Dr. Curry Warms the Southern Ocean

The climate science community is in spectacular denial about what is actually occurring in Antarctica. It is a classic case of “The Emperor’s New Clothes.”
They pile the BS on deeper and deeper with every new theory. Ozone, Grace, blue maps turning magically red, penguins drowning, ice sheets collapsing. It is a three ring circus.

A veritable Cirque du Soleil of contortions on a minimalist data set. But why oh why do these people engage in such diversions, maybe for the grants?

So, more ice around the Antarctic means warmer southern oceans. Heck, I didn’t need them to do a study for me to know their position. ‘Its colder now, but the trend, we’ll invent one and ……..blah, blah, blah’. Dr. Curry remains a mystery to me. At times, her statements seem to be wavering towards being skeptical. Then she puts a study such as this in.
Willis raises the point, “So if I had been a referee on the paper my first question would be, why do the authors think that any analysis based on that HadISST data from 1950 to 1999 has any meaning at all?”
Looking at Figure 2, 1979-1999 certainly doesn’t have any meaning relating to 1870-1978 and especially at the 60-70 parallels. How someone can contrive any special meaning contextualizing in this manner is puzzling to me.
Surely there was something of more substance than this in the paper?

Willis:
Have you shared this with Judy? It seems so blindingly obvious she must have some reason for spending time on the paper. It sure doesn’t explain the paradox of growing Antarctic sea ice to me.

it seems that there is a dire shortage of big men in climate science today.

that’s because “climate science” is simply a marketing machine for the global warming industry

Willis,
Just an initial impression without having read the paper – but the first 4 diagrams include 2 model simulations.
The implication seems to be that the models mimicking purely natural forcings predicted cooling whereas the actual measurements where roughly zero to very slight warming.
Hence the idea that something other than nature has “increased” the temperature in recent times [ie prevented the decrease that aught to have occurred].
A huge amount of current science seems to hang on these models – as if they are real data.

I have the impression that all those National these that and the other that are controlled directly by government have gotten a directive of : ” show up the heat” otherwise the money will run out?
In Greece we have a National Meteorological Office , and a National Observatory of Athens. They are independent in their measurements and the latter is academically controlled. The observatory predictions for the Athens region are two to three degrees lower than the Meteorological Office, but it is the latter that is reported on television.
Example: yesterday they were predicting in Athens temperatures up to 41 degrees and national measures were taken with directions for public air conditioned locations for the indigent. Well, one can go and look at the 24 hour direct measurements of the Observatory, at http://www.meteo.gr/Gmap.asp , and it nowhere went over 38 even right in the center. Maybe they are predicting car temperatures sitting in the sun.
The 40 was reached in Larissa only, as far as I checked, an inland town in the middle of a fertile plain, which cools nicely at night too.
It is evident that the people controlling information want to make the public believe it is hotter than it is.
Maybe it is the same all over the world.

So this is what J. Curry does in her spare time. It’s wonderful how a warmist can unhinge his brain to swallow a contradiction larger than his head.

AGW, not Climate Change, always was and always will be a Government sponsored fraud. The term Climate Change is a bob both ways.
I would be asking JC to show us both hands. Me thinks there will be a little finger missing. This ‘paper’ appears to be a ‘sorry’ note. As in an apology.
regards

Hey, you could be nicer to Judith Curry – she’s one scientist who tries to engage courteously with the community, sceptics and warmers alike, in a civil fashion.
GIGO seems far too dismissive a conclusion – perhaps a question to Judith inviting her to explain might be in order.

Tortured data confesses. Read all about it! Whoever said that climate was a Cinderella science which was taken up by second rate scientists (like James H) got it about right. Present company excepted of course.

I can’t make any sense of this without reading the full paper. It just does not look right as Willis had pointed out. I think it fair to say the EOF mode has some problems, not the lest of which is a lack of data. These models often look straight forward but like all these thing the devil is in the details. I think one needs to be paying very close attention to the currents in the area too. When you try and relate two different fluid dynamic systems, neither of which is fully understood, together the results can be anything but accurate.

chris1958 says:
August 17, 2010 at 10:11 pm
Then Dr. Curry should respond by showing a cooling model, and it’s hindcast/forecast.
The problem is, nobody wants to look at the other possibility, i.e. – global cooling causes cooling.
Reminds me way too much of the economic bubbles: so few were daring enough to challenge, so many were swept away in the frenzy of the day.
Where are the cooling models?

I think it would be courteous to ask her first.
Responding to:
trbixler says:
August 17, 2010 at 8:34 pm
Judith has been more open minded than many others. But the underlying motive many have for defending AGW has most often seemed to be ideological, either political (collectivist/socialist/communist) or religious (~Pantheism) or both. But there are rent seekers as well. Missing is objectivity, critical to scientific discovery.

I look forward to the mechanism that turns precipitation into sea ice.

Dr Curry, bringing a climate vindaloo to a Southern Ocean data table near you.
:0/

Dennis Nikols, P. Geol. says:
August 17, 2010 at 10:29 pm
“…I think it fair to say the EOF mode has some problems, not the lest of which is a lack of data. These models often look straight forward but like all these thing the devil is in the details. I think one needs to be paying very close attention to the currents in the area too. When you try and relate two different fluid dynamic systems, neither of which is fully understood, together the results can be anything but accurate.”
You are right Dennis, but it’s worst than that. The sea surface is a turbulent system, driven by the wind above and the currents below. It does not have an homogeneous temperature over large areas, rather it is fractal with pockets of warm water in areas of cooler water at all scales. Many many temperature measurements are needed to even get within an accuracy of +/-1K, which clearly we have not got.
Much of the cargo cult science behind CAGW depends on ignoring deterministic chaos and using meaningless linear trends which are cherry picked to provide the expected result.
As this paper from Dr. Curry shows, confirmation bias and the need for continuing government financial support, provides some amazingly wrong conclusions/results.

Willis Eschenbach: August 18, 2010 at 12:16 am
Does anyone truly think that we can estimate the temperature of that huge swath of ocean to a ±0.06°C accuracy by taking a couple of temperature readings per month over thirty million square miles of ocean, bunched in the summer months, with each reading being taken in a very different place all around Antarctica?
Figure 1 (b)
http://www.sciencemag.org/content/vol295/issue5558/images/large/se0720194001.jpeg
shows almost no shipboard readings in the area between NZ and Chile, and the majority of the readings from the ’60s and ’70s are concentrated off the east coast of Australia, Argentina, and the Cape of Good Hope. Trying to construct an accurate picture of temperature over the entire area using those scattershot data points would seem to be — *frustrating*…

richard telford: August 18, 2010 at 1:06 am
Its an easy analysis to do, probably less than 50 lines of R code, and could prove your point. But perhaps you prefer to wallow in your logical fallacy.
Where is the logical fallacy in saying you can’t accurately estimate the temperature of thirty million square miles of ocean to ± 0.06°C when you have measurements for less than half the area, and most of those were taken during the summertime?

Finally, I am not arguing from incredulity. I am applying the reasonableness test.
——–
A rose by any other name …

thompson Denis et al 2009 GEOPHYSICAL RESEARCH LETTERS, VOL. 36, L20704, doi:10.1029/2009GL040104, 2009
This (paywalled) paper also pretends to find warming in Antarctica. The main line is that Gomez dome at the base of the antarctic peninsula shows similar temp rise to a station at the tip of the peninsula where climate is clearly determined by surrounding ocean (so they report it is continental temp rise) .
The principal components extracted from this data shows a 20th c. warming that has already peaked around yr 2000. Thus far the science seems reasonable although they do not even mention the fact the rise appears to have peaked.
They then venture in the computer model fairly land. They compare the rise they find with computer simulations for an earlier period where no data is available and conclude the because the _selected_ 50 year periods do not show similar trends in the models the rise is UNPRECEDENTED and there is a need for more research .
There is no analysis of whether differently selected 50y periods show a rise or, more importantly, whether the models they used AGREED with their core samples for the period where they do have data.
Liz Thompson of BAS is a competent scientist and on the face of it the core data analysis seems sound. However, it seems that they seem to think they have to put the “unprecedented” spin in the conclusion and abstract to get it published in GRL and get funding for next year.
Of course this paper has already been cited by Met office as proof that the warming has been recorded on all continents of the world including Antarctica.
This, plus the Curry paper, will be used in the next IPCC report to “prove” continued unprecedented warming.
Maybe we are witnessing the emergence of womann-made GW.

I notice on the lattitude band graph, that all lattitudes show a upward blip post 1940. So is this raw data or homogenised, ie has it had the “GISS correction” applied to it as they did with other temperature data, if so it looks like it created a blip where none actually existed.
I also agree with Willis re quantity of data, not much happened down that way prior to the IGY in 1957 and even then shipping only went south in summer months so what periods of the year were covered and what was the quality /quantity of data that was relied on for this study?

It is certainly reasonable to question the accuracy of SST temperature readings over the entire dataset. Yet, even forgetting that, there appears to be a high in temperature readings around 1985. So that would give 25 years of southern ocean cooling during the peek rise of CO2. Where is the heat to assume an increase, or am I looking at the graphic all wrong?

Willis:
I believe that it does us no favours to name call or otherwise insult/belittle those that hold views contrary to our own. It takes maturity to rise above this. Arguments are always best kept cold and dispassionate and such stance assists analysis of the issues raised.
Several people have suggested that Judith should be asked to comment upon the points raised by you and to set out her interpretation on the data sets you refer to and to explain the reasons behind her interpretation. I concur with this.
Willis you say that you hope that Judith will “turn up” and comment. I consider that it would be much better to provide her with your comments and data set and request her comments. If she either refuses or simply decides not to avail herself of the opportunity, this would speak volumes. It would enable you to say that I have raised these points but she has been unwilling or unable to explain her position thereby inferring that she has no good answer to the points raised.
I consider that the best way forward in this ‘battle’ is to engage in a courteous and constructive manner with ‘opponents’ and I would ask you to consider contacting Judith in a courteous manner specifically asking her for her comments and further thoughts etc.
I consider that peer review should always contain the comments of those who hold a contrarian view so that always both sides of the argument are present thereby assisting the reader to make up thier own mind what merit there may be in any study/paper.

Interesting paper released by the Australian Academy of Science leaves little room for doubt – perhaps they should have read some of the comments above(?)
http://www.science.org.au/policy/climatechange2010/index.html

richard telford: August 18, 2010 at 2:01 am
“Finally, I am not arguing from incredulity. I am applying the reasonableness test.”
——–
A rose by any other name …
So tell us, richard,
1 how can one *reasonably* claim to get a ± 0.06°C accuracy from an instrument (a water-temperature gauge in a ship’s boiler intake line) that can only be read to the nearest .5°C and may have an instrument error of another whole degree C, and
2. how can one *reasonably* claim to have measured the decadal temperature change in a standard 5° x 5° grid *at all* without ever having taken one single measurement in it?

Willis:
Judy has been very courteous and has responded to any email I have sent her. I fear that we are so used to the discourteousness and dismissiveness of folks at RC that we sometimes forget ourselves. You now have an established track record of finding interesting issues so I would think that academics like Judy would be more than willing to discuss and clarify.

Kwik said “I don’t understand.
Do they really expect 40 degrees increase in tempeature in Antarctic?”
Sure they do. There on the southern peninsular. Right next to the runway and the furnace outlet. And right next to the thermometers.

Willis Eschenbach says:
August 17, 2010 at 11:02 pm
I see your point about the southern oceans manifested in an SST temp series. They are mostly copied according to specific day/mo/ from one year to the next. Only near S. America, Australia and Africa do they change. For most of the Antarctic extent, they do not change as well. A vast area of Earth either does not have any temperature data, or is remarkably stable over decade time frames.

Looking at the temperature data, for several of the bands the 2010 temperature is nearly the same as the 1870 temperature. Even if all of the temperature data were valid, where is the warming, accelerated or not?

We’ve failed to take these matters up with Dr Jiping Liu, the individual that perhaps should be asked to respond to our questions and Willis’ comments first. Ever since the news broke of this GT study we’ve been addressing Dr Curry, however, Dr Liu would seem to be first chair on this study.
From his website –
_____________________
http://shadow.eas.gatech.edu/~jliu/v0/jliu_research.html
Climate Modeling
Model Development
“Recently, the implementation of a sea ice component that accurately simulates sea ice mass balance, ice extent, interfacial fluxes, and associated feedbacks with the atmosphere and the ocean into coupled global climate models (CGCMs) has received considerable attention in the global climate modeling community. One factor motivating such improvements is the enhanced climate sensitivity at high-latitudes, typical CGCM projections of future climate change. Another factor is the increasing observational evidence of significant recent changes in various aspects of polar climate, which require mechanistic understanding and appropriate representation in climate simulations.”
“I have been collaborating with Gavin Schmidt, David Rind, Douglas Martinson and Gary Russell to improve the NASA GISS CGCM. I have incorporated more realistic sea ice dynamics (viscous-plastic rheology) and thermodynamics (snow/ice albedo, penetration of solar radiation in sea ice, sea ice salinity budget and ice-ocean boundary formulation) and subgrid scale ocean processes (Gent & McWilliams mesoscale eddy isopycnal mixing with Visbeck scaling mixing coefficients and Wajsowicz viscosity diffusion) in the NASA GISS CGCM, leading to improved sea ice and ocean simulations.”
Model Evaluation
“I evaluated the extent to which CGCM (GISS, NCAR and GFDL) can capture observed polar-global climate teleconnections. The results have been guiding me to improve atmosphere-sea ice-ocean interactions in CGCMs and to design follow-up experiments to deduce the underlying mechanisms responsible for the polar-global climate teleconnections.”
“I have been participating into the Arctic Regional Climate Model Intercomparison Project (ARCMIP) and collaborating with Judith Curry to analyze, compare and evaluate the modeled cloud and radiation fields from different models utilizing in-situ measurements (i.e., SHEBA) and satellite-based products (i.e., ISCCP/FD and CASPR)”

Hey Dr.Curry, WE NEED THAT WARMTH OF YOURS DOWN HERE, PLEASE, IF YOU FIND IT, SEND IT RIGHT AWAY!

The study was published.
I don’t see any comment that Willis was sent a pre-publication copy for approval, and I doubt that his opinion mattered to them. He does not have a union card. That might change — but I doubt it.
I don’t see how it can be rude to comment publicly on a public (published) document.
Perhaps the issue is that he has now gained some prominence in “auditing” the validity of the numerical and logical analysis. When people become concerned about what he might have to say then they will send him courtesy copies in advance of publication and solicit his comments. To do so, however, would recognize his membership in the club or on “The Team” as it were.
Think about it. 🙂

I like the phrase argument from incredulity, because that’s my first reaction to almost everything that the alarmist climatologists claim to have measured. How can anyone believe that what alarmists claim to be true can be estimated accurately with scattered local measurements at a few convenient spots? The most believable number they have is probably the amount of CO2 produced by fossil fuels, and that only because it comes from economic statistics. But even there, just because something’s been successfully sold doesn’t mean it’s been successfully burned — a very poorly known fraction of the stuff is spilled or allowed to evaporate, and what person or organization is likes to confess to being sloppy and wasteful? Everything else is randomly (in the sense that we cannot mathematically predict where it must be) and fractally distributed over the earth’s surface. Take, as just one example, the estimates of how much CO2 enters and leaves the earth’s atmosphere over the course of a year. Start with CO2 absorption by living organisms. You have smallish areas where relatively large amounts of absorption per acre occurs (e.g. jungles) and very large areas where relatively little absorption per acre occurs (e.g. blue-water sea surface). Can you, for that latter type of surface, multiply that very large area in acres by the small and inaccurately known average absorption per acre and get any sort of trustworthy estimate of the total CO2 absorption? Remember, that small and hard to measure average CO2 absorption per acre has to be well known for a very large area before the small and large numbers can be multiplied together to get an answer worth taking seriously. And of course the CO2 metabolism of living organisms changes with the weather, among many other things, so the fractal distribution is changing fractally with time — and that source of uncertainty exists everywhere living organisms grow. The amount of CO2 absorbed or released by the earth’s oceans through gas exchange is supposed to be known because we know the composition of sea water and the relevant temperatures. Really? Can calm beakers of sea water in a scientist’s laboratory, held at a steady temperature, be used to predict the gas-exchange behavior of the earth’s oceans — oceans that heat up non-uniformly during the day and cool non-uniformly during the night, oceans that are criss-crossed by everlasting wind and waves and at irregular intervals are whipped to a froth by passing storms? Rain falls from clouds, picking up CO2 on the way down and depositing it, dissolved, into the ocean. Can you imagine trying to estimate accurately the amount of rain falling into the earth’s oceans? Please don’t make me laugh! The amount of CO2 produced by natural sources (e.g. volcanos above and below the sea) is also randomly and fractally distributed in a way that changes randomly and fractally with time. It’s a good thing we have such accurate knowledge about all the sources of vulcanism under the earth’s oceans. Oh, you don’t thing so? Neither do I. It is no accident that Steve McIntyre led the charge against bogus climate science. With his background in mining he knows from case study after case study how difficult it is to determine the true distribution of an economically valuable mineral before actually constructing the mine. Mining companies routinely drill hundreds of small, narrow, relatively cheap boreholes in the land they intend to mine, then look at the concentrations of the mineral of interest inside the boreholes, and from that try to decide whether the proposed mine makes economic sense. This is over areas of only a few hundred to a few thousand acres as opposed to the entire surface of the earth. The borehole sampling would be analogous to surface measurements of CO2 absorption or temperature by little local gas-meters or thermometers — if the CO2 absorption or the temperature, like the minerals under the ground, were not changing with time. And yet these mining companies, with all the economic incentive they have to get the answer right, time and again are surprised when they go ahead and build the mine, finding much more or much less than expected. The random, fractal nature of the mineral distribution defeats their best efforts to discover through little, local measurements the true amount of the mineral present under those few hundred or few thousand acres. Climatologists never do the equivalent of constructing a mine to check out their estimates and, as mentioned, what they try to estimate often changes randomly, rapidly, and fractally with time. Yes, I am incredulous that alarmist climatologists know with any significant accuracy what they claim to have estimated.

Southern Ocean Warming Issue
In this study, we only examined the dominant SST variability in the Southern Ocean using the Empirical Orthogonal Function (EOF) analysis. We performed the EOF analysis on the area-weighted annual-mean observed SST south of 40 °S for 1950–1999. The leading EOF modes of HadISST and ERSST show positive values extending from middle to high latitudes of the Southern Ocean, and negative values in the Antarctic and sub-Antarctic zones. The corresponding principle components (time series) show a substantial upward trend (statistically significant at the 99% confidence level). Thus, the observed SST pattern in the Southern Ocean during the second half of the 20th century is dominated by a broad-scale warming that accounts for one third of the total variance (28% for HadISST and 29% for ERSST).

The leading SST EOF modes show that the strongest warming is in the middle latitudes of the Southern Ocean, and the warming is reduced poleward.

We confine the analysis to the period 1950–1999 due to large uncertainty associated with sparse in situ SST samplings in the Southern Hemisphere for the first half of the 20th century (Smith TM, Reynolds RW, 2003). Also, in this study, we identify the dominant coupled pattern between the ERA40’s precipitation minus evaporation(which starts from 1950s) and HadISST SST south of 40 °S.

Yes, I also noted the rather unusual colour sequence in the orthogonal whats-it plots.
The pink of light red would not nomally be between blue and yellow. This inversion has the effect of making the contour closest to zero (0-0.1) look “slightly hot” and lots of pink in southern ocean in plots A and esp B .
Since the tones do not look very continuous even on the legend strip , they won’t give a correct visual on the plots.
Odd choice. But since nearly all latt. bands seem to be cooling anyway this seems like a stack of smoke and mirrors to arrive at accelerated warming.

Bill Tuttle says:
August 18, 2010 at 4:45 am
So tell us, richard,
1 how can one *reasonably* claim to get a ± 0.06°C accuracy from an instrument (a water-temperature gauge in a ship’s boiler intake line) that can only be read to the nearest .5°C and may have an instrument error of another whole degree C, and
2. how can one *reasonably* claim to have measured the decadal temperature change in a standard 5° x 5° grid *at all* without ever having taken one single measurement in it?
————–
Since no demonstration I make against the creed would ever be accepted by the true believers here, I instead offer a simple way that Eschenbach could try to falsify the claims made in the paper instead of relying on incredulity. He appears to prefer his incredulity, perhaps his readers do to, than risk finding out that his test of reasonableness is little more useful that extispicy. The other reason for not attempting the analysis myself is that Eschenbach finds so many thinks incredulous that there is not time in the day to test them all.
I will however point out that the average of several measurements is much more precise that that of the original measurements.

HOCKEY STICK ALERT
Jiping Liu, “The corresponding principle components (time series) show a substantial upward trend”
“We have got you covered …………… move away from the methodology, slowly, with your hands up “

I browsed comments and saw that no one had yet commented on the use of an “orthogonal” model. I have not read the paper, but I want to highlight how the use of an “orthogonal” model, as I understand the term, should be close to a fatal flaw.
The paper attempts to take a few data sets – the “empirical” aspect – and develop a model: how much to weight each of the sources of data when using them to predict some outcome. So, ultimately, the method has the same intention as a linear regression: figure out how to best weight two predictors to predict values of a third.
By using an “orthogonal” model, the calculations will not compensate for covariance (with a lower-case “c” so maybe I should say “covariation,” or “co-mingling” ) in the predictors. In contrast, an “oblique” model has the assumption that the various predictors are co-dependent.
For weather and climate phenomena, the various empirical sources of data have to be assumed to be inter-related, no? Temp relates to rainfall, barometric pressure relates to rainfall, and so no? No?
I know this from factor analysis in behavioral and social sciences – we might try to predict some behavior based on related predictors such as using both annual income and years of education. We know these two predictors are dependent upon each other to some degree, so we would simply develop any “factor analysis” with a model that allows the predictors to be oblique, to have some common variance.
Here is the problem – and I have read Mann 1998, and I have not yet analyzed it enough to determine how orthogonal or oblique they allowed their principle components analysis to be – but this could be an issue in Mann 1998 as well- —
With orthogonality, you force the weighting to be done with the aspects of these inter-related variables that are unique, thus magnifying the influence of a modest portion of the variance of any predictor / not taking advantage of the full variance available to predict.
I believe that, mathematically, this will end up necessarily with a result that excessively amplifies at least one predictor / contributing “empirical” variable.
These models are developed to predict the future. Just as it is a small issue to be 3 degrees off-course on a short trip but it is a huge problem on a long trip, the prediction will incerasingly be “off” as we use an erroneous model to predict the future: predict temp, rainfall, ice, whatever.
Do I have this use of the term “orthogonal” correct?
Can anyone else put this issue into clear language? I have tried to be clear, but it is not easy.

richard telford: August 18, 2010 at 7:30 am
Since no demonstration I make against the creed would ever be accepted by the true believers here, I instead offer a simple way that Eschenbach could try to falsify the claims made in the paper instead of relying on incredulity.
In other words, you can’t address my questions without looking foolish.
He appears to prefer his incredulity, perhaps his readers do to, than risk finding out that his test of reasonableness is little more useful that extispicy. The other reason for not attempting the analysis myself is that Eschenbach finds so many thinks incredulous that there is not time in the day to test them all.
Take a deep breath and try again. Your typos preclude understanding your point.
I will however point out that the average of several measurements is much more precise that that of the original measurements.
I will however point out that your statement is not invariably correct, particularly when those measurements are spread so widely apart in both time and distance, and when instrument and calibration error combined are 250% greater than your claimed accuracy.

“”” For the last half of the 20th Century, as the atmosphere warmed, the hydrological cycle accelerated and there was more precipitation in the Southern Ocean surrounding Antarctica. This increased precipitation, mostly in the form of snow, stabilized the upper ocean and insulated it from the ocean heat below. This insulating effect reduced the amount of melting occurring below the sea ice. In addition, snow has a tendency to reflect atmospheric heat away from the sea ice, which reduced melting from above. “””
Well the above was C&P directly from the Georgia Tech page.
So I read it; ten times actually; and it forms a picture in my brain. In this picture I have the Southern Ocean which is a favorite place for round the world sailors; and it consists of salty water and other stuff. So there are clouds above this salty water; and if it is cold at times it snows over the Southern Ocean and the snow falls in the salty water; and being fresh water the snow floats on the salty water. Funny thing is I have never heard of any round the world sailor who ever mentioned finding snow on the water while down there in the Southern Ocean; but I’ll Take Dr Curry’s word for it if she says there is snow on the water.
Now this snow is sitting like a blanket on top of the cold top salty water; and below that there is this warmer deeper salty water.
Suddenly out of nowhere we have some sea ice and the hot salty water down below it can’t get to the bottom of the sea ice to melt it because of the snow that is floating on top of the salty water.
Maybe I have it all wrong; perhaps the snow is on top of the sea ice rather than the salty water; and the hot ocean salty water is below the sea ice, but can’t get heat to the bottom of the sea ice to melt it, because of the snow sitting on top of the sea ice.
I’m going to go and get a cup of Capuchino; and then I’m going to come back and read that paragraph from Georgia Tech for the eleventh time; because I know there’s a trick in there somewhere !

Dr Curry and Dr Liu – Many thanks to both of you for coming to the forum, rolling up your sleves, and sitting down with the WUWT Crowd. It is perhaps a very disturbing initial experience but one that offers so much in return. There are very few places where you are welcomed with open minds, arms, and hearts and challanged and torn apart for what you say at the same time -quite like your own immediate families. Be assured that we hold great respect for each of you. Rest assured we will challange everything you say. What’s the benefit to you? Let’s call it an educational experience that very few have the guts to even think of attempting. I assure you that you will be the better for it. In giving so much of yourself you will gain so much in return.
PS: When dealing with the WUWT Crowd you will no doubt meet a few you’ll wish you hadn’t. They too teach.

Judith Curry says:
August 18, 2010 at 8:18 am
Dr. Curry, you have always been open minded and fair in your work and published papers. While some here have been derogatory towards you in light of the Eschenbach critique, many have also defended you. Anthony has a very open policy in regards to comments, so both will appear.
I hope you take into consideration that not all commenters are Anthony or Willis, and that some are going to be more ascerbic than others in their comments, but your commenting is always a welcome addition to the posts.
Thank you for clarifying your paper, and for commenting. I look forward to your posts as they seldom (if ever) seem to be defensive, and mostly are explanatory.

I can’t resist commenting on this, even though I have seen some reference to it in Anthony’s original thread:
From the GT Press Release:
“This increased precipitation, mostly in the form of snow, stabilized the upper ocean and insulated it from the ocean heat below. This insulating effect reduced the amount of melting occurring below the sea ice. In addition, snow has a tendency to reflect atmospheric heat away from the sea ice, which reduced melting from above.”
This just does not make sense to me. Does snow behave differently in Anarctica than it does in the US? In my experience, when snow falls on water, even water at 0 degrees C, the water absorbes into the snow crystal, making it ice and water, the whiteness of the snow and thus its reflective properties are nulified. Snow’s insulative properties are at least partly due to the air incorporated into the structure it forms when packed, which again gets nulified when it fills with water.

“”” Bill Tuttle says:
August 18, 2010 at 8:54 am
richard telford: August 18, 2010 at 7:30 am
Since no demonstration I make against the creed would ever be accepted by the true believers here, I instead offer a simple way that Eschenbach could try to falsify the claims made in the paper instead of relying on incredulity.
………………….
…………………..
(Richard)
I will however point out that the average of several measurements is much more precise that that of the original measurements.
(Bill)
I will however point out that your statement is not invariably correct, particularly when those measurements are spread so widely apart in both time and distance, and when instrument and calibration error combined are 250% greater than your claimed accuracy.
Well an average is just that; an average. Unless the quantity being measured is stationary; meaning that it’s real value is presumably not changing; then one would expect the value of that quantity to change from time to time, and hence from measurement to measurement. So each of those measurements if they had infinite precision would be expected to have different values. and those real quantities could be represented with high precision by the measured values, if the measurement method is accurate enough. There’s no reason to expect the average of several readings of a variable quantity to be more correct than each of those readings was at the time it was taken.
We can say; due to an argument by Galileo Galilei, that at some time between the first measurement, and the last measurement, there must be an occurrence of any possible value that lies between the extreme high and low values that were observed; and hence there must be a point at which the variable quantity had a value that exactly equals the average of the measured values.
But unfortunately; we have no idea when that occurred; so the most we can say is that at some point the value will equal the average; but there’s no reason to believe that is a better number to use, than the actual observed values were at the time they were measured.
If the quantity being measured is not a variable but some fixed value such as Boltzmann’s Constant for example; then perhaps the average of a number of measures of it could be more precise; well unless our measurement system had some systematic error that we were not aware of. But I would hardly say the average was much more precise, since if the errors were randomly distributed; then at best we would expect the error of the average to diminish by the square root of the number of measurements; assuming of course that nothing changes during that number of observations.
Statistics can give you a different answer from what you measured; it isn’t automatically a better answer.

Regardless of the paper’s insights and ultimate research value, it is legitimate to complain about the title – “Accelerated warming of the Southern Ocean and its impacts on the hydrological cycle and sea ice” – in light of Dr. Curry’s comments:

With regards to my concerns about the quality of the SST data, I have only been digging into that the last several months. The concerns are greatest for the period prior to 1950, and there is relatively little data in some parts of the Southern Ocean prior to the satellite era. The general pattern of warming in the mid latitudes and slight cooling in the high latitudes is the key issue with regards to SST. We explore how the atmospheric hydrological cycle, ocean temperatures and sea ice extent interact. We then do a “what if” experiment with scenarios of 21st century warming to consider how this interaction would change in a perturbed environment.

The two don’t match up in at least a couple of ways:
1. Differential warming, maybe; “Accelerated” warming is unproven until the data are sufficiently tested for reliability and uncertainty.
2. It should be “modeled impacts”, not just “impacts” which implies actual measurements over time rather than computer-generated “what-ifs”.
We should be patient to hear a defense of the paper, but frankly, it looks like it came out of the oven undercooked with the nutritional values improperly labeled.

Jiping Liu: Thanks for stopping by.
You wrote, “The corresponding principle components (time series) show a substantial upward trend (statistically significant at the 99% confidence level). Thus, the observed SST pattern in the Southern Ocean during the second half of the 20th century is dominated by a broad-scale warming that accounts for one third of the total variance (28% for HadISST and 29% for ERSST).”
That may be true, but the SST anomalies of the mid-to-high latitudes of the Southern Hemisphere (90S-40S), the Southern Ocean in you paper, have been declining since the mid 1990s.
http://i36.tinypic.com/am83mb.jpg
Also, since the vast majority of the Southern Ocean data is manufactured by NCDC and Hadley Centre, aren’t you really performing PC analyses on data that was created using similar statistical methods? A quick look at the spatial coverage of ICOADS data reveals that almost all of the data prior to the satellite era was manufactured. And since the ERSST.v2 dataset does not use satellite data, the NCDC is still infilling most of the data.
http://i37.tinypic.com/t8x4ox.jpg

I think we should mind our manners lest this thread degenerate into the insults, name-calling, vitriol, ad homs, etc. of RC, CP and OM. I, for, one, appreciate the willingness of Dr. Curry and Dr. Liu to make their work available and contribute to this blog. There are certainly points in the paper that need to be discussed/explained but this should be done in a civil, collegial,and objective manner.

Judith Curry: You wrote, “With regards to my concerns about the quality of the SST data, I have only been digging into that the last several months.”
In my comment above to Jiping Liu, I linked maps of ICOADS observations from 1950 to 2000 for the Southern Oceans, south of 40S, for Januarys in 1950, 1960, 1970, 1980, 1990, and 2000. The source data is basically nonexistent:
http://i37.tinypic.com/t8x4ox.jpg

Judith Curry and Jiping Liu,
Thanks for joining the discussion. I am sure that you have developed a sufficiently thick skin to deal with the less pleasant comments.
The press release is misleading. There is no “paradox” – your own data shown above and in fig 3a of the paper shows cooling near the poles, and the land data shows no warming except on the peninsula. So there is no need for contorted arguments about how warming can produce more ice!
The title of your paper is also misleading. There is no ‘acceleration’. Again, your own data, fig 2a and fig 3c in the paper, show that the warming is DEcelerating – in fact virtually flat in the last decade. The only acceleration is in the computer models that merely reflect the beliefs and assumptions of the climate scientists who designed them.
Then there is the central issue that Willis raises. Why do your graphs show warming when his, directly taken from the data, show no warming? Why is it that climate scientists are so reluctant to present the raw primary data, preferring instead to put it through some mangle like EOF, SVD or PCs? If there are holes in the data, fine, show this. It is almost as if you are deliberately trying to arouse suspicion. When will climate scientists learn that is their own behaviour that is primarily responsible for increasing levels of skepticism?

George E. Smith says:
August 18, 2010 at 9:28 am
Maybe I have it all wrong; perhaps the snow is on top of the sea ice rather than the salty water; and the hot ocean salty water is below the sea ice, but can’t get heat to the bottom of the sea ice to melt it, because of the snow sitting on top of the sea ice.
I’m going to go and get a cup of Capuchino; and then I’m going to come back and read that paragraph from Georgia Tech for the eleventh time; because I know there’s a trick in there somewhere !
______________
How about a sea ice sandwich with snow underneath to insulate from the hot water and snow on top to reflect the sun? Where the ice comes from in this scenario is a mystery. But life is like that – full of mysteries. I don’t think rereading the paragraph will help, but I could be wrong. Life is like that too.

Thank you, Dr. Curry for coming by and please know that many of us appreciate any clarification you can offer. It’s IMPERATIVE to the discussion such critique/defense. I strongly urge patience on your part and polite civility on the part of our commenters.
This isn’t yahoo.
Mark

Thanks Dr Curry for letting us have the whole paper. I hope you don’t mind that I have saved it in my Climate Folder; I’m rather choosy as to what I save in that. I also printed it out; since I like to have actual dead tree to scribble notes on. But rest assured that it won’t propagate anywhere else from me; and If I mention anything from it, I will give proper credit.
Also I liked the Abstract more than I liked the Georgia Tech Page statement; which seemed like gobbledegook to me.
I’m not overjoyed to learn that much of the paper is computer simulations; but at least I can read what you actually are presenting rather than some newspeak rendition.
And again; thank you for accommodating us in this way.
George

“…first Empirical Orthogonal Function of the temperature data…”
I get the force of Willis’ argument, but can anyone enlighten me what “EOF” means in layman’s terms?

Sounds like ocean current “weather” to me. Without knowing the heat distribution of the global ocean from bottom to top it’s difficult to say much about what’s going on system-wide.
Unless precision sea-level measurements are showing an accelerating thermal expansion then any extra warming in one volume of water must be offset by cooling in a different volume. This is really the only way we have of measuring total heat content of the global ocean.

Anthony/moderators!
I think you should consider some major snipping. Dr Curry has been courteous to those who frequent your blog and the snarkyness and name calling is inappropriate. Even the use of the term garbage WRT the data is bad form. What do people think this is: Real Climate?
REPLY: I agree, and unfortunately there have been some whom have spoiled the debate here and I have not been able to moderate as much as I’d like due to other obligations. And, unfortunately it is a catch22 post facto, going forward however, I will aggressively remove such comments or portions thereof, from both sides, that don’t advance the discussion in a courteous way. – Anthony

Now this seems like a new approach 🙂
1) Model the “natural internal variability” with a computer model.
2) Capture the actual variability as “calculated” by CRU Hadley Centre.
3) Subtract 1) from 2) and hey presto you get AGW.
So now they are proving AGW by hind casting.
Sure beats rolling the dice to make a forecast that can be shown to be wrong.
100% AGW guaranteed in every packet of AGW Pops…
Try new formulae AGW Pops so you are not left out in the cold!
Coming to your favourite MSM outlet soon!

@Willis
There’s a positive one degree step change in the raw data that happened between 1940 and 1945 then basically no change or, if anything, very slight cooling since then.
What’s up with that?
Large step changes like that are often artifacts of changes in the way data is collected and/or collated. It seems very unlikely that a billion cubic miles of water heated up by 1 degree over the course of a few years. There’s something rotten in the state of that raw data for sure.

1. richard telford says:
August 18, 2010 at 10:29 am
1. oeman50 says:
August 18, 2010 at 9:47 am
I can’t resist commenting on this, even though I have seen some reference to it in Anthony’s original thread:
From the GT Press Release:
“This increased precipitation, mostly in the form of snow, stabilized the upper ocean and insulated it from the ocean heat below. This insulating effect reduced the amount of melting occurring below the sea ice. In addition, snow has a tendency to reflect atmospheric heat away from the sea ice, which reduced melting from above.”
This just does not make sense to me. Does snow behave differently in Antarctica than it does in the US? In my experience, when snow falls on water, even water at 0 degrees C, the water absorbs into the snow crystal, making it ice and water, the whiteness of the snow and thus its reflective properties are nullified. Snow’s insulative properties are at least partly due to the air incorporated into the structure it forms when packed, which again gets nullified when it fills with water.
——————–
The paragraph you quote is perhaps not quite sufficient to explain what is going on. The argument concerns not the insulative properties of snow, but the effect of snow on salinity. Precipitation (either snow or rain) falling into the ocean will reduce the salinity at the surface, making it less dense than the bulk of the ocean below. This salinity gradient will inhibit water column mixing, and the thin fresher layer is easily cooled. So the sea ice is bathed in relatively cold water, which insulates it from the relatively warmer more saline water below.
The second effect of snow is that snow falling on the sea ice it increases the albedo of the sea ice.
——————–
I can understand the effect of snow on the albedo of ice, but that is not the issue. It is the effect of warmer water sea ice is floating on. I have looked up the thermal conductivity of pure water (0.561 W/mK) and of seawater (0.563 W/mK). This small difference in no way makes fresh water an “insulator” from sea water or makes it easier to cool, in fact it appears to be the opposite.

I think Willis hit the nail on the head saying the older raw data is worthless in regards to being representative of the entire southern ocean while the more recent, reliable raw data shows a cooling trend if any trend at all.
[snip]

I have read and re read the comments form top to bottom several times and cannot see any comments that could be termed unacceptable or rude or beyond what could be termed reasonable comment.
I have seen numerous comments thanking Judith Curry for her participation and I have seen many extremely valuable posts examining her paper, I myself have tried to keep my comments within reasonable limits as have many others.
Am I missing something here?
I see a genuine and interesting dialogue between interesting and intelligent people holding to a rational and on the whole polite discourse, if someone can can show me just what comments are rude or unacceptable or hurtful then I would be grateful.
This particular thread is IMHO one of the best I have read for some time with so many valid and interesting posts, if I were a scientist then I myself would be happy to submit my work to this kind of forum.
Many thanks to all the posters on this thread, I believe it is your participation that makes a large contribution to the popularity of this wonderful site.

I have an hour to try to catch up on this, it is difficult to find the signal amidst the noise here. Please see the plot at the top of the thread, spatial trends of SST in the Southern Ocean for the period 1950-1999. As described in the paper, there is warming in the midlatitudes, with slight cooling in the higher latitudes. Willis’ plot is on a scale where you can’t eyeball anything. But I’m glad Willis raised this issue, since the actual SST data (not just the EOF) is needed for this discussion.
oeman50 says:
“This just does not make sense to me. Does snow behave differently in Anarctica than it does in the US? In my experience, when snow falls on water, even water at 0 degrees C, the water absorbes into the snow crystal, making it ice and water, the whiteness of the snow and thus its reflective properties are nulified. Snow’s insulative properties are at least partly due to the air incorporated into the structure it forms when packed, which again gets nulified when it fills with water.”
Snow in the high southern latitudes can fall on the open water, on sea ice, and on the glacier; its fate is different in each case. When it falls on water, it immediately melts.
When it falls on the glacier, it accumulates and undergoes various metamorphic processes that increases its density, and it sticks around unless it melts during the summer. The snow that falls on sea ice will accumulate as on the glacier (or on land), but melt as part of the seasonal cycle. The more snow on the ice, the slower it is to melt.
Precipitation in the Southern Ocean can be either liquid (rain) or solid (snow). At colder temperatures, the precip is more likely to be snow than rain; as temperatures warm there is a greater likelihood of rain. The snow versus rain is seasonal, snow in the cold season and rain in the warm season, with the warm season lengthening in a warmer climate.
When precip falls on the open ocean, it doesn’t matter too much whether it is rain or snow (there is a small latent heat of melting the snow), both have the same freshening effect on ocean, which reduces the density of the upper ocean, and stabilizes so the heat below doesn’t reach the surface. Note, the antarctic sea ice is patchy, you have a bunch of ice floes surrounded by open water, with an overall ice concentration typically of 80%. When precip falls on the sea ice itself, it does matter substantially whether it is snow or rain. Snow raises the reflectivity of the ice, so it is harder for the sun to warm the sea ice. Rain falling on snow covered sea ice will actually accelerate the melting owing to increasing the density of the snow which reduces its reflectivity.

oeman50 says:
August 18, 2010 at 11:34 am
I can understand the effect of snow on the albedo of ice, but that is not the issue. It is the effect of warmer water sea ice is floating on. I have looked up the thermal conductivity of pure water (0.561 W/mK) and of seawater (0.563 W/mK). This small difference in no way makes fresh water an “insulator” from sea water or makes it easier to cool, in fact it appears to be the opposite.
—————
Conduction is slow – convection is more important. The density gradient introduced by the freshwater stabilises the water column and reduces convection.

With regards to data quality in the Southern Ocean. There is very little before 1950, which is why we did not attempt to discuss the period prior to 1950. During the satellite era, the data is relatively good, although there are still problems. Data during the 1950’s and 1960’s is relatively sparse. There are several efforts underway to get more ship data (which exists in the form of handwritten logs) into the climate data bases. Jiping Liu is working with Chinese scientists to get Chinese ship data into the archives. So I think there is hope for improving the data situation in the near future, but the Southern Ocean is definitely a data sparse region.

The 1940 period is very interesting. Globally there is a bump in surface temperatures that peaks in 1940 (land and oceans, both hemispheres). There have been attempts to explain this away by aerosol forcing, which is mostly a bogus argument IMO, particularly for the southern hemisphere. If you want to read more about challenges associated with the sea surface temperature data sets, read this summary by Rayner et al. https://abstracts.congrex.com/scripts/jmevent/abstracts/FCXNL-09A02a-1662927-1-Rayneretal_OceanObs09_draft4.pdf

I just want to echo the sentiment that we should be polite and respectful to Drs. Curry and Liu. Some of us might have reservations about this paper, but there’s no reason why we can’t be civilised about it.
Oh, and I’d still love to know, like Lucy Skywalker, what “EOF” means. I’ll bet we aren’t the only ones, and not knowing that may well be impeding the best quality discussion here. So TIA to anyone who can put me out of my misery! 🙂

Empirical Orthogonal Functions
In statistics and signal processing, the method of empirical orthogonal function (EOF) analysis is a decomposition of a signal or data set in terms of orthogonal basis functions which are determined from the data. It is the same as performing a principal components analysis on the data, except that the EOF method finds both time series and spatial patterns. The term is also interchangeable with the geographically weighted PCAs in geophysics.
see: http://en.wikipedia.org/wiki/Empirical_orthogonal_functions

The EOF mode identifies regions that are closely related and with strong gradient (spatial variability), and the PC indicates the amplitude of EOF as it varies through
time (temporal variability). More detailed information can be found at andvari.vedur.is/~folk/halldor/PICKUP/eof.ps.gz

BarryW says:
August 18, 2010 at 11:02 am
Anthony/moderators!
I think you should consider some major snipping. Dr Curry has been courteous to those who frequent your blog and the snarkyness and name calling is inappropriate. Even the use of the term garbage WRT the data is bad form. What do people think this is: Real Climate?
REPLY: I agree, and unfortunately there have been some whom have spoiled the debate here and I have not been able to moderate as much as I’d like due to other obligations. And, unfortunately it is a catch22 post facto, going forward however, I will aggressively remove such comments or portions thereof, from both sides, that don’t advance the discussion in a courteous way. – Anthony
____________________
There is great irony in this exchange. Major snipping would, in fact, make this exchange MORE like Real Climate. What, do we have to be reverential if a real climate science might be listening in? It’s high time the academics who have fostered the CAGW came forward and defended their work in the trenches. Most of us are eager to learn, eager even to be taught. But that requires a transparent and free exchange of ideas. It’s the allowance of divergent points of view that make this blog worthwhile. Take that allowance away and you have nothing.
Oddly, the only comment that seems close to inflammatory that I’ve seen on this thread is the title. It infers rather warmly that Dr. Curry cooked the books. Anything said after that has to be seen in the context of that. But even that is a worthwhile point of view and I’m sure Willis and everyone else is willing to accept persuasive alternative explanations.
I should add that for the colour challenged (8% of males) that all those coloured figures and charts are meaningless. They are plague among us.
Also for my part I’m wondering if the M&W2010 statistical analysis is applicable here, since the data points are so sparse is it not possible that nothing can really be made of them? And has a proper statistical testing been done to rule out that possibility?

With regards to the moderation, this thread is fine, it is the comments on the previous thread (Paradox of Antarctic Sea Ice) that were inappropriate. Thanks to Anthony for addressing the situation, particularly the latter half of this thread is good.

“”” Malaga View says:
August 18, 2010 at 12:52 pm
Empirical Orthogonal Functions
In statistics and signal processing, the method of empirical orthogonal function (EOF) analysis is a decomposition of a signal or data set in terms of orthogonal basis functions which are determined from the data. It is the same as performing a principal components analysis on the data, except that the EOF method finds both time series and spatial patterns. The term is also interchangeable with the geographically weighted PCAs in geophysics. “””
Izzat a ten sill-abble word for Fourier Analysis; or its equivalent. I’m fully aware of the procedure of representing some continuous function as a sum of orthogonal or ortho-normal functions; such as Sinusoids or Legendre Polynomials; Bessel Functions and the like. Is that basically what you are referring to or is this EOF something different.
It seems obvious to me that a function of multiple variables such as time and space (how about the global temperature map) can be represented by a “spectrum” of “frequencies” in each of those variables. And of course the theory of Sampled Data Systems, requires specific rules be followed to represent such continuous functions properly by a series of discrete samples ; specifically the Nyquist Criterion.
But other than these considerations; what is it that EOF has to offer; if anything. I dunno; which is why I am asking ?
George

Willis, Judith, Jiping – thanks for having the courage to hold an honest, real world related “warts and all” discussion on a public blog. IMO progress comes from such things! 🙂

Too many assumptions, along with the use of poor quality data and the fiction of computer climate models, make this paper less than convincing.
Nothing new about recent temperatures seen in the Antarctica either, as can be seen from the graph here:-
http://a.imageshack.us/img807/8771/jones7990n.jpg
As with climate everywhere, temperature in the Antarctic oscillates up and down over the years, and was doing so long before mankind started burning significant quantities of coal and oil.

I have read the paper and I have some problems:
The paper states:
“As shown in Fig. 4b, SST under sea ice in the 2090s is a few tenths degree warmer than that in the 2000s,”
Most of the sea ice lies at >65 degrees S lat. According to Willis’s graph of actual temperatures, this area is at -1 to 0dC. Looking at figure 4b, it does not seem that a “few tenths” translates into “enough tenths to put us above freezing.”
Also, according to NOAA: “Multi-year ice is less common in the Antarctic than the Arctic, and is usually confined to the western Weddell Sea and isolated embayments at other locations around the coast. Ocean currents and the atmospheric circulation result in a net divergence of sea ice around the continent, causing most of the ice to melt in the summer as it drifts into warmer waters, or as the upper ocean heats up as the open water areas within the pack absorb solar radiation…
The ablation season of Antarctic sea ice is rarely associated with the presence of melt water on the surface of the ice.”
So an increase in rainy precipitation is going to have very little effect on the sea ice, since there won’t be much sea ice. The average summer temperature in Antarctica is -6dC. Even on the balmy Peninsula, the temp is around 2dC in the summer. So, there will have to be quite a bit of warming to bring rain past the coastline. It will still be cold enough during the winter for the precipitation to fall as snow (that whole absent-sun-thing). So there will be plenty of snow insulation in the ice-growth season.
The increase in Antarctic sea ice is occuring during the SH winter. There is very little change in the sea ice extent during the melt season. see: http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/seaice.area.antarctic.png
In fact, if you look at only the last two years, minimum extent does not seem to predict maximum extent.
I also have a big problem with the snow-insulating-water idea. Unless this just got mangled in the press release. The sea ice forms from freezing sea water, not congealed snow. In the summer, it is melting ice that cools the water. Snow is not nearly dense enough to come close to the heat transfer of ice. Even rain falling in those latitudes would be PRETTY COLD WATER. And, since rain drops are also more dense than snow, I’m guessing there wouldn’t be much difference in the effect on SST.

Bill Tuttle
1 how can one *reasonably* claim to get a ± 0.06°C accuracy from an instrument (a water-temperature gauge in a ship’s boiler intake line) that can only be read to the nearest .5°C and may have an instrument error of another whole degree C,
assuming the errors are randomly distributed (which is reasonable for the problems you posit) they will balance out with more measurements and the more measurements you use the less likely the error of any individual measurement has to distort the cumulative value. a fair coin has a 100% “error rate” when tossed once (will only get within 50% of the actual random distribution of heads-tails in coin-tosses of a fair coin and will never produce that distribution on a single toss), however a trial of 10 coin flips has an “error rate” of about 15%, flip the coin a million times and the variation from the 50% heads-tails ratio is so small as to be not worth worryiong about – the coin none-the-less has a 100% error rate on one toss. this does require that the errors be randomly distributed, but that is a reasonable assumption in the case of thermometers. whether or not you can actually get a ± 0.06°C accuracy from the measurements in this paper is a different issue ( i have difficulty accepting that claim although not worth my time to test it), but trials using measurements with randomly distributed errors have smaller errors as more and more of measurements accumulate.

Well let me cast my self adrift on the thin ice (Southern Ocean style) where only fools (like me) don’t fear to tread.
For those non mathematicians who have no idea what Orthogonal Functions are; here’s my stick on a sandy beach explanation.
Everybody understands the function y = sin(x) which goes from 0 to +/-1 back and forth as (x) goes from zero to 2.pi.
Draw yourself a sin “wave” and you can immediately see that the total area for a complete cycle is exactly zero.
OK; now we could aslo have a function y = sin(2x), or y =sin (3x) and so on ad infinitum.
This set of functions; y = sin (nx) (for n = zero to infinity) is a set of orthogonal functions.
What the heck does that mean ?
If I take any one of these functions (y = sin(7x)) and multiply it by itself to get y = sin^2 (7x), and now plot a graph of that, I will discover that the total area included in one complete cycle of x = zero to 2pi is not zero but has a definite non zero value.
If I take any two different ones of these functions and multiply them; y = sin(7x).sin(13x) and integrate that, I find that the area is still zero so long as the two functions are different.
So that is the general property of “Orthogoanal Functions” The integral of the product of any two different functions of the set over some svery specific interval (0 to 2pi for the sin function), is zero, but the integral of the square of any one function of the set is finite; over that same interval of course.
There are a whole slew of mathematical functions that have this property; and any such set can be used to synthesize ANY continuous function over some interval.
The most common form of this of course is the Fourier analysis of cyclic functions into a series of “harmonics” ;but Bessel Functions, and Chebychev Polynomials (special case of sinusoids) LaGuerre Polynomials and on and on; have all been use to synthesize other functions. I’m not going to go into exactly how you figure out the correct sequence to use; but the common method relies on that orthogonal property, because I can multiply every term in my series by one of the functions and do the integrations; and every darn one of them is going to disappear, except for the one I used as my multiplier; and that will get me the correct coefficient for that term.
People who calculate fields in electron Optics and such like make a lot of use of Orthogoanl Functions. Ortho-Normal simply means that the integral comes out to something nice like 1.0 for example.
But empirical has me bamboozled so you are on your own out there now so be careful !

Tamara,
The sea surface temperature undergoes an annual cycle (warmer during the summer). During the winter, the surface temperature of the water between the ice floes stays very close to the freezing point of sea water (about -1.7C). During the summer, the ocean warms to slightly above freezing, particularly after the ice has melted (much of the energy goes into melting the ice). In a warmer climate, the winter temperatures remain the same, but the melt season is longer (and the sea surface temperatures get somewhat warmer during the summer).
Snow that falls on sea ice accumulates on the surface of the sea ice. If the snow gets very heavy, it can weight down the flow and the snow can get wet and form snow/ice. But mostly the snow sits on top of the sea ice until the snow melts in the summer. The more snow on top of the sea ice, the slower the ice will melt. Nearly all of the Antarctic sea ice melts each summer, except for that which gets trapped in various embayments.
The rainfall doesn’t do all that much to the sea ice (other than speed up the summer melt), it is mainly the absence of snow accumulation that speeds up the ice melting.

Dr Curry,
why didn’t you investigate the impact on the deep water upwelling by an increase of the wind stress?
You state that the anular mode circulation around Antarctica is stronger now than in the past. But you failed to analyse which impact that atmospheric circulation change could have on sea temperature and salinity, And this analysis is a pre-requirement before to look for some P-E change.

Willis, I don’t disagree that there are issues with the Southern Ocean data, although there is arguably sufficient data in the midlatitudes of the Southern Ocean, which is where the warming is observed. Also, climate models have been simulating a decrease in Arctic sea ice and no decrease in Antarctic sea ice for the latter half of the 20th century, in general agreement with observations.
Is this paper the “last word” on this subject? Of course not. I hope it stimulates more data set building and hypothesis testing, using data and models. This is a topic that deserves more attention.

In looking at the first two so-polar pictures, I see that the horizon seems to be in the middle of New Zealand; which would put it at about 40 deg S. I know that Wanganui is about 36-37 deg S.
And the color chart says that the “trend” is about 0-0.1 deg C per decade.
So that tells me that these plots are not SSTs; but some sort of “anomaly”
And there’s the rub.
It’s impossible to tell from anomalies, whether the snow should melt when it lands in the ocean or not.
So I guess we have to look elsewhere for SSTs to find out where the freeze line is or was.
I’m also curious; now that everyone wants to know what EOFs are what has been written as to the validity of EOF analysis; if the sampling rules for sampled data systems are violated.
It seems to me that the entire HADCrut or GISSTemp data is corrupted by aliassing noise due to gross undersampling both in time; but especially in the spatial sampling; so the Arctic, and Antarctic are not renowned for their surplus of sampling stations.
So I’m curious as to the validity of whatever EOFs purport to deduce in the Antarctic ?

Anthony,
I respect completely, on your blog, that you have decided to exercise (what appears to me) some special case of commenting style/restraint for the post of Judith Curry & Jiping Liu as opposed to many other posters who took a full spectrum of style/unrestraint commenting without complaint and with good grace.
I am confident you have your well thought-out reasons. Though, honestly, I do not understand why you do so.
I am sorry if it may be inappropriate to ask whether you will use this special approach to all future WUWT posts going forward?
Again, I appreciate the open and fair venue you have provided here.
John
REPLY: It simply got out of hand on the other thread http://wattsupwiththat.com/2010/08/16/georgia-tech-on-resolving-the-paradox-of-the-antarctic-sea-ice/
…which I’ve closed comments on, and redirected commenters here, now that we have the full paper and additional supplemental data. Sometimes it simply pays to remind people to be courteous in the heat of the moment. – Anthony

Judith Curry wrote:
“[…] midlatitudes of the Southern Ocean […]”
Dear Dr. Curry,
This could be a source of important misunderstanding, so please quantify “mid” (as used in the preceding context) as precisely as possible. Thanks if you can clarify.
Sincerely,
Paul Vaughan, M.Sc.
–
A note to “nonalarmists” who [humorously] seem alarmed by the paper:
Drawing attention to the Southern Ocean is the way to go. I wish Drs. Curry & Liu much success in raising the profile of this important neglected region. I wish Dr. Liu far more stable personal income.
–
Willis, You make several good points, but I would advise against going so far as suggesting that Southern Ocean data is “nonexistent” before the satellite era. It is, however, fair to laugh heartily at the pre-satellite HADISST “data” for 60-90S. Those folks would be well-advised to take out some of their garbage.

Dr. Curry,
With respect to a previous question I asked: I understand your arguments relative to winter antarctic ice, however the summer ice minima has risen aproximately 1.5 million sq km since 1980 from CyrosphereToday. While the area has not gone up as much, it too is showing a noticeable rise which means the ice is not melting back as your theory assumes?

“”” Pompous Git says:
August 18, 2010 at 4:02 pm
EM Smith wrote:
“Well the above was C&P directly from the Georgia Tech page.
So I read it; ten times actually; and it forms a picture in my brain. In this picture I have the Southern Ocean which is a favorite place for round the world sailors; and it consists of salty water and other stuff. So there are clouds above this salty water; and if it is cold at times it snows over the Southern Ocean and the snow falls in the salty water; and being fresh water the snow floats on the salty water. Funny thing is I have never heard of any round the world sailor who ever mentioned finding snow on the water while down there in the Southern Ocean; but I’ll Take Dr Curry’s word for it if she says there is snow on the water.” “””
Well that is totally amazing.
I could sware that I wrote those exact words just a few minutes ago; and lo and behold E.M. Smith goes and posts the exact same thing; that is truly amazing.
I think I should go out and buy a lottery ticket.

Bill Tuttle, richard telford, George E. Smith, Max:
As I just noted to Willis above…
*************
ICOADS is the source dataset for ship and buoy readings used in the Hadley Centre and NCDC SST data. The NCDC and Hadley Centre then infill the remainder of the data. I provided a link above to maps I created that show the limited observations for Januarys spaced a decade apart from 1950 to 2000. There’s nothing special about January, other than it occurs during the Southern Hemisphere summer, so one might expect some traffic occassionally at high latitudes. There is, as you can see, very little source data for the latitudes included in Liu and Curry. Here’s the link to the maps:
http://i37.tinypic.com/t8x4ox.jpg
*********
So it’s not a matter of averaging 50% of the grids. You’d be lucky if you could find data in 10% of the grids on a good month.
http://i48.tinypic.com/2i7spwy.png
And I’ll add another note, since 1982, the Hadley Centre has also used satellite data which has much better coverage from 90S-40S. But the NCDC still relies in ship and buoy data. They included satellite data in their ERSST.v3 data then deleted it a few months later due to, assumedly, political pressure, because the satellite data lowered the Global SST readings by about 0.03 deg C and changed the order of record years. More on this in my post detailing the SST datasets:
http://bobtisdale.blogspot.com/2010/07/overview-of-sea-surface-temperature.html
Regards

Judith Curry says:
August 18, 2010 at 5:29 pm
We posed a hypothesis, we tested it using data and model simulations, which support the hypothesis. If the data are bad and the model is wrong, that doesn’t falsify the hypothesis, it reduces the support for the hypothesis
=====================================================
Judith, why would you form a hypothesis based on what would happen if it got warmer in the southern ocean? and then title it “Accelerated warming”, when it’s not?

If the data are bad and the model is wrong, that doesn’t falsify the hypothesis, it reduces the support for the hypothesis. So it doesn’t make any sense to say that our hypothesis is incorrect because there are holes in the sea surface temperature data set.
I find this statement baffling. As I understand it the problem with the hypothesis is that the observations and data show the area is cooling and the ice is growing. Is this not contrary to the hypothesis and therefore falsifies it.
Forgive me if I misunderstand. I am not a scientist and any understanding I have of science comes from reading Karl Popper when I was a student.

Judith Curry: You wrote, “So it doesn’t make any sense to say that our hypothesis is incorrect because there are holes in the sea surface temperature data set.”
Just a clarification: It’s not that there are small holes in the data. The vast majority of the SST data used in your study is infilled by Hadley Centre and NCDC. Refer to the following maps created by NCAR. They illustrate the ICOADS data availability for the period of your study. They show that less than 10% of the months and grids have data available.
http://www.cgd.ucar.edu/cas/guide/Data/coads.sst.f2.html
http://www.cgd.ucar.edu/cas/guide/Data/coads.sst.f3.html
And I’ll agree, the methods used by NCDC and Hadley Centre are reasonable. But since the SST data is so sparse south of 40S…
http://i37.tinypic.com/t8x4ox.jpg
…aren’t you really performing an EOF analysis of the statistical methods NCDC and Hadley used to infill the data. In other words, if you use HADSST2 data, which does not infill all of that missing data, would you get the same results?

richard telford says:
August 18, 2010 at 7:30 am
…
I will however point out that the average of several measurements is much more precise that that of the original measurements.
Say what? You can divide a total by a count and come to an “average” with any number of digits. Thus “more precise”. But not accurate. The ACCURACY is limited by the original data. Moreover, by the LEAST ACCURATE of the original data. So “0.06°C” is nonsense right out of the box.
You can’t reduce errors by averaging them, without really grotesque assumptions about the distribution of errors (as many upside as downside, by equal amounts, etc., etc.)
The ocean measurements, and their differences, are ACCURATE to a couple of degrees C. That’s all you get, however many digits you extend your total/count long division exercise to.

The biggest thing I find troubling about this paper is the somewhat arbitrary 1999 cut-off date.
There are very few months of un-contested, accurate SST data.
The more months the better in this case.
An alternative approach in this case might be to stop in 1999 and then forecast the model to today and check the degree to which the model forecasts that period.