UPDATE: 8/18 10:30AM I spoke with Dr. Judith Curry by telephone today, and she graciously offered the link to the full paper here, and has added this graphic to help clarify the discussion. I have reformatted it to fit this presentation format (side by side rather than top-bottom) While this is a controversial issue, I ask you please treat Dr. Curry with respect in discussions since she is bending over backwards to be accommodating. – Anthony
===========================================================
[Update] My thanks to Dr. Curry for showing the graphic above, as well as for her comment below and her general honesty and willingness to engage on these and other issues. She should be a role model for AGW supporters. I agree totally with Anthony’s call for respect and politeness in our dealings with her (as well as with all other honest scientists who are brave enough to debate their ideas in the blogosphere). I also commend the other author of the study, Jiping Liu, for his comments below.
However, as my Figure 2 below clearly shows, any analysis of the HadISST data going back to 1950 is meaningless for the higher Southern latitudes. The HadISST data before about 1980 is nonexistent or badly corrupted for all latitude bands from 40°S to 70°S. As a result, although the HAdISST graphic above looks authoritative, it is just a pretty picture. There are five decades in the study (1950-1999). The first three of the decades contain badly corrupted or nonexistent data. You can’t make claims about overall trends and present authoritative looking graphics when the first three-fifths of your data is missing or useless. – willis
===========================================================
Guest Post by Willis Eschenbach
Anthony has posted here on a new paper co-authored by Judith Curry of Georgia Tech, entitled “Accelerated warming of the Southern Ocean and its impacts on the hydrological cycle and sea ice”. The Georgia Tech press release is here. Having obtained the paper courtesy of my undersea conduit (h/t to WS once again), I can now comment on the study. My first comment is, “show us the data”. Instead of data, here’s what they start with:
Kinda looks like temperature data, doesn’t it? But it is not. It is the first Empirical Orthogonal Function of the temperature data … the original caption from the paper says:
Figure 1. Spatial patterns of the first EOF mode of the area-weighted annual mean SST south of 40 °S. Observations: (A) HadISST and (B) ERSST for the period 1950–1999. Simulations of CCSM3 (Left) and GFDL-CM2.1 (Right): (C, D) 50-year PIcntrl experiment (natural forcing only),
Given the title of “Accelerated warming”, one would be forgiven for assuming that (A) represents an actual measurement of a warming Southern Ocean. I mean, most of (A) is in colors of pink, orange, or red. What’s not to like?
When I look at something like this, I first look at the data itself. Not the first EOF. The data. The paper says they are using the Hadley Centre Sea Ice and Sea Surface Temperature (HadISST) data. Here’s what that data looks like, by 5° latitude band:
Figure 2. HadISST temperature record for the Southern Ocean, by 5° latitude band. Data Source.
My first conclusion after looking at that data is that it is mostly useless prior to about 1978. Before that, the data simply doesn’t exist in much of the Southern Ocean, it has just been shown as a single representative value.
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?
Next, where is the advertised “Accelerated warming of the Southern Ocean”? If we look at the period from 1978 onwards (the only time period with reasonable data over the entire Southern Ocean), there is a slight cooling trend nearest Antarctica, and no trend in the rest of the Southern Ocean. In other words, no warming, accelerated or otherwise.
Finally, I haven’t even touched on the other part of the equation, the precipitation. If you think temperature data is lacking over the Southern Ocean, precipitation data is much worse. The various satellite products (TRMM, SSM/i, GPCC) give widely varying numbers for precipitation in that region, with no significant correlation between any pair (maximum pairwise r^2 is 0.06).
My conclusion? There is nowhere near enough Southern Ocean data on either side of the temperature/precipitation equation to draw any conclusions. In particular, we can say nothing about the period pre-1978, and various precipitation datasets are very contradictory after 1978. Garbage in, you know what comes out …
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
I will thank Jiping Liu and Ms Curry for showing up to defend their work.
I will also suggest that the major “Southern Ocean Currents” around Antarctica are famously apt to eddy, spin off independent “rings” of warmer (or even colder) current, and generally mess up the temperature at any given location. Five degrees C for the season, five more for the present current ( the current, at the present moment, I mean) and even 3 or so degrees from the north edge to the south of the same current. I’d think it very odd to attribute temperature anomalies to climate before the currents were accounted for. And I think the current is not at all easy to model.
But it’d be handy. Ben Franklin would be proud of y’all.
I think that should be the final word on the debate:
When you say garbage in garbage out, you are not understanding the scientific process. 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. 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 know I can’t think of an answer to that.
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?
“”” Bob Tisdale says:
August 18, 2010 at 5:19 pm
Bill Tuttle, richard telford, George E. Smith, Max: “””
Bob, Not quite sure why you included me in this comment. I haven’t actually got much into the data in the paper yet; other than to register my disappointment that the SST stuff was actually anomalies or trends; which I guess are one step removed from anomalies in understanding what the SST is.
I don’t have a good mental picture of what sea water temperatures typically run in the Southern Ocean so I was a bit bamboozled by whether snow landing on the water could persist.
But now that I have a better picture of the situation; that we are dealing with snow on top of already ice.
In which case I would caution that if the sun is up, then snow becomes a whole lot less reflective PDQ; as in just a few hours.
If the sun is not up (inside the Antarctic Circle) then I imagine the snow reflectance could persist at high values for some time.
And if the sun does not get up outside the Antarctic circle; then run like hell for some high ground !
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.
john b says:
August 18, 2010 at 6:50 pm
=============================
John, I’m going to propose a hypothesis that if sheep laid eggs (model) one out of every ten sheep eggs would have a dog in it (data).
Just because I’ve never found any sheep eggs, or sheep eggs with dogs in them, does not falsify the hypothesis, it reduces the support for the hypothesis
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?
OK Bob, I think I found the handle. Thanks for posting that. There’s a whole lot of holes out there in that Southern Ocean. I never quite realized just how far flung New Zealand actually was.
I’m becoming increasingly puzzled that the General Theory of Sampled Data Systems is not a part of the orientation lecture for Climate Science 101.
First we have Dr Mann boring a hole in a tree; to get a one dimensional sample of a three dimensional object; and presuming to be able to conclude what the Temperature was as represented by that hole in one.
I paid $75 to get a legal 8 x 10 black and white drawing of a very famous Larsen “Far Side” Panel.
I believe its official title is “What’s in My Yard.” The punch line (in part) says…. “Say Emily, could you go to your window and describe what’s in my front yard ?”
In the panel, Gladys is on the phone in her front living room; and her front window is entirely blocked out with an enormous eye jammed up against the glass looking in at her.
Ever since I first saw that panel, I have considered it the very best illustration of the failure to observe the Nyquist Criterion, when dealing with sampled data.
Unfortunately it is never legal under any circumstance no matter what, to use any “Far Side” panel in a lecture, or slide presentation or on the web; so I can’t post it here. Gary Larson simply does not allow ANY usage of his works; but you can legally buy any one of his products and get a legal print of any panel for your own personal gratification.
I believe there is some sort of general theorem of Pattern Recognition; which essentially says that pattern recognition is impossible. More specifically it says something like:- Given any finite number of (n-1) dimensional projections of an (n) dimensional object; it is always possible to construct a counterfeit ( n-dimensional) object which is different from the subject object; but which can project the exact same set of (n-1) dimensional views.
So for example if you have the plan, front, and side elevation drawings of a house; you can always design a different house that has the exact same three views. Obviously the more views you have, the faster the probability of misidentification drops.
That problem is what created the myth of “the face on Mars” which is a three dimensional picture of a four dimensional object. (the thrree geometrical dimensions plus the particular illumination source that created the image.)
It’s a problem of great interest to people who get involved in camouflaging things to disguise them as something else.
Well we need to get the Camo off the southern ocean so we can see what it really looks like.
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.
George Smith;
You earlier said you doubted the “dryness” of Antarctica, given the ice sheet buildup. Nonetheless, it has very little snowfall. But what there is remains. (Some areas are actually barren frozen rock, that get no snow. ) What Antarctica lacks is melting. There isn’t even much sublimation, due to the coldness of the air and its very low dewpoint. (There are actual frost-falls sometimes, right out of the clear sky and air.)
Paul Vaughan says:
August 18, 2010 at 4:06 pm
Apologies for my lack of clarity. I was trying to say that the HadISST Southern Ocean data prior to 1980 is nonexistent in the more southern section, and badly compromised in the northern section.
Bob Tisdale says:
August 18, 2010 at 4:51 pm
Thanks, Bob. What is the source for your ICOADS data and graphs?
w.
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.
sandyinderby says:
August 18, 2010 at 1:41 pm
I don’t pretend to understand the science and models but this this makes sense to me:
—
Judith Curry says:
August 18, 2010 at 12:16 pm
Nevertheless, the models are useful tools for trying to understand how the climate system works.
—-
I think that this is a good position for all climate scientists to take. Dr Curry has risen in my estimation for this single statement.
~~~~~~~~~~~~~~~~~~
Sandy;
If and Only If — you’re reel, reeeel careful making your models, and are very open and transparent about where you’re using straight lines and guesses to approximate curves and missing information.
And there’s the rub. These “scenario” models are incredibly clumsy. You wouldn’t believe. Average sun all day all night on a flat earth with evenly spread average temperature with evenly spread CO2 and water vapor, etc., etc. I call them “Tinkertoy models”. Gerlich and Tscheuschner, German radiative physicists, call them “computer video games”. The only things they “explore” are the preconceptions and imaginations of their designers and programmers.
What I find intriguing about this thread is the way it is highlighting current scientific methodology.
This is probably true in much wider spheres than climate science but it appears to be something like this.
Climate is extremely complex. The best tools we have for describing something complex are intricate computer models that attempt to mimic that complexity.
So we build models and compare them with measurements. [In the case above – 2 models were picked because they weren’t bad in a couple of their predictions, but the implication is that there were many aspects in which they were out or had no useful outputs. ]
The data we use may be wrong and the model may be missing important components or have incorrectly structured interactions, but it is the best we have, so we report progress on these endeavours. We do this in the sure belief that over time the models will improve and with them our understanding of how the whole thing works.
I actually don’t have a problem with this – as long as it remains in a lab or amongst the peer group fiddling with this. And until recently, that is how the scientists working on these problems would have felt things worked.
But now – as one of the posts above has shown – the media makes a big message about the “science” telling us what will actually happen. Not only is this method exposed to public media filtering and imaginative phrasing, but it is picked up in a cause to save the earth.
People like Willis then come along and try to apply the real world common sense tests to it and feel that it is shooting way past its ability to actually say anything at all about what will happen.
The dilema for todays computer modelled science is how to present its own ignorance clearly & realistically.
For those areas of research outside of the public policy area, life goes on as normal and slow improvements might be made.
Perhaps the lesson is to begin prominantly displaying the ignorance in the headings of such papers rather than allowing headings or summaries that suggest they have predictive value.
But perhaps the other problem is that modern man is so proud of his abilities, that he does not really see just how much he doesn’t know. The lesson from history is that we can have things completely wrong for very long periods of time, but we always believe that our generation does not suffer from such faults and inadequacies like our forebears did.
Willis Eschenbach says:
August 18, 2010 at 7:48 pm
Bob Tisdale says:
August 18, 2010 at 4:51 pm
Willis,
Bob did a post on his blog about each of the datasets with links to them here:
http://bobtisdale.blogspot.com/2010/07/overview-of-sea-surface-temperature.html
In it you will find the link to ICOADS website, but I went through and got the direct page to the data, however it is in netCDF files only that I can see:
http://www.esrl.noaa.gov/psd/data/gridded/data.coads.ltm.html#plot
richard telford says:
August 18, 2010 at 7:30 am
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
I will however point out that the average of several measurements is much more precise that that of the original measurements.
_________________________________________________
As an engineer, I fought the “PRECISION” versus “ACCURACY” issue my entire career. You can make measurements very “precisely” and still be very “inaccurate”.
You can measure something “precisely” to three decimal points, but if the object has high variability, the “precision” is meaningless since it is “inaccurate”. The hole a bullet makes in gelatin is precisely the size of the bullet. But if it hit the wrong gelatin pack then the shot was not very accurate. Averaging 20 readings taken to no decimal places and taking the result to three decimal places gives you a “precise” answer … unfortunately it is not “accurate” since the original measurements were not done to the same level of precision.
The o.o6 degrees C referenced in this study implies a 0.005 degree confidence interval. I doubt it.
(I know how y’all hate Wikipedia but here is a simple definition of accuracy and precision: http://en.wikipedia.org/wiki/Accuracy_and_precision
Judith Curry’s hypothesis does not fail with lousy data or computer models which lack resolution. Instead, her hypothesis fails due to simple meteorology and wind blown snow.
Those great big icicles which develop into shelves are not deposited from above as much as they come from inland. The reason for this is simple offshore flow. Since the water around Antarctica is usually warmer than the near water land surface, air over the water rises then descends over the land causing wind to blow off shore. The resulting wind blows snow out to sea (kind of like when snow is blown from Montana to Nebraska, for anyone who has lived there). Some of the snow stays on the shelf while some of it lands on open water and increases sea ice or simply melts.
In the event that the water becomes warmer, the offshore flow increases in strength and causes more snow to build up along the coast and develop more sea ice. This will not likely to be the case in the area of the horn due to it’s geography.
So, in the case of warming, more sea ice should be expected around Antarctica in times of ocean warming.
Hypothesis fails.
Judith Curry says:
August 18, 2010 at 12:06 pm
“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.”
I can eyeball a difference of 2 degrees easily, which your paper shows and Willis’ plot does not. The actual data is needed, does your contention of scale mean you distrust Willis’ plot? This should be a no-brainer, either there has been substantial and accelerated warming in the Southern Ocean or not. Models are useful in real world applications, but they can’t turn data that goes one way into data that goes the other way.
precision = variance
accuracy = bias
Three quick questions for Wayne Delbeke:
What is the “law of large numbers”?
What is the “central limit theorem”?
How is radar satellite altimetry performed to produce accurate values for sea level rise in the millimeter range?
An article on this theme: How can annual average temperatures be so precise?
Dr. Curry, thank you for taking your time to answer the questions posed by many on this blog, including the author of this post, and for clarifying your positions with respect to your study. My question, if you have time to answer is this:
Why or how did you decide on this specific hypothesis to address the growth of sea ice in the Antarctic? Did you look at other mechanisms that might explain it, and if so, why were they ruled out in favor of the hypothesis used in your study?
And what I’ve seen in the last half of this thread has been very inspiring. Thank you all! Especially Anthony for the best blog on the web, bar none.
Dr Curry says that the model was designed to find warming. I guess that they found it. No surprise there. An Old friend of mine that was chief engineer for a very important physics research lab said that scientists would come to him with a grant to prove a theory. He would design the “experiment” to get the needed proofs, never failed. How many proofs sited in physics were created by experiments that could not fail?
How many proofs of AGW are based on bad or “adjusted” data? Solid long term unadjusted temperature records show no long term warming, even cooling.
If Dr Curry wants my respect she needs to work toward a solid temperature set that reflects reality. Then we can look to see if there are human caused changes other then UHI.
O. Wolf;
If the numbers were large, you’d have a case. But they are few, and biased-selected, and just generally low quality. FAIL.
Brian E.;
If the CRU-Crew used competent professional modelers and forecasters and statisticians, there’d be some hope of gradually increasing quality of result. But they don’t. They are a DIY bunch whose amateurish kluges are regularly condemned and dissed by people who know better.
Defending my own position, and wishing that Willis would look at the other references, I did some digging (15 minutes with Google, actually), and found some more Southern Ocean temperature data — in zones where Willis says there is hardly any, covering the period when Willis says there is hardly any; and supporting Southern Ocean warming where the plots added to the thread by Dr. Curry show there is Southern Ocean warming. Since I’ve got links to all the papers I found (well, abstracts at least), I suggest reading them isn’t that difficult.
Meredith, M. P., and J. C. King (2005), Rapid climate change in the ocean west of the Antarctic Peninsula during the second half of the 20th century, Geophysical Research Letters, 32, L19604, 10.1029/2005GL024042.
http://www.agu.org/pubs/crossref/2005/2005GL024042.shtml
Abstract:
The climate of the Western Antarctic Peninsula (WAP) is the most rapidly changing in the Southern Hemisphere, with a rise in atmospheric temperature of nearly 3°C since 1951 and associated cryospheric impacts. We demonstrate here, for the first time, that the adjacent ocean showed profound coincident changes, with surface summer temperatures rising more than 1°C and a strong upper-layer salinification. Initially driven by atmospheric warming and reduced rates of sea ice production, these changes constitute positive feedbacks that will contribute significantly to the continued climate change. Marine species in this region have extreme sensitivities to their environment, with population and species removal predicted in response to very small increases in ocean temperature. The WAP region is an important breeding and nursery ground for Antarctic krill, a key species in the Southern Ocean foodweb with a known dependence on the physical environment. The changes observed thus have significant ecological implications.
M.J. Whitehouse, M.P. Meredith, P. Rothery, A. Atkinson, P. Ward, R.E. Korb, 2008: Rapid warming of the ocean around South Georgia, Southern Ocean, during the 20th century: Forcings, characteristics and implications for lower trophic levels, Deep Sea Research Part I: Oceanographic Research Papers, Vol. 55, no. 10, Pages 1218-1228, ISSN 0967-0637, DOI: 10.1016/j.dsr.2008.06.002.
http://nora.nerc.ac.uk/6188/
Abstract (sorry for the length):
The Southern Ocean is known to have warmed considerably during the second half of the 20th century but there are few locations with data before the 1950s. In addition, assessments of change in this region are hampered by the strong seasonal bias in sampling, with the vast majority of data collected during the austral summer. However, oceanographic measurements near South Georgia span most of the last century, and we here consider almost year-round data from this location over an 81-year period (1925–2006). We observe significant warming between the early and late 20th century, with differential warming between summer and winter months and an indication that late 20th century summer temperatures peaked ~6 days earlier. To quantify the long-term warming trend in this highly variable data, a mixed model utilising a Residual Maximum Likelihood (REML) method was used. Over the 81-year period, a mean increase of ~0.9 °C in January and ~2.3 °C in August was evident in the top 100 m of the water column. Warming diminished below 100 m and approached 0 at 200 m. Thus the long-term warming around South Georgia is substantial—more so than documented previously for the circumpolar warming of the Southern Ocean. We examine potential causal effects of this trend, including local atmospheric and cryospheric change, the influence of upstream waters and the role of coupled modes of climate variability such as El Niño/Southern Oscillation and the Southern Annular Mode (SAM). It is likely that all of these play a part in the observed temperature increase. However, the role of the SAM is strongly indicated, via its likely role in the circumpolar warming trend in the Southern Ocean, and also by the atypical response of the South Georgia region to changes in heat fluxes associated with the SAM. Furthermore, the combination of a regional decline in ice extent and strong upstream warming likely explains a significant part of the strong seasonal variation apparent in the warming trend. In addition, we consider the implications that long-term warming has for South Georgia’s lower trophic levels. For Euphausia superba, at their northern limit, we find a significant negative relationship between summer South Georgia water temperatures and mean summer density of E. superba across the southwest Atlantic sector of the Southern Ocean. Simple abundance and growth rate relationships with our long-term temperature data appear to show declining habitat suitability for E. superba. In contrast, the warming trend is likely to favour other macro- and mesozooplankton species that occupy the more northerly parts of the Antarctic Circumpolar Current, and it is likely to promote phytoplankton growth.
(South Georgia Island is located right at 55 S. The Antarctic Peninsula is located considerably further south. Interestingly, the HADISST plot appears to be somewhat in conflict with the South Georgia vicinity data described above. )
On my own blog (findable with Google), I’m about to post a link to a freely available online copy of Levitus, Antonov and Boyer 2005: “Warming of the World Ocean, 1955-2003”. Perhaps worthy of reading, too.