to cool to its minimum and so far it has taken about 300 years for the IPWP to recover.  Remember that BEST is scaled by 0.55 to make this splice meaning that warming of the land areas would be about twice the warming of the IPWP.  At the 1955 end of the Oppo reconstruction, it was still showing a weakly damped oscillation with a period of roughly 62 years.  That is the same rough period as the Atlantic Multidecadal Oscillation (AMO) which is about twice the period of the Pacific Decadal Oscillation (PDO).  In the southern hemisphere the ~62 year oscillation is muted by the more efficient mixing and can be interpreted as a centennial scale oscillation in the Pacific which if you consider a mix of  62 and 31 year recurrent recovery oscillation would appear to be a 93 year oscillation.  You can pick nearly any period that is a multiple of above or below ~31.5 year base frequency you like, if you are in to being a cyclomania. 

Is this fit convincing?  That depend on on how faithful you are to your personal theories.  Given the magnitude of change is roughly +/- 0.5C degrees "Globally" in a thousand years, I doubt it much matters. 

Sol y Vol - Dealing with the NH Amplification

Probably the trickiest part of determining the impact of hemispheric imbalance is dealing with the amplification of the surface temperatures used as the primary metric.  On the whole the data is remarkably accurate and well organized.  What the data is supposed to represent versus what it actually indicates is a bit of a problem though.  The surface air temperature is measured approximately two meters above the ground surface at an average altitude of ~680 meters.  The less dense air tends to amplify the variability plus there are many more land issues than ocean issues.  Most of the warming is over land and in the northern hemisphere oceans at first glance, but both the land and the NH oceans due to the bottleneck between latitudes 30N and 60N tend to amplify overall temperatures.

To isolate imbalance impact that amplification has to be considered.  This chart uses the BEST and Hadley Center Sea Surface Temperature on a 1980 to 2010 baseline to scale the data to a Southern Hemisphere reference, HadSST3 SH.  The BEST global requires a scaling factor of 0.55 meaning it is amplified by 181% or nearly doubled and the NH SST needs a 1.3 scaling factor since it has not warmed as much as the southern hemisphere.  The residuals of both BEST and HadSST3 NH minus the SH reference are used to "zero" in the gain.  Since the BEST data is so noisy I use a 27 month smooth to simplify things.  Because of the sensitivity differences that cause the varying gains, this scaled version should improve estimation of the actual energy imbalance required to produce the fairly large natural internal variability.  It takes a lot less energy to create large swings in temperature because temperature is not really a very good way to determine "energy" imbalances. 

The well mixed Southern Hemisphere oceans do have an advantage in the surface temperature metric department with less "active" land mass to create noise.  The Antarctic and moist tropics land masses are much more stable than the NH especially the 30N to 60N region.

I am sure there will be a few questions on this scaling, but the Toggweiler Shifting Westerlies paper mentions how much impact on surface temperature a shift in the Inter-Tropical Convergence Zone (ITCZ) can have when compared to CO2 forcing.  There have been quite a few issues created by ignoring internal variability and "amplification" due to inconsistent regional climate "Sensitivity" that will prove to be rather embarrassing sometime in the near future if my thoughts are correct. 

Sol y Vol Smoothing Choices

When I do a rough fit of the Sol y Vol combination of sun spot number derived Total Solar Irradiance (TSI) and the Crowley and Unterman volcanic sulphates forcing estimate there are a lot of choices than can be made.  I selected a 27 month base smoothing period because of internal lags in the tropics.  There is a huge seasonal swing of near 100 Wm-2 due to Earth's eccentric orbit that has to have some reasonable settling time.  13 month is a standard smoothing choice which would show more detail than 27 months.  The longer the time period selected the more information that is lost.  Any time frame I select will be question by someone with an agenda or someone thinking I have an agenda.  Such is life.

This is a comparison of the BEST global data using the 27 month cascade smoothing and the Vol part for the northern hemisphere.  Believe it or not, that is not a bad fit.  Since there is considerable uncertainty in the earliest part of the temperature data and determining volcanic forcing from sulphate deposits in glacial ice in not an exact science, this is the kind of correlations that should be expected in a complex system.  Volcanic forcing due to sulphates in the atmosphere obviously cannot cause surface cooling before they occur, but there is no reason to expect that every volcano happens to have the same atmospheric conditions when they erupt or that there are smaller volcanoes that may not be included in the data that may have some impact  Also every perturbation of the system can have a different impact on temperature depending on the dynamic state at the time of the perturbation.  The quick and dirty settling time estimate for the southern hemisphere indicates 45 years until 50% of the hemispheric imbalance works its way through the system.  90 years later there can still be a pulse of 25% of the initial impact on the ocean surface temperature which would just contribute more noise to the surface air temperature recorded by BEST.  So I tend to personally prefer the general detail offered by the 27 month smoothing.

This uses a 13 month cascade smooth.  There is a little difference in the fit but not enough to write home about unless you are trying to isolate a specific period.  In both cases I showed the impact of each filtering stage.  I blurred the BEST monthly data to show the realistic range of uncertainty.  With the blur regions of higher variability show darker bands indicating the fit in those areas is likely to be more uncertain that in the less variable bands.  In the land data there is also lots of "other" factors, glacial and snow cover advance/retreat which can be impacted by natural warming, i.e. recovery from LIA conditions, black carbon from the use of fossil fuels, agriculture albedo impact due to snow removal, dust and smoke from slash and burn clearing, crop selection and rotation, water cycle changes due to everything from beaver hat fads to irrigation and swamp drainage.  Under all this uncertain expecting a tight fit over the whole period is just a bit goofy.  The oceans tend to not have as many issues.

This is the Hadley Center SST version 3 with the 13 moth cascade smoothing.  The overall fit is generally better but with issues more likely related to ocean settling times.  The Vol used is still NH only which misses half the globe with more heat capacity.  We are also missing 100 years of data which had more volcanic impact that would likely have some delay moving through the system. 

The same data with the 27 month smooth shows less detail which may be a problem for some, but should be easier to "roughly" fit.  The extra smoothing make is it simpler to determining the weakly dampen settling periods.  Once that is done there can always be an attempt to fill in some of the finer points, but I think starting with too much detail too early just leads to missed opportunities. 

Note that these last two charts have light blue and red bands.  The blue band is the mean of the entire period and the red the mean of the 1951 to 2010 period.  Both are blurred to simulate the nominal instrumental error range.  No smoothing method will ever replace accuracy. What the 27 month stacked or tiered smoothing does though is create a simple band pass filter.  By having a larger number of tiers you can isolate the more synchronize 27 month events.

This is a bit of an extreme example where I used 17 tiers to isolate the ~62 year AMO/Global pseudo- cyclic oscillation.  If the period happens to be fairly constant you can make a prediction based on this recurrent pattern.  I wouldn't count on it though since volcanic perturbations may not be so predictable :)

Sol y Vol - Hemispheric Interplay

The Sol y Vol hemisphere imbalance doesn't explain everything.  It does provide a lot of clues though.  The well mixed southern hemisphere which is mainly slow response oceans have a more predictable response to an imbalance, like a volcano.

Even with all the noise generated by land surface temperatures you can get a reasonable fit to a 45 year dampened decay curve, e^-t*cos(2pi()t where t is 45 years.  There is no reason to expect that period to be locked at 45 years, but the general range should be somewhat consistent, you should expect it to take about 90 years for the decay oscillations to become insignificant provided there are no major perturbations during the settling time. 

If the northern hemisphere were as well behaved, Climate prediction would be a piece of cake.  It is not though.

As the SH settles back to its desired mean, more energy is available for the NH via the ThermoHaline Circulation (THC) producing a variety of possibilities.  Exponential growth is not sustainable so there will be a new tipping point restarting the sequences. 

Sol y Vol - the GISS Rough Fit

Surface temperature is not a great indication of energy change in the Earth system without considering a long enough time frame for all the internal settling to diminish to some reasonable back ground level if you are considering climate.  For weather, the decadal swings are important, for climate they are more of a PITA.  But since "surface" temperature is a majority of the data available, it has to be used.  Sea Surface Temperature is a much more reliable metric though with much sparer coverage than the ~25% of the Globe with long term land surface temperature data.

This chart compares the GISS land and ocean hemisphere data sets to the Sol y Vol imbalance or NH minus SH difference.  Since the imbalance "rides" on a secular trend the Sol y Vol was adjusted to fit the existing trend in the GISS data.  There is no easy way to determine how much the imbalance adds to the the secular trend, but this comparision shows what the imbalance can add to the variability around that trend.  This is just a rough fit since there is considerable room for in and out of phase interaction in the surface temperature averaging method.  If a baseline for surface temperature anomaly includes more or less of an imbalanced period, the seasonal signal could cause about a half of a degree error mainly due to the NH dominate land mass which should be obvious in the different variance of the NH and SH monthly GISS data. 

To get the Sol y Vol imbalance used in this chart I added an 11 year centered smoothing to the Sol y Vol hemisphere "forcing" prior to differencing to remove most of the 11 year solar cycle and in the left axis is the GISS temperature anomaly while the right axis is approximate imbalance energy in Wm-2.  The imbalance energy I am sure can use a more rigorous treatment, but with the average imbalance on the order of 18Wm-2, this roughly +/- 1.5 Wm-2 imbalance estimate would be  close to an 8% variation or about +/- 0.3 C degrees.  That is about 3 times the estimated "global" temperature anomaly impact of the Atlantic Multidecadal Oscillation (AMO) which is one of several oscillations that exist due to internal energy imbalances. 

This post is mainly to quiet the "global" surface temperature junkies that tend to gravitate to the noisiest of data to prove some point or the other.

Using detrended data above you could estimate a lower effective imbalance energy.  The difference should be due to internal mixing efficiency change.  A small shift in the average westerly winds especially in the southern oceans would be more than enough to account for the discrepancy (Toggweiler 2009).

Sol y Vol II - It's Not Forcing - It's Finesse

Most of Climate Science is over simplifications of extremely complex combinations of processes.  There is an old joke about a mechanic that looked at a car engine, took out a small hammer and tapped a part under the hood.  That fixed the problem so the mechanic charge the car owner $50 bucks for the repair.  The owner complained that was a lot of money for a single tap.  The mechanic explained, "It's not for the tap.  It's for knowing where to tap."  The tap just freed some mechanical part allowing it to function as it is supposed to.  Vibrations and perturbations tend to reset parts of system so they can begin to do the things they are supposed to do.  If you just look at the "forcing" you may not notice that the brakes are locked.

The chart above compares the Sol y Vol hemispheric imbalance described in this link with the Hadley Center land and ocean extended tropical surface temperature estimates.  If you are just looking for a gnat's ass fit you will miss the in phase and out of phase regions where system efficiency would shift from one mode to another.  In an imbalanced mode, the mixing of the system changes so that the oceans may store more or less energy.  If the oceans store more energy, the surface temperature would tend to decrease along with any decrease that would be indicated by pure "forcing".  You don't get the same response in a balanced phase as you would in an imbalanced phase.

Since different regions of the globe have higher mixing efficiencies, those regions would have a faster ocean response and slower atmosphere response than regions with less efficient mixing. This is especially important for simple radiant models of "forcing".

Internal mixing is mainly anisotropic or horizontal instead of vertical.  Radiant models are general up/down dominate with poor consideration of anisotropic energy transfer on longer time scales.  It can take twenty years or more for an imbalance between the hemispheres to settle out to a value close enough to zero to be ignored, especially if the Climate Scientist is looking for that gnat's ass level of precision.

Once the Climate Scientist believes he/she has found that level of precision, if they did not account for the impact of natural internal variability, their work is likely to be a "Crock of Shit" to use a fairly new Dendroclimatetology term.  This series is going to attempt to avoid that status by offering "envelopes" of possibilities.  

These would be some of the envelopes.  This by the way is not a particularly novel approach.  There are dozens of previous works that deal with non-linear systems.  This is more intended to be an Eye Opening approach so more Climate Scientists can avoid the Crock of Shit results :)

Sol y Vol - The Adventure Begins

With the new publication by Crowley and Unterman on hemisphere volcanic aerosols more of the pieces have begun to fall into place.  That unfortunately means I should be a little more meticulous since I am moving from a quick and dirty "rule of thumb" hunt for clues into a more refine the evidence mode.

The Crowley and Untermann 2013 Global and Hemispheric Volcanic Climate Forcing data is located at the NCDC link.  I am not using their forcing estimates "exactly".  I am using their data by hemisphere and doing my own adjustments with the goal of estimating the Hemispheric imbalance impact on internal oscillations.  The forcing estimates I will be looking for are not "global", but equatorial. 

Briefly, I have used a spreadsheet to interpolate the hemispheric "forcing" into monthly pseudo-data with 50% decay over a period of 27 months.  The 27 months, 2.25 years +/- .25 is a common internal settling frequency near the equator caused by unbalanced seasonal solar forcing.  The 50% decay rate is an initial estimate to be used as a reference.  The data is then scaled down by a factor of ten again as an initial estimate to be used as a reference.

The hemispheric volcanic estimates is then combine with the G. Kopp TSI reconstruction from 1610 to start of the SORCE satellite TSI composite data.  The TSI data converted to anomaly using the current 1361.1 average TOA TSI.  The Kopp-Comp is not scaled priopr to be combined with the interpolated NH and SH volcanic forcing estimates.
Then since filtering is always a matter of personal taste I am using a three tier 27 month cascade filter.  This requires removing 27 months from the start and end of any filter time series.  This should be similar to the filtering recommended by Gregg Goodman.

This is the results of all that fiddling with an "Estimated" Wm-2 anomaly.  Since there is some question with the relationship between volcanic aerosols effecting sun spot number count, the Sol y Vol series is simply an estimate of combined impact with no attempt to isolate either just yet, so let's call it a Redneck composite.

Since I am specifically looking for hemispheric imbalances I am using various latitude bands instead of "global" surface temperature data.  ERSSTv3b has a variety of pre-selected bands like the Tropics with Extratropics.

This is the results of the 3 tier smoothing on those pre-selected bands.  Unfortunately, some of the bands I would like to use are not "standard" for all data sets.  The extended tropics, 30S to 30N is used by the Hadley Climate Reseach Unit and can be easily created from the ERSSTv3b 30 degree bands available.  The 30S-30N band simplifies calculations since it is approximately 50% of the global surface area and less prone to seasonal noise.  Volcanic aerosols have a greater seasonal impact since they tend to scatter solar energy.  That don't happen at night and it can impact the seasonal cycle removed in smoothing the "global" surface temperature anomaly data adding to real "noise" generated in data manipulation.

Another reference I use is 0.8dT(CO2) which is 0.8*ln(CO2/CO2ref)/ln(2) where CO2 ref is the average CO2 concentration of the baseline period.  For 1980 to 2012 that is ~370 ppmv.  The 1980 to 2012 baseline is selected to provide the longest satellite era baseline.  Next year it will be 1980-2013.  If there are seasonal cycle issues, eventually they will resolve themselves with more data.  I am not using an overall baseline because of the differences in uncertainty at the ends of the data.  I am also not using 1951 to what ever because that includes more volcanic, solar and internal oscillation noise.

The CO2 reference here is used to compare data uncertainty.  It is fit to the less noisy of the two latitudinal band since this is a search for imbalance impact. 

When temperature and forcing are charted together the left axis is temperature and the right axis is "estimated" Wm-2.  If there were no internal mixing lags all three SST regions should either fit the Sol y Vol better and if there were no significant impact due to the Sol y Vol imbalance there would be no correlation at all.  The trick is determining how much of what is actually signal.  That is where the less commonly used latitudinal bands combinations will come into play like the 30S-0 and 0-30N Land and oceans minus oceans used in the previous chart. 

Hopefully, this is clear enough for the curious so I can move into the more fun stuff :)

Sol y Vol with the Extra Tropical Oceans

This is the current fit of the combined Crowley and Unterman 2013 volcanic sulphates with the ERSSTv3b extra-tropical zone oceans.  I have not included the sliver of ocean above latitude 60 mainly  because of uncertainty in ice edge.  The serious scaling involved factors of 0.1 for each volcanic data set with 27 month exponential decay to 50% then combining the volcanic data with the Kopp TSI reconstruction to which the SORCE composite TSI was spliced.  That produces the base ~ Wm-2 estimates.  In order to "fit" the 30N-60N SST data required a 0.5 scaling factor.  All told, the NH bottleneck amplifies solar and volcanic forcing by a factor of ~2 which tends to throw most of the "global" forcing estimates off since that 30N-60N SST also is amplified by the land surfaces.  In other words, approximately 25% of the global surface is causing issues with the "global" mean surface temperature metric.  The surface/atmospheric energy is real and the weather oscillations note worthy, but as far as the "Global" energy budget goes the 30N-60N region is a serious PITA compared to the quiet well mixed southern hemisphere. 

It's Complicated - That is Why They Call it CHAOTIC

Most folks have not problem at all understanding that "weather" is chaotic, i.e. has lots of turbulent mixing interactions.  "Climate" for some reason is assumed to be less chaotic, because it takes longer.  You get a glimpse of "climate" on what you think is a reasonable time scale and start thinking you can figure it out because it looks so simple compared to "weather".  If "weather" is unpredictable say 10 days in advance, a climate "day" may be 10 times 60 years = 600 years.  If you are only looking at 120 years worth of data then you only have one fifth of the information you need to leap to conclusions.

Hopefully, this is a simple enough example to show what you are up against.  This uses the ERSSTv3b data for the extended tropical bands, 30 degrees to the equator.  Each of the two band include ~25% of the global surface area.  The northern hemisphere band contains more land area than the southern.  So after differencing the land & Ocean combined data from the Ocean only data you have a "weak" signal that needs to be corrected for areal differences.  The average land to ocean ratio for these bands is 1 to 4 so to isolate the temperature difference you must multiply by 3.49 for the NH and 4.36 for the SH.

Differencing removes a large amount of the common "noise".  Then I smoothed each data set with a 27 month moving average, just over two years to allow for seasonal variations, to make things a little less "noisy".  As you can see the two hemispheres have different responses with different lags.  It may take 20 or more years for an imbalanced response to settle back to "normal" with one hemisphere alternately dampening and amplifying the settling curve of the other.  Over twenty or so years "mild" disturbances zero out, but larger disturbances that impact the greater thermal mass of the deeper oceans may take centuries or longer to zero out.  The longer lags may only be a few tenths of a degree, but if you are trying to predict "climate" to a few tenths of a degrees, you are stuck with another significant source of "noise".  Since the ~25% of the bands are land surfaces which amplify the "noise" by a factor of 4 you get ~0.4 C of "noise" that may be irreducible on realistic time scales.  There is just as much chaos in the data as there is in the system.  The tighter you make your tolerances the worse the data "noise" becomes.

I added a CO2 reference curve to the chart for, wait for it, a reference.  Now watch this closely.

This chart was the width of the CO2 reference expanded to roughly the claimed margin of error of the temperature data sets.  That is probably the most realistic precision you could ever hope for in this "noisy" combination of system and data.  Imagine that, the lower "no feedback" climate sensitivity estimated by the skeptical low ballers, Lindzen, Spencer and Kimoto seems to make sense.  I wonder why?

Volcanic Aerosol Imbalance? Whoda Thunk it!

My using reconstructed Solar TSI as a general orbital forcing reference reveal a few oddities that I didn't expect.  Ocean you get back to the earliest instrumental data, solar and volcanic forcing were just a little bit to close together for comfort.  So with a quick search, Crowley et al. 2013 came to the rescue.  They have a new 1200 year volcanic reconstruction by hemisphere.  There ya go.  That orientation is in Norther Hemisphere cooling is the negative region.  I haven't worked out the scaling yet, but that will make a lot of changes in my current mind set on responses to forcing.  Here is the link to the Crowley et al 2013 data on the National Climatic Data Center website.

As you can see here, the 1750 to 1900 section was a bit out of synch with what should be reality.  Now I get to start all over on the less than ideal aka "sub-optimal" volcanic aerosol reconstruction used by BEST. 

 In addition to the hemisphere imbalance there is also that circa 1243-1255 volcanic "GLOBAL" Little Ice Age trigger. 

Solar with Various Cumulative Lags

When using solar TSI as a reference for internal charging efficiency, the ENSO shifts, there are a lot of shorter term oscillation periods to contend with.  Longer term the periods tend to blur making it difficult to isolate an exact lag period.  This chart compares a variety of cumulative lag periods and I have "blurred" the 31.5 year period to show how it includes most of the other lag choices.  The "blur" is about +/- 0.15 Wm-2/C degrees depending on the scaling used to compare with any particular temperature record or reconstruction.  The different lags represent different charging efficiencies.  If the ocean mixing efficiency is poor, the surface response is larger with a shorter period.  If ocean mixing efficiency is good, the surface response is weaker and the oceans gain more energy.

Since the actual temperature data for the longer time frames is not very accurate relative to the "impact" trying to be determined, the blur provides a sanity check for the more devoted Cyclomaniacs.

The trailing cumulative lag averaging is just a cheat for comparing data sets.  Most of the data is already averaged and seasonally adjusted to death so a proper integration of smoothed and seasonally adjusted data is most like mathturbation of the finest kind.  A. M. Selvam among others have mentioned that cumulative impacts should be considered, my simply Cumulative Lags do just that.  If they need to be fine tuned once more reliable data is available, then so be it.  

Update:  This chart uses the 31.5 year cumulative average plus I added the BEST volcanic forcing with BEST global and ERSSTv3b tropics and extratropics all spliced to the Oppo 2009 Indo-Pacific Warm Pool reconstruction.

From this Volcanic forcing can have a longer term impact requiring several decades to 100s of years for all of the regions to catch up.  This used a 1980 to 2010 baseline and of course the uncertainty in the data increases moving back in time.  Some anthropogenic forcing will likely be required once a reliable baseline "normal" can be determined. 

Charging Gear Shift - a Solar/ENSO Charging Path

For whatever reason, there was a Little Ice Age that appears to be about 0.9 C cooling than today or possibly 1940.  Depends on your choice of "normal".    Solar tends to fit the the rise from the LIA, but not perfectly enough for the carbon minions.  So I have put together two charts of Solar Possibilities.

One period that seems to be common is the 27 month or 2.25 year lag in the tropics.  There is also the Hale cycle which is roughly 20-21 years and I included a 3rd harmonic, 6.75 years. I used the longer 20.25 year curve to blow up the margin of error in the instrumental.  If it was not for the 1945 la nina, the fit would be very good.  But El Nino/la nina events occur and the stronger ones before a shift or regime change.

Oops! Spreadsheet error.  11.25 years should be 47.25 years making 15.75 a 3rd not a 5th.  I have to revise the text below or do a do over, but 47.25 years with a 15.75 year shift results in a 31.5 year "pseudo" oscillation in the NH.  Since the actual solar ~11.25 year cycle ranges by approximately 2.25 years from ~9 years to ~13.5years there would be a range of NH "pseudo" oscillation periods.

This one is still based on the 2.25 year base lag but with a 5th, 11.25 years which is close to the "standard" solar cycle and a 7th, 15.75 years which would be an 11.25 plus a 4.5 lag.  The internally generated 15.75 pseudo cycle would be a 3rd of a longer 47.25 years settling pseudo-cycle about mid-way between the solar 3rds.  Any lag that synchronizes with a fundamental would produce some amplification.  This one is kind of interesting because the 11.25 47.25 and 15.75 curves show two paths.  An ENSO free path, slow charging and an ENSO path, fast charging.  Not that this is what happens, but this is the kind of thing that can happen in a bi-stable system.  You try to rev it up too fast and it shifts to a slower gear.

I left the solar in Wm-2.  If you consider a 0.8Wm-2 ocean imbalance, once the "charging" is complete or hits a charging speed bump, the imbalance would add to the surface energy instead of being absorbed.  With an average ocean temperature of about 18 C, stopping a 0.8Wm-2 imbalance would basically increase the average surface temperature by .15 C degrees.  That increased surface temperature would interact with the atmosphere potentially amplifying the impact by 2 or up to 0.3 C degrees.  For example if instead of just staying "average" that energy is transported poleward it would have an impact of 0.17C potentially amplified to 0.34C degrees.  So the reduction in imbalance could produce between roughly 0.2C for a round number and 0.34 C.   Solar in this case would not be driving climate as much as just recharging climate or adjusting the climate charging rate.

There are plenty of "other" potential feedback items like water vapor and ice/snow melt, but if solar was at one time assumed to cause the 1940s temperature peak, shifting into a lower gear could have extended that solar forcing to ~1985. 

Since there is not real way to determine how the internal oscillations/damped response curves will synchronize with so many irregular forcing/feedback mechanisms, the path shift or gear shift is just one of many possibilities.  It does look interesting though.

Since solar/volcanic impact varies regionally, I may be able to compare a few different regions to see what periods may be more likely to synchronize.

UpDate: Since I had the spreadsheet error I revised the last chart to make is less confusing hopefully and included the southern hemisphere ocean band fro 30S to 60S for comparison.  There is more involved than just solar, so I scaled the ERSST ocean data by 0.75 to improve the fit.

The longer lag appears to be more dominate in the well mixed southern oceans.  As I said, this is not "The" response curve just a possible combination of internal mixing lags with varying forcing.

Just to be complete here are other ocean regions.

 This last one 30N-60N is the problem child.

And this one is the longest instrumental record.

 I cheated a little bit by showing the approximate error margin, about +/-0.125 at the end and over +/- 0.25 at the beginning.