New Computer Fund

Thursday, November 28, 2013

Stop Assuming so Much

When I mentioned that the AMO and PDO are defined oscillations best used for weather not climate I feel the deer in the highlights look.  The AMO and PDO are effects.  Something causes them on longer time scales.  If all you are concerned with is weather patterns, they are fine, but if you are trying to predict climate, without even knowing what time scale is best or climate, you don't just assume things are fixed oscillations. 

Ocean heat transport is the "cause" of the oscillations.  The chart above is the percentage ocean by latitude.  65N has the least ocean and most land so it is a ocean heat transport "choke" point.  The Thermohalide circulation and Coriolis effect along with equator to pole temperature gradients pump energy poleward.  The choke point limits that transport amplifying the impact of ocean heat content in that region. 

This chart compares the 30N-70N ERSSTv3b ocean surface temperature with the BEST land only data for the same region.  The BEST data is scaled by a factor of 0.24 in order to match the trends of both data sets. 

Fans of the AMO will have noticed how similar the 30N-70N SST looks.  This compares the Kaplan AMO with the 30N-70N SST and the yellow curve is the difference between the two. 

Surprise, surprise, the difference bears a remarkable resemblance to the Pacific Decadal Oscillation which has to be scaled since it has been defined as a weather oscillation based on the Aleutian Low.

They are not perfect fits, but the AMO combined with the PDO properly scaled pretty much replicate the 30N-70N SST.  If you have the 30N-60N SST though, why do you need to replicated it with couple of combined weather pseudo-oscillations? 

Since the AMO and PDO are the results of changes in ocean heat transport, not the causes of ocean heat transport it is a little twisted to consider either "causing" climate to do anything.  They do impact weather which really should be considered a different subject. 

Tuesday, November 26, 2013

The Atlantic Multi-decadal Oscillation Misconceptions

Vaughan Pratt, one of the more qualified commentators on the Climate Etc. blog made another one of those comments that just mystify me, that is impact on the AMO is only +/-0.1 C degrees.  The AMO is an oscillation by definition not design that is detrended for easy illustration more than anything else.  The 0.1 +/- C is a result of the use not the phenomenon.

Once you take an anomaly after detrending your mean or average is locked in place then with a little extra smoothing like annual anomaly you pretty much have lost all of the reality of the "thing" being used to create the oscillation index.  This is a rough mask of the North Atlantic from 20N-70N and latitude 20E-90W.  The average anomaly in orange has all of the seasonal signal and the anomaly in blue has the average seasonal cycle for the full period from 1854 to 2012 removed.  The average of the full period is 16.2 C degrees and the median is about a half degree lower at 15.7C degrees.  The comparison seems to indicate that the "AMO" signal drifts about +/- 0.5 C which is about 5 times larger than Dr. Pratt's +/-0.1 C degrees.  Dr. Pratt appears to have underestimated the impact of the "AMO" because of the smoothing methods he is trying to explain and the general confusion over what is and is not a climate "oscillation".

Since Redneck's aren't statisticians or logicians in the formal sense, all I can do is question the "common sense" in assuming something is "normal" and negligible when from the looks of it the formal statisticians and logicians seem to have underestimated the potential impact by a factor of 5.

Dr. Pratt does seem to be more impressed with the Pacific Decadal Oscillations (PDO) because that has a larger swing so by his logic it can have more impact.  The PDO is based on how North Western Pacific fish stocks respond to climate oscillations which varies more in the North Western Pacific than it the whole northern Pacific.  Same basin range of fluctuation just a different region picked out for different fish.  If he looked at the entire northern oceans from 20N to 70N he would find that the whole shebang fluctuates pseudo-cyclicly.

I used both axis to highlight things with this one for the northern oceans.  It has been noted that the northern hemisphere with a larger percentage of land tends to amplify temperature changes which some seem to think indicates that it's "worse than it looks" because they assume they know what "normal" is supposed to be pretty much like they assume that defined oscillations mean more than they are supposed to mean. 

I have to admit though that thanks to Dr. Pratt and Greg Goodman I now know how to smooth the crap out of any time series. 

Sunday, November 24, 2013

A Guest Post on Volcanoes? I Don't think So.

If you take the "standard" volcanic forcing estimates and try to isolate the impact of volcanoes on climate you will find that things just don't add up.  That is because the concept of a "standard" forcing is flawed.  Without going into a great deal of math, try to clear you mind and think of what you have sans all the theories.

You have a ocean with a sea surface temperature that is always greater than the average temperature of the oceans.  You have an average temperature of the oceans which is about equal to the land surface temperature.  You have a northern hemisphere which is about 3 degrees warmer than the southern hemisphere.  If you change how well the sea surface temperature mixes with the deeper oceans or divides itself between the hemispheres you will change the "average" surface temperature.  There is no "forcing" required in the radiant physics sense, just changes in the mechanical mixing efficiency.  A small imbalance in "forcing" can have a greater impact than a larger "global" "forcing" change.  It doesn't matter if that forcing is positive, negative, due to volcanoes, the sun or unicorns, imbalances will always have a greater impact on shorter time scales than uniform forcing. 

Why?  Because a "uniform" forcing reduces the potential of imbalances, decreasing the mixing efficiency actually slowing the rate of warming.  Since the Earth land and oceans are not symmetrically distributed around either the equatorial or polar planes, there will always be some imbalances and temperature gradients, uniform forcing just serves to reduce the degree of imbalance. 

My writing a guest post on "Volcanic Direct and Indirect Effects on Climate" would be a complete waste of time because the chosen radiant frame of reference doesn't allow the communication of the basics completely ignored by radiant physics based climate theory.  For "global" warming in the radiant sense everything "globally" would warm at the same rate, slow as molasses.  One full overturning of the oceans takes on the order of 1700 years.  It would take an "average", based on the limited data available, of ~316 years for the "globe" to warm 0.8C.  By using land based and surface skin measurements warming can "appear" to be greater, but once you back out the internal pseudo-oscillations, about 316 years per 0.8 C degrees.

This chart shows the combined volcanic and solar impact on degree latitude bands of the oceans using the ERSSTv3b data downloaded from KNMI Climate Explorer.  You can see the complex recovery paths of each band with the Northern Hemisphere having the fastest recovery producing an overshot of the mean and the combined oceans regions "hunting" for a new "equilibrium" or quasi-steady state condition.  The only spot the bands are even close to being in synch is during the 1910 period.  The most interesting of all the bands is the 65S to 55S band which has the highest mixing efficiency.  The strongest responses are related to the lowest mixing efficiencies with 35N-45N and 45N-55N located at ocean heat transfer choke points being the strongest.  That choke point causes the land surfaces in that region to amplify the impact of the reduced mixing efficiency resulting in land temperatures being amplified by ~1.8 times the SST change.  Some portion of that amplification is likely due to "other" causes, but without the choke point, land warming would be more uniform. 

The impact of the change in mixing efficiency has been highlighted in a number of papers focusing on climate of the past.  In their paper, On the Relative Importance of Meridional and Zonal Sea Surface Temperature Gradients for the onset of of Ice Ages and on Pleocene-Pleistocene Climate Evolution, Brierley and Fedorov estimate impacts of 3.2 C and 0.6C for respective impacts.  It is not like the information is not out there, it is that the radiant "uniform" forcing models are in conflict with the reality.

To "explain" how imbalanced forcing can both warm or cool depending on region and timing requires a audience capable of listening, not an audience wedded to a failing theory. 

Saturday, November 23, 2013

Just for Fun - Battle of the Surface Temperature Reconstructions

The Berkeley Earth Surface Temperature (BEST) program supposedly has a "global" combined land and ocean temperature series that is ready but just getting some last minute tweaks and reviews.  I have been looking for it to hit the news but I keep getting tired of waiting.  GISS land and oceans surface temperature appears to have a baseline/seasonal cycle selection issue that I have been wanting to see how much it might impact trends, especially the "pause".

The difference should not be much, less than the stated error margin, which normally would be no big deal.  Since climate change is a political hot potato though it seems every milliKelvin is a battle ground.  So I built my own simple "Global" surface temperture record using the full baseline and seasonal cycle for periods where both hemisphere actually had data. 

Tah dah!  As advertised there is not much difference since GISS loti uses the same "global" ocean data ERSSTv3b and what difference there is is mainly near the end where long range interpolation used by GISS might tend to over emphasis Arctic Winter Warming.  BEST uses kriging which should be more reliable than simple interpolation provided you avoid using unicorns in the sky for a reference.

In a previous post I showed how the inclusion of the Antarctic data had added to the variance of the southern hemisphere.

I had also shown the difference in the northern hemisphere which I suspected was due to Arctic Winter Warming.  So now I have just upped the accuracy a tiny bit by removing the seasonal cycle from both the land BEST and ocean ERSSTv3b data an baselined to the entire period which is supposed to be the way it should be done.  This kind of sucks though because every year the entire reconstruction would need to be adjusted to the newer, longer baseline. 

Since I used the actual temperatures instead of anomaly I also have a "global" land and ocean diurnatal temperature range.

And there is the "Global" Tmax and Tmin with all its seasonal cycle glory. 

While I am pretty sure that my reconstruction is pretty close, it depends on the current actual land/ocean ratio and mine is pretty old, it would be Best to wait for BEST before screaming that GISS might be off by 0.05 C.

Friday, November 22, 2013

Inches of Stuff

In the US we are still stuck on inches, feet and miles.  The French who traded with lots of different nations that used lots of different units decided they wanted a simpler "universal" system and developed the metric system based on the length of the meridian through Paris in a blatant case of prime meridian envy.  While the miscalculated that a touch, their metric units made it to the big time.

Still we in the US like things like inches.  If there was an inch of rain, there was about 2.42 centimeters of rain.  If that inch of water applied a pressure, then it applied about 0.0024558598569 atmospheres of pressure.  One atmosphere of pressure is then about 407 inches of water or about 33.92 feet of water. If you build a water barometer it would have to be at least 34 feet tall which is a little inconvenient so most liquid barometers are filled with Mercury so they would only have to be about 30 inches tall. 

I bring this up because I have used both Mercury and water (spirits most of the time so no algae grows) to test pressure in air systems.  One inch of water as a differential pressure measured using a Pitot tube is about 4005 feet per minute and the air velocity increases as the square root of the differential pressure.

There are the typical caveats about STP, standard temperature and pressure which is typically one atmosphere and ~70 degree F, another archaic units we in the US cling to more out of tradition than necessity. 

Around the Globe, the average wind velocity is about 12 miles per hour which is (12*5280feet/mile)/60minutes/hour or about 1006 FPM.  You can just imagine that 4005FPM is about .758 miles per minute or about 45 mph.  The average differential pressure for 12 mph would be lower by the square root of the pressure ratio or the differential pressure the square of the velocity ratio.  (1006/4005)^2=0.063 in. W.C. which is inches of water column. This is part of the Fan Laws which is more correctly called the Affinity Laws.

What is cool is that the fan laws are so simple that whatever units you prefer, you only have to remember one reference velocity and pressure then you can calculate your butt of with the best of them.   There are things you have to be aware of, if you really want to be a fluid flow hot shot, but the Fans Laws are the backbone of fluid dynamics.

To measure air flow differential pressure you need some standard like a Pitot tube and a U-tube or inclined manometer.

The Engineering Toolbox has one of their typically stellar write ups on the Pitot tube plus this diagram.

You have the static pressure hs, total pressure ht and what you want for velocity is hd or the differential pressure.  Once you have all that tested, as long as the system is constant, meaning you don't restrict or improve flow downstream of the "traverse" point you have an air flow monitoring station or reference if you have a reliable static pressure probe installed.

If you want to get fancy you can install an orifice plate or a venturi flow meter which can amplify the differential pressure for more accuracy or just a stack of bricks if you are a cheap skate with a differential pressure across the bricks.  Nothing to it.  Then the reference is completely independent of the total static pressure.  Note that total static pressure is not the same as just old static pressure since it includes the velocity pressure.  You can use total static or static as long as both sides of you bricks measure the same thing.  The "same" thing tends to throw some people off.

Orifice Plates and Venturi Flow meters use known changes is velocity pressure to measure the air flow.  If you know the area change and the performance characteristic of the meter and fluid, you have a high quality and very accurate measuring station based on the difference in total static pressure and static pressure.  Since they are based on the difference, the measurements are independent of the total pressure up to the limit where the operating temperature and pressure impact the characteristics, Density, Viscosity and Phase of the fluid plus there can be some laminar versus turbulent flow issues if you over size or under sized the orifice. 

Neglecting the velocity component of the total static pressure was a common problem with variable air volume components back in the day.  If you have 1 inch of static pressure available and required a velocity pressure of 1 inch (45mph velocity), you never got what you wanted.  You needed the 1 inch plus the losses downstream of the restriction.  So a lot of systems never would work quite right without some TLC, like enlarging the entrance duct to allow static regain which is nothing more than reducing velocity so that the velocity pressure reverts to static pressure, smoothing approaches or increasing the fan RPM, BHP and system TSP which increased operating cost and noise levels which was not a good thing.  Some people would over compensate which resulted in an extremely quiet but expensive system.  It is always better to get it closer to right the first time.  So know your fan laws if you want to play with fluid flows.

The reason I have this post is because one of the local denizens is making the simple mistake of assuming that as long as the static pressure is fixed, the average velocity pressure is fixed.  Nope.  Velocity pressure is dependent on the differential pressure which has a lot more involved than just atmospheric pressure and temperature.  

For example the Southern Oscillation Index (SOI) is based on the monthly average sea level pressure between Tahiti and Darwin using the formula 10x((average monthly Tahiti MSLP)-(average monthly Darwin MSLP))/(long term standard deviation of difference for month) which amplifies the monthly deviation by a factor of ten.  Since the MSLP is typically in millibars and there are a few minor variations on depending on baseline selection and the need to use the 10 multiplier, but the base differential pressures are the same.

Tahiti is located at Latitude:17 32 S Longitude:W 149 34 W  and Darwin at 12 28 S 130 50 E

Using the KNMI Climate Explorer and the Kaplan Surface pressure reconstruction I put together this chart for the Tahiti and Darwin areas, this is not the exact match to the SOI since the areas are 20 degree longitude bands by 10 degree latitude bands roughly centered on each reference area.  Each of the series has a long term trend related to general shifting of climate related to the Inter-Tropical Convergence Zone (ITCZ) or shift in the Earth's Thermal Equator.  J. R. Toggweiler with the Geophysical Fluid Dynamic Laboratory (GFDL) has an excellent easy to read paper on the subject called Shifting Westerlies

Toggweiler is mainly an ocean modeler and tries to related a good deal of his work to paleoclimatology which for some reason tends to through people off.  Paleo is our reference to the past and if it happened before, especially recently, it is likely to happen again.  So hundred year and longer climate shifts being confused with some other "forcing" like CO2 for example, will likely happen again. 

Now as far as "weather" goes, using the fans laws to predict wind velocity gets you in the ballpark but all those other things that have to remain equal are lurking.  For example a super typhoon/hurricane might have a central pressure of 890 mb which compared to "normal" of about 1010 mb would be a 120mb low or ~48 in. W.C. of differential pressure.  Using the basic fan law formula the maximum velocity would be about 311 mph or about 275 knots.  How tight the pressure gradient is makes a large difference, but I you have four foot of water pressure differential you have yourself some serious winds.

Differences in saturation vapor pressure is still a pressure differential which is enough to enhance summer sea breezes and the afternoon thunderstorms common on the coasts.  It only takes an inch of pressure differential and the wind itself creates as lower local pressure which increases surface evaporation helping the storms intensify.  Luckily these dynamic processes tend to destroy themselves or life might be way too interesting.  It don't take much to get the ball rolling though.

Little Things that Bug Me

On the whole I try to stay out of the temperature series debates because on the whole that are pretty darn impressive.  Having an error in the ballpark of +/-0.25 C for a whole planet is a remarkable accomplishment.  The are still blemishes, but they require squinting which now has become completely over done with the Pause debate and the new Cowtan and Way kriging into the stratosphere method. 

While all this is going on the most obvious blemish is completely swept under the rug, the Southern Hemisphere divergence.

The data for the Antarctic didn't really start until the late 1950s.  The coldest place on the planet has the least data for a good reason, it is the coldest place on the planet.  Even though the Antarctic is only around 4% of the entire globe including sea ice extent, it has been a battle ground for sneaking in a tiny bit of "global" warming for good measure.  Sometimes tweaking can get a bit out of hand.

Since I have mentioned how baseline selection can really change the look of the data and changing the baseline can help isolate common breakpoints, I though I would show what baseline selection could do with the "global" surface temperature data starting with the southern hemisphere.  GISS land and ocean is a combination of the GISS land surface series, dTs and the ERSSTv3b now with no satellite inspired cooling bias.  KNMI Climate Explorer has the entire ERSSTv3b data available to 1854, though it is rarely seen starting before 1880.  So I masked the southern hemisphere and removed the seasonal averages from 1854 to 2012 to create a full series baseline.  GISS dTs is already in anomaly so I just rebaselined  both to 1880-2012 then using the land ocean area percentages created a quick and dirty southern hemisphere time series.  I have checked a few times, but there still could be a spreadsheet error I may have missed.  As you can see there is a large difference in the variance starting in the late 1950s which the longer baseline period and seasonal cycle appear to remove.  I can't vouch for the data from KNMI, but it appears to be in order. 

Since I did the SH I might as well do the NH.  Same situation with a more uniform reduction in the variance using the full series baseline and seasonal cycle.  Picking a shorter baseline typically produces a "dimple" at the baseline period which is barely noticeable if you look hard.  The difference in terms of error is really not very much, ~0.12 which is the advertised uncertainty.  The largest difference though doe happen to be right in the "pause" region most likely due to Arctic Winter Warming getting more weight than usual.  Since the simple combination I made treats sea ice as ocean not land, it would have a noticeable impact if one was inclined to threat ocean as land.  I am more concerned with actual  heat in the system instead of every wiggle in some location -30+ degrees below zero, so this is more the type of "average" surface temperature anomaly I would expect.

As I said, the difference is not much to write home about, but since everybody and their brother trying to boost trends in their favor, I just had to toss this out as one of those flat forehead moments.  The data has an uncertainty range for a reason gang.  

Sunday, November 17, 2013

Charging Analogy with Rough Numbers

I use a charging analogy because it is simple for most to understand and it is related to R resistance to charge and C the capacity to be charged or recharged.  The neat thing about a planet is that R should be closed to fixed and since CO2 equivalent forcing is to be added, R is what should change.  If you know R, then you can start teasing other information out. 

In the legend you have the equation which is the same for each curve and a rough value for C, the capacity being recharged combined with the years required.  So instead of a real C, you have an effective C.  The R-value selected is 0.0192 which is an extremely rough guestimate based on what I considered best fit.  I selected the ice free portion of the oceans to reduce some of the noise and the Oppo Indo-Pacific Warm Pool reconstruction starting in 1700 to extend the instrumental back in time.

I did not try to make an exact fit.  There is a lag that is questionable in the Oppo data due to the natural smoothing of their deep ocean core samples plus their binning method.  I don't expect the IPWP to be perfect so I led the fit a little just so the curves would be easy to see.  The starting values are nearest tenths of a degree which given the data may be at best +/- 0.25 C degrees is another "ballpark" estimate.

What is interesting with just this rough estimate are the C values of 15.4, 45.6 and 73.2 years respectively.  Pseudo-oscillation periods could easily be related to the charging rate explaining some of the more "pseudo" aspects of these "oscillations" that really aren't.  It may be nothing, but the values are interesting. 

There is a smoothing note on the ERSSTv3 65S-65N data in the legend.  That is for the 27 month stacked smoothing I find works well, leaving some of the more interesting short term "oscillations" and/or noise.  The data can be downloaded at KNMI Climate Explorer and the Oppo et. al. 2009 Indo-Pacific Warm Pool data at the NOAA NCDC Paleo website

The charger analogy is a ridiculously simple model which I think might be very helpful with this ridiculously complex problem.  The model with these rough estimates leaves about 0.2C degrees of the SST warming for the area selected for "other" than natural response to perturbations.  Since the 65S-65N SST doesn't include the land or polar amplification which can be due to both natural and "other" causes, don't get too crazy reading more into the rough estimate than is there. 

Climate Time Series Smoothing and What is Statistically Proper

In the chart I have the "raw" data for the Indian Ocean sea surface temperature via Climate Explorer from the ERSSTv3 data set plus an aggressively smoothed version compared to the "raw" Indian Ocean 0-700 meter vertical temperature anomaly.  Raw is in scare quotes because there is not such thing as manageable raw data in climate.  Every data collection method involves some form of natural and decision based smoothing.  Since is statistics you should never determine a correlation based on "smoothed" data, there is actually no "proper" way to determine any correlation in mixed climate related times series.  That means that you have to "play" with smoothing choices based on your best understanding of the situation.  Different people have different "understandings" of the situation which would bias their choices in smoothing. 

Since the majority of the "raw" data using is seasonally adjusted temperature anomaly you open another can of worms with "unbelievable" confidence intervals.  With anomaly the actual deviation is not very sensitive to the baseline period selected to create the anomaly nor is the trend, but absolute value that the anomaly represents can have significant difference dependent on the range of absolute values being averaged to create the anomaly series.  Since you are really considered with the energy not so much the temperature that should represent that energy, the F to T^4 relationship severely limits the range that can confidently be assume to have "negligible" error.

This is also aggressively smoothed zonal SST data where the original data is actual temperature with the seasonal cycle intact. 

This is what the "raw" data looks like without "all" of the aggressive smoothing.  Notice the 5S-5N series in darker green.  That is the big kahuna with 5N-15N and 5S-15S fighting for the position.  Since the "raw" data is monthly, it has already been smoothed to some extent.  Even the daily data is smoothed.  Once you get to a number of samples per day or hourly you are actually getting to raw data which quickly becomes completely unmanageable.  In this chart the smoothing is one 27 month moving average "selected" because there "appears" to be a recurrent ~27 month cycle.  I made a choice.  That costs one degree of freedom.  I "selected" the data series and width of the zonal bands.  There goes two more degrees of freedom.  Since the data was "smoothed" by others, there is a degree of freedom or two that has to be considered there.  Then if I compare this data to another data set I should have to consider all the degrees of freedom of that data set plus one because I chose that data set.  So now that I am up to about five degrees of freedom, I should be able to find just about anything I like. 

I am in a pickle.

Now we have the data and we have to make choices so how do we avoid fooling ourselves?  I think with lots of comparisons and lots of humility.  No matter what choices you make there will always be someone that perceives your choices are biased, because they are.  That is unavoidable.  So there becomes a battle for "consistency". 

I may for example compare any or all of those SST regions to a Paleo reconstruction that has its own natural and collector based smoothing.  Ocean core samples build over many thousands of years and selecting a high or low frequency reconstruction.  I compare a low frequency paleo series with higher but not highest frequency instrumental I get one correlation and then any smoothing of the instrumental will improve the correlation.  William Briggs has an excellent post on that pitfall.  But even knowing the pitfall, some smoothing can be helpful if properly noted and considered.

So I think there should be a better "degree of bias" that encompasses the combined degrees of known and unknown freedom in the statistical food chain. 

Just having a brain fart.

Tuesday, November 12, 2013

The Chicken or the Egg - Which Came First?

It is really fun watching different folks try to use "Global" data to try and figure out things like the "Effective Diffusivity" of the oceans to an atmospheric "forcing".  Dr. Roy Spencer has a new paper out and the regular denizens are attempting to rip it to shreds.  The paper uses a of course and compares the rate of warming of the oceans at different depths.  This is a nice approach but what exactly is causing what?

Since I am more into the internal oscillations than the "forcing" problem.  I thought I would take my usual contrarian view and see if there are some more fun twists that "forcing" and having the oceans respond and unbelievable fast rates despite their huge heat capacity and mass advantage.  One could think that possibly the atmospheric forcings are responding to the oceans. Most of the energy in the atmosphere is provided by the oceans and the majority of greenhouse gas in the atmosphere, H2O is also a product of the oceans.  This is of course bassakwards thinking relative to a bassackwards theory.

I am using the Climate Explorer version of the ERSSTv3 data since it provides simpler masking of the regions I like to look at.  Since I am not fond of the volcanic impact on seasonal issues my baseline is 2000 to 2012 which is relatively volcano impact free and since prior to nuclear submarines there was probably not much reliable data for sea surface temperatures above the polar circles in winter, I am limiting my masks to 65 degrees.

those familiar with Dr. Vaughan Pratt's AGU poster or his guest post on Climate Etc. "Multi-Decadal Climate to a milliKelvin", might recognize the Sawtooth shape of the 45N to 65N SST anomaly.  Of course to be an exact match I would have to creatively smooth the data to knock off some of the rough edges.  According to this there was a huge 45N-65N impact in the early 1900s that was felt "Globally" but the 45S-65S may have had a delayed response to part of that impact by 20 years or so.  The 45N-65N region rebounded quickly with the 45S-45N region taking 20 to 30 years to recover.  The oceans have a huge thermal mass which would take a while to "recharge".

Just to show that there are no Rabetts up my sleeves hear it the data in degrees C with no adjustments or raw plus a smoothed version overlay.

Based on the ERSSTv3 SST data in C degrees, the 45N to 65N region has almost fully recovered from the volcanic activity and lower solar irradiance between 1890 and 1918.  With the recovery of the SST, water vapor should increase with SST and since the atmosphere responses to "any" forcing, even the removal of negative forcing, the "global" surface temperatures which are amplified by the asymmetrical land/ocean distribution would rise.

Since the oceans from 45N to 65S respond slower due to more efficient mixing and the THC which force feeds the 45N-65N oceans, their recovery would be slower.  If you ASSUME that atmospheric forcing is completely caused by something other than natural recovery, then you could find yourself believing that the oceans recharge like the energizer bunny and everything but the SAW has to be due to the Great and Powerful Carbon.  Should you wonder why the SAW exists, you might find an alternate theory :)

Sunday, November 10, 2013

The Tropics and Solar

One of the problems with conventional "Forcing" logic is that all "forcings" are not created equal.  Unless you actually hunt for the correlations, you are liable to miss something significant and end up wandering behind the little animals.  Solar forcing in the tropics can have different impacts in just a few degrees latitude.  Clouds would respond more quickly in one band, the ITCZ and may not respond at all in another.  Then since there are seasonal variations and shifts in climate patterns like the Madden-Julian Oscillation and ENSO, you end up with plenty of noise to make things really interesting.  Expecting a perfect correlation would be a bit foolish.

Above are the Tropical SST bands in degree increments using a 1854-1883 baseline with no seasonal cycle removed.  I smoothed each with stacked 27 month averaging to remove most of the seasonal noise but left enough to show how the seasonal impact increase as you move away from the "Thermal" Equator.  The Thermal Equator is about 5N to 15N which shifts the "ideal" ITCZ latitudes a little northward.  Since the ITCZ can shift, mixing at the equator is not consistent so lagged responses to solar or any other forcing would not be consistent.  There is currently a fairly consistent 27 month lag between solar forcing variation and SST response in the tropics plus there is a fairly consistent 8.5 year (102 month) lag in the upper ocean mixing layer.  When you consider both lags you get the correlations between solar and the tropical SST bands.  The highest correlation is 69% in the 5N-15N Thermal Equator zone.

A 69% correlation is not all that exciting.  All It means is that I can be around 69% confident that my smoothing is in the right ballpark of the possible lags that would influence SST when forced by solar variations.  That is better than a coin toss, but not all that exciting for the average stats guy.  Given the complexity of the system, it would stimulate some interest in an engineering kind of guy.

If I just smooth the daylights out of solar using only the 27 month lag the correlation is under 60% so the additional 102 months lag combined with the 27 month lag is moving in the right direction at least.

The dTSI I am using it the G. Kopp reconstruction up to the satellite era with the SORCE TSI composite filling in the end of the reconstruction.  Since there is quite a bit a variability some smoothing should be used, but it is easy to go overboard.

The 27 month smoothing I used is based on the obvious at least to me 27 month lag in the ENSO and tropics from other trial fits. 

While the long term correlation is interesting, using a running 50 year correlation produces "Climate Art" more than anything concrete :)

This is correlation to the a less smoothed absolute temperature of the bands not anomaly which results in the larger seasonal fluctuations.  Not to worry.

The correlation between the SST bands and GISTEMP is not a lot better until the more modern data :)

Most of the issues would be due to volcanic and all the other climate influences which will require a little more sophistication than approximating lags in a complex system.

65S-55S 55S-45S 45S-35S 35S-25S 25S-15S 15S-5S 5S-5N 5N-15N 15N-25N 25N-35N 35N-45N 45N-55N 55N-65N
0.73 0.71 0.76 0.78 0.79 0.81 0.82 -0.64 -0.76 -0.73 -0.71 -0.70 -0.64

 The full GISSTEMP period correlation is somewhat better with the equator and SH tropics having respectable correlations to global temperature.  The inverse correlation in the NH is also interesting :)

Saturday, November 9, 2013

It is Still the Sun Stupid.

Just a quick post on the silliness about the Sun not having ANY (less than a tenth of a degree) impact on climate.  That is just an indication of how ridiculous people ca be once obsessed with pet theories.

I have shown this simple comparison a few times.  It is just tropical "SST" and the satellite era solar TSI composite from SORCE.  Most rational folks would look at that and say HUH?, looks like solar might just have some impact on climate, DUH. 

Then when you compare the TSI reconstructions with the tropical "SST" you might say, HUH, except for that crazy shift around 1940, looks like solar might have some impact on climate DUH!

Well, the "SST" data isn't SST data.  It is a mixture of measurements of the ocean bulk mixing layer from just under the surface to as much a 10 meters below the surface.  "SST" would not imediately indicate changes in temperature due to seasonal and longer solar variations since the SST north and south of the equator is isolated by Coriolis effects.  It looks to take about 27 to 30 months for the equator temperatures to equalize which just happens to be about the same frequency as the Quasi-Biennial Oscillation (QBO).

So what about that weird down in 1940 and up around 1980?  Do ya think that it might take longer for other areas to equalize thermal imbalances?

Most certainly it does.  The 30 north and south latitude bands are the "ideal" location for the Hadley and Ferrel cell convergence.  The Hadley, Ferrel and Polar cells drive most of the poleward heat transfer and exist because of Coriolis Effects and temperature gradients.  Hemispheric thermal imbalances cannot exist forever so the cell dynamics shift gears from time to time to restore some reasonable level of hemispheric thermal "equilibrium".  All of this is driven by energy from the sun and rotational energy of the Earth spinning quietly in space.

If you want to blame any other changes on CO2, land use or unicorn farts, you need to figure out the basic stuff first.  Have fun. 

Thursday, November 7, 2013

SST versus OHC - Some Explanations on Apples and Oranges

The objective of using volume based SST instead of area is to reduce some of the noise and to have a better comparison to Ocean Heat Content and paleo Bottom/middle Water Temperatures.  So the volume based total of the masked major basin areas will not be a perfect match for "normal" SST for a variety of reasons.

This compares the Masked volume temperature anomaly to the HadIsst 65S-65N "Global" temperature.  There is an average ~0.1C of residual "error" between the masked and "global".  There is virtually no trend in the error which is a good thing, and the "average" error is in the nominal standard error of the data.  Considering my masking is less than perfect, this is about what I would expect.  I can later fine tune the masking if need be, but for my purposes this is fine for now.

There is not a huge difference between the area and volume calculated basin temperatures.

The average difference is about 0.01C and there is a trend due to seasonal cycle baseline more than likely, which is amplified a little by the different basin weighting.

With the volume weighting for just the five major basins I should be using only 87.1% of the total ocean volume.  Since I am using 65S-65N instead of 60S-60N I am actually using more, which is part of the larger error in the comparison first comparison.  I am not concerned with that error because it should show up as more "anthropogenic" warming/cooling or uncertainty in regions that don't have stellar data to begin with.  The "rough" estimate of OHC is not going to be all that accurate anyway and is mainly going to be compared to paleo which has a much larger uncertainty range.

When I compare with Hadsst3 or ERSSTv3b, the error is larger and has a slope.

Since the early part of ERSSTv3b has a larger error margin and the seasonal cycle selection by ERSSTv3b includes a period with higher volcanic activity, this is pretty much to be expected and highlights the reason for the 2000-2012 baseline. 

Since I an not particularly interested in shorter term noise this chart compares the individual basin volume contributions to the estimated "Global" ocean temperatures to the GISS 65S-65N 250km interpolation "Global" surface temperature using 11 year trailing averages.  The ultra-long range interpolations to eke out polar noise just adds useless complexity to an already complex problem.  The actual liquid portions of the globe oceans should be a much more reliable reference for total energy.

So just for early warming, the "temperature" anomaly I am using should not be directly compared to "Global" anomalies without consideration of the differences. 

Tuesday, November 5, 2013

SST versus OHC - Simple Correlations

In the first part of this series I actually broke down and dug out the data I needed instead of waiting on Climate Science to catch up.  Using the KNMI Climate Explorer website (h/t to Bob Tisdale) you can mask regions of most of the popular databases without too much heart burn.  Based on the results of the first trial I revised my ocean basin masking a touch.  The new masking includes the regions from 65N to 65S with the standard Atlantic/Indian break at 20E and Indian/Pacific at 147E.  For the Northern Hemisphere the Atlantic/Pacific break I used is 80W and the Southern Hemisphere 70W.  The NH break could be  tightened up a touch but appears to be good enough for government work.  The Climate Explorer website is user friendly should anyone want to fine tune things.

The most radical thing I am doing is adopting a new baseline.  Since 2000 to present has the most state of the art data acquisition plus fewer perturbations, volcanoes, wars, etc. etc. that can impact climate and climate data acquisition this shorter than ideal baseline period should make up for a number of issues.  With this baseline I am removing the seasonal signal for the 13 full years and using the last year as zero.  Since I am more concerned with longer term impacts and not all the short term noise, for the graph above I used an 11 year trailing average for both the SST and the 0-700 meter vertical temperature anomaly.

While this might look a bit like a mess, it appears to be a pretty good representation of the mess.  With a little imagination you can "see" the data doing pretty much what the climate models do, diverge in a somewhat random walk the further you get away for the baseline and zero.  Unlike the climate models there are common responses to perturbations, volcanoes, ENSO etc. with differing regional response times.  These internal variation in response time is the biggest fault of the climate models.  The models are too complex and sensitive so the modelers have to slow the responses down so much they can't get enough realistic detail.  That is a bit ironic since the "purpose" of the models is to replicate the "system" response to perturbations on century and longer time scales. 

The biggest WOW in that chart should be the Indian ocean.  The tiny little Indian Ocean is the Climate big Kahuna sea surface temperature wise with a 85% correlation to GISS global land and oceans.  You could through out the majority of paleo and stick to the Indian Ocean and you would be able to amaze your Paleo friends.

Then if you blow off all of the diligent work by the Minions of the Great and Powerful Carbon you can combine the ocean basins using their percent volume instead of surface area and you have a better metric for "global" warming.  Combining both the basin SST and the NODC vertical temperature anomaly using the percentage volume, you get his correlation which is exactly what you should expect.  There is a different response curve for the surface than there is for the high heat capacity 0-700 meter volumes.  The two tend to regroup within a decade or so for larger perturbations that have asymmetrical impact.  For "Global" perturbations energy doesn't have top flow around the globe to replace one region's loss or gain, the system uniformly and efficiently deals with "Global" perturbations. 

This brings up what should be the second WOW, the North Atlantic 0-700 meter temperature/OHC.  While the Indian Ocean surface temperature has the highest correlation to GISS, the North Atlantic 0-700 meter has the highest correlation to GISS. 

This compares GISS with the NODC Indian Ocean and North Atlantic 0-700 meter Ocean Heat Content.  No scaling or smoothing just the annual data plot against independent axis.  The Indian Ocean  Thermohaline through flow is about a third of the North Atlantic's Gulf Stream.  It takes the Indian Ocean longer to recharge.  That slow recharging and less sloshing around with other basins since the North Indian Ocean is basically blocked, makes the Indian Ocean region a naturally smoothed surface air temperature reference. 

From this comparison you can easily see how well the Indian Ocean Basin SST correlates with the normal Global surface temperature.  Normal meaning 65N-60S with only the 250 meter interpolation so the anti-phase polar responses don't complete screw up the correlations.  You can also see there is little difference in the responses of the northern Pacific and Atlantic even though there is some desire by many to promote the Pacific Decadal Oscillation to some sort of magical "Climate" index.  The PDO has a huge impact on salmon fishing and US/Canadian longer term weather patterns but doesn't really "drive" global Climate. 

ENSO on the other hand should be an excellent Climate and Weather index as long as it is not detrended.  The northern and eastern Indian ocean is the hot side of ENSO which would be the better multi-decadal/century scale index while the eastern and central tropical Pacific the up to multi-decadal index.

For century scale considerations, the Indian Ocean temperature response should provide more information on the actual magnitude of volcanic and solar influence on Climate.  Looking at the response of the huge thermal mass Pacific ocean which has the worst correlation to surface air temperature tends to cause a great deal of confusion.  It responds so quickly in recovering that the longer slower THC recharging of the North Atlantic gets completely overlooked. 

The next step will be "calibrating" hemispheric volcanic forcing to the ocean basins using the volume instead of surface to make quick and dirty energy balance estimates. 

Monday, November 4, 2013

SST versus OHC - Part I the Setting Things Up

The correlation between SST and the 0-700 meter vertical temperature anomaly is pretty good in most cases.  There is a fairly larger error especially near the beginning of the record, but allowing for that matching the 0-700 meter temperature trends to SST trend should provide a fair reference for increases in basin temperatures which can be used to approximate OHC changes for the entire SST record.

Using the KNMI Climate Explorer I have rough ocean basin temperatures.  My "Pacific" includes the most of the Caribbean and the Gulf of Mexico which I may some day revise unless someone else does before I get the chance.  If not I can rescale the basin which I may have to do anyway.  This is more just a first stab than a final result so be patient.

For area, volume and percentages I have the following adapted table:

Area+ % Ocean Volume % Ocean Avg. Depth Max Depth

(m) (m)
North Atlantic 41,490,000 11.5 146,000,000 10.9 3,519 8,486
South Atlantic 40,270,000 11.1 160,000,000 12 3,973 8,240
Indian Ocean 70,560,000 19.5 264,000,000 19.8 3,741 7,906
North Pacific 77,010,000 21.3 331,000,000 24.8 4,298 10,803#
South Pacific 84,750,000 23.4 329,000,000 24.6 3,882 10,753
Total: 361,900,000¤ 100 1,335,000,000 100 3,688 10,803

 + Boundaries between oceans vary depending upon agency, making comparisons with other published estimates difficult.
¤ Total surface area of Earth is 510,082,000 sq. km. The oceans cover ~70.9%.
* Southern Ocean area and volume calculated from ETOPO1 Bedrock version (includes Weddell and Ross seas without ice cover).
# Deepest ocean depth is in the Marianas Trench, measured at 10,911 meters. Maximum depths from ETOPO1 are not expected to exactly
 match known measured maximum depths as ETOPO1 represents average depths over ~4 sq. km areas.
Eakins, B.W. and G.F. Sharman, Volumes of the World's Oceans from ETOPO1, NOAA National Geophysical Data Center, Boulder, CO, 2010.

Scaling of the SST by basin to the NOAA NODC Vertically Averaged Temperature Anomaly is simply matching the linear regressions for the SST with the VAT for the overlap period from 1955 to 2012 then the scaled regression for the full 1870 to 2013 SST period can be used to approximate changes in 0-700 meter temperature and then 0-700 meter OHC.

Since the Pacific is nearly 50% of the global ocean volume, the comparison would indicate that at least half of the global oceans haven't had an "unprecedented" CO2 induced warming pulse.  The warming would appear to be mainly in the North Pacific which is nearly 25% of the global oceans.

The Indian Ocean does show some acceleration for that 19.8% of the global ocean volume.

 In the Atlantic Ocean only the North Atlantic shows an acceleration for that 10.9% of the global oceans.  So for Anthropogenic Global Ocean impact you have acceleration mainly in two basins with a third trying to kick in some more.  The North Atlantic Basin should be one that would show the most land use/snow field reduction impact since most of the world's river flow volume ends up in the northern Atlantic.

Using surface temperature as a proxy of sorts for deep ocean temperature is far from perfect.  Since most of the deep ocean warming is due to mixing which reduces surface temperature, there will be plenty of shorter term fluctuation that are out of phase.  When the deeper ocean cools, that energy can be transferred to the surface somewhat bring the two in phase.  So how much useful information may be gained is a total crap shoot, but it is a fun comparison.


S. Atl N. Atl S. Pac N. Pac S. India N. Indian Average
%Global 12.00% 10.90% 23.40% 21.30% 9.75% 9.75%
1871-2012 0.25 0.29 0.06 0.22 0.11 0.08 0.17
1956-2012 0.29 0.66 0.12 0.24 0.21 0.21 0.29
Ratio 1.19 2.26 1.98 1.08 1.96 2.44 1.71
kg 2.82E+19 2.90E+19 5.93E+19 5.39E+19 2.47E+19 2.47E+19 3.66E+19
J/K 1.13E+23 1.16E+23 2.37E+23 2.16E+23 9.88E+22 9.88E+22 1.47E+23

Instead of just scaling the OHC I thought it would be interesting to actually calculate the rough variation.  There are a few issues with my mass calculations likely due to not including the shoreline slope since the kg appears to be a little less than a percent high compared to the Rosenthal et al. 2013 values, but pretty close for this initial run.

Starting with the Pacific, there is a lot more variability in the north which can be smoothed out, but without smoothing more that with a 5 year moving average it looks like this.

This is how it looks back to 1900.  It is kind of interesting to compare the 1900 to 1940 estimated energy regain to the current regain.  The noisy north Pacific would appear to contribute to "weather" while the stoic southern Pacific holds the climate in check.  The scaling could be tweaked but there is a lot of noise.

The Indian Ocean which has the highest correlation with GISS provides me with a little blind test.  Since the table from NOAA doesn't have the Indian separated by north and south I had to ratio the India to 18% north and 82% south to get this fit.  With that ratio there is only half of the Joules estimated with the SST as there is with the NODC OHC data.  Before checking on the actual distribution I will scale the Atlantic just to see how the coarse fits match NODC before getting too wrapped up in precision.

With the Atlantic the coarse fit is excellent for the north and the south has similar scaling required as the Indian Ocean. 

One of the major issues with the SST masks I used is the southern and Arctic oceans are not included.  Including the polar oceans will require a more detailed masking which can also reduce the Caribbean and Gulf of Mexico areas being in the Pacific region with the initial quick and dirty masking. 

This ends part one of this series. 

On Data Quality and the Interpretation of the Data

The "Globe" has warmed by ~0.8 C degrees since 1900.  Some portion of that "Global Warming" would be "normal" and some portion could be "abnormal"  Since 1940 when there was a peak in "Global" temperatures, there has been ~ 0.4 C of warming above that peak which is considered "abnormal" for whatever reason.  The "average" accuracy of the measurements to determine the degree of warming is ~ +/- 0.25 C degrees so since 1940 the "Globe" may have warmed by 0.15 C degrees to 0.65 C degrees. Roughly 0.5 C of the total 0.8 C of warming since 1900 is "uncertain".  One group interprets that data as definitive evidence that mankind has caused greater than 50% of the "Global Warming" however much of the "warming" is real.

Since ~1998 the rate of "Global Warming" has decreased.  The peak rate of warming was the period from 1976 to 1998 which just happens to correlate with new and improved methods of measuring "Global Warming" and combining, as in "reanalyzing" the data to reduce "uncertainty".  By creatively manipulating the data we have determined that we now have data with "almost unbelievable accuracy".

For some odd reason, "almost unbelievable" just doesn't have the same grab that it once had when your parents warned you that if something sounded too good to be true it probably was too good to be true.  Convincing people that the "new and improved" methods may not be all they are cracked up to be is not all that easy.

Since the rate of "Global Warming" decreased, "Global Ocean Heat Content" has replaced "Global Mean Surface Temperature" as the go to metric for inspiring urgent action to save the world from the Great and Power Carbon Monster.  Kevin Trenberth and company have even produced a new and improved "Reanalyzation" of the "Globe Ocean Heat Content" to stoke the fires of "mitigation now".

The OHC has been reanalyzed so well, that a recent paper comparing paleoclimate proxy derived Pacific OHC indicates that warming in that region since the start of deeper ocean data collection has been10 times (the press release for the paper said 15 times which requires some creative rounding) faster than any other period during the past 10,000 years.  That would be "almost unbelievable".

What's a guy to do?  Trusted Climate Scientists are saying we are all going to fry. 

Let's look at some of the other data.  Since the OHC data is in ocean basins both as vertical temperature anomaly and the OHC in Joules times 10^22, I need to separate the ocean surface temperature into the same or at least similar basins.  If I did that right, above are the SST anomalies for the southern oceans from latitude 60S to the equator.  Notice how the South Pacific has risen 0.4C since 1870 and appears to have a decreasing slope as it approaches present.  The Pacific Ocean is the largest so it would tend to have less impact from internal "natural" or otherwise variability.

If I look at the northern oceans I see the same start at 0.4 C below present with a more ragged by similar reduction in slope for the Pacific.  Starting around 1990 the Pacific ocean SST rise almost completely stopped and by the looks of it the Atlantic and Indian Oceans "caught up" around 1998. "Catching up" would not be an indication of Anthropogenically caused "Global Warming".

The deeper ocean data starts in 1955.  These are the vertical temperature anomalies for the northern ocean basins.  From 1955 the huge Pacific ocean 0-700 meter temperature rose by about 0.075 C degress.  There was more than three times the warming the the smallest of all ocean basins the Northern Indian Ocean.

In the southern oceans the South Pacific warmed by about the same 0.075 C degrees while the South Atlantic and Indian Oceans warmed by about twice as much. 

Prior to the more "activist" Climate Science, there was a period formerly known as the "Little Ice Age" (LIA) where "Global" temperatures where thought to be lower than today due to a variety of "natural" possible influences like volcanic activity and solar variation.  Since the more "activist" Climate Science the fka LIA and preceding "warm periods" have been renamed as anomalies and could have virtually zero impact on modern climate because the "physics" of "Global Warming" indicates that "only" CO2 and the Great and Powerful Carbon could possibly have caused so much "rapid" climate change, 0.15 C to 0.65 C, over the past 60 odd years. They are right.  Once you remove all other possibilities, that much warming, the high end at least, is "almost unbelievable".

If you look at the Pacific Ocean combined 0-700 meter virtical temperature anomaly you can see the earliest part of the data has a large error range which is to be expected given that data was collected prior to "Activist" climate science.

Post "Activist" Climate Science the Pacific Ocean Heat Content which uses the Pacific Ocean vertical temperature anomaly to calculate the increase in Joules now has better accuracy prior to 1960 than it does today.  That is not only "Almost Unbelievable" it is "Too Good To Be True".  The data was either completely botched or "inspired" by some "Activist" Climate Scientist(s) with an agenda.

With every data set indicating less than "unprecedented" warming, one has to wonder why the sudden focus on a botched data set.  Normally, someone trying to double check their work would compare as many data sets as possible to avoid an embarrassing error.  Could it be they "liked" the results so much they were willing to risk embarrassment?  Then perhaps "Shame" is too 20th Century.

Data Sources: KNMI Climate Explore and NOAA National Oceanographic Data Center

Sunday, November 3, 2013

Climate versus Weather Correlations

S. India S. Atl S. Pac N. Indian N. Atl N. Pac
1951-2012 Giss dTs 0.80 0.60 0.63 0.80 0.66 0.38
1880-2012 Giss dTs 0.81 0.74 0.70 0.79 0.67 0.55
1951-2012 GISS Gbl 0.79 0.60 0.68 0.80 0.66 0.45
1880-2012 GISS Gbl 0.86 0.79 0.76 0.83 0.73 0.65

This is a correlation table I put together as part of the on going "setting up the problem".  Finding correlations and recurrent patterns is a piece of cake.  Convincing others that they are relevant, not so much.  In order to get the data for the table I had to used the KMNI Climate Explorer site and I masked the ocean regions listed above.  I used 60 to equator for the latitude range and for the ocean breaks 70W, 147E and 20E which is not perfect but seems to be close enough.  The GISS global land and oceans and land only (dTs) are downloaded from the GISTEMP site.   The table is the results of correlation tests for the two time periods for both GISS data sets to the masked HadISST 1 SST regions bounded by my selected latitude and longitudes.

The reason for the table is making a stronger case for the Indo-Pacific Warm Pool as being a strong global climate teleconnection location.  The South and North Indian Oceans have the highest "raw" correlation to the GISS data sets.  I will leave it to others pick apart the relevance of the correlation values, but all the data sets where monthly so there is close to 1600 points for the 1880 to 2012 longest correlation.  This is the first time I have done this by ocean basins so I was a little surprised that the North Atlantic didn't do a bit better.  Remember though that the mask area only extends to 60N which likely has some impact.  Trying to splice to paleo, the higher latitudes just produce more Weather noise and I am really looking for 60 plus year Climate correlation to splice with the often 50 year and longer naturally smoothed ocean cores

With paleo splicing my intent is to use these regions smoothed to whatever paleo sample time frame to make more use of the lower frequency paleo which appears to have more reliable information than many of the attempts at high frequency resolution. 

I just posted this for guys that have been picking up on the Oppo et al. 2009 Indo-Pacific Warm Pool reconstruction which is nicely interpolated into 10 year bins just for paleo to instrumental spicing.