Just a little something he threw together
Guest Post by David Middleton
First the breath-taking headlines…
- Scientists Report Faster Warming in Antarctica, New York Times (WUWT commentary)
- West Antarctic Ice Sheet warming twice earlier estimate, BBC
- West Antarctica warming much faster than previously believed, study finds, NBC
- Western Antarctica is warming three times faster than the rest of the world, Grist
Oh noes out the wazzoo!!!
What could possibly have caused such an out-pouring of Mr. Bill impersonations?
Apparently this did…
Central West Antarctica among the most rapidly warming regions on Earth
David H. Bromwich,1, 5 Julien P. Nicolas,5, 1 Andrew J. Monaghan,2 Matthew A. Lazzara,3 Linda M. Keller,4 George A. Weidner4 & Aaron B. Wilson1
Nature Geoscience Year published: (2012) doi:10.1038/ngeo1671
Received02 May 2012 Accepted15 November 2012 Published online23 December 2012
Abstract
There is clear evidence that the West Antarctic Ice Sheet is contributing to sea-level rise. In contrast, West Antarctic temperature changes in recent decades remain uncertain. West Antarctica has probably warmed since the 1950s, but there is disagreement regarding the magnitude, seasonality and spatial extent of this warming. This is primarily because long-term near-surface temperature observations are restricted to Byrd Station in central West Antarctica, a data set with substantial gaps. Here, we present a complete temperature record for Byrd Station, in which observations have been corrected, and gaps have been filled using global reanalysis data and spatial interpolation. The record reveals a linear increase in annual temperature between 1958 and 2010 by 2.4±1.2 °C, establishing central West Antarctica as one of the fastest-warming regions globally.
[…]
The manufactured “record reveals a linear increase in annual temperature between 1958 and 2010 by 2.4±1.2 °C.” That’s a 50% margin of error on the reconstruction that supposedly corrected the recording errors.
I haven’t purchased access to the paper (nor do I intend to); however, the freely available supplementary information includes a graph of their reconstructed temperature record for Byrd Station. It looks very similar to the NASA-GISS graph that doesn’t show any significant recent warming trend.
The NASA-GISS data (GHCN & SCAR) for Byrd Station are in two segments: 1957-1975 and 1980-2012. The 1957-1975 series depicts a moderately significant (R² = 0.19) warming trend of about 1.0 °C per decade. The post-1980 series depicts a statistically insignificant (R² = 0.01) trend of 0.3 °C per decade.
Bromwich et al., 2012 get their 2.4 °C of warming from 1958-2010 (0.4 °C per decade) by stitching together the fragmented data sets. If I just combine the two NASA-GISS series, I get a trend of about 0.4 °C per decade…
But, almost all of that warming took place before 1988. And Byrd Station has seen no warming (actually a slight cooling) since 1991.
Furthermore, the corrected temperature record of Bromwich et al., 2012 appears to actually depict more cooling since 1991 than the uncorrected data…
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
As Adam Gallon says:
December 27, 2012 at 6:37 am
And we’ve got a weather station that’s showing a maximum summer temp below -10C!, and that was in the early 1970s.
Now, when I was at school, we were taught that ice melts at 0C.
==========================================================
I am with Adam, or am I missing something ? What about the fact that it does not matter what the average temperature is, it matters how many hours the ice sheet is above 0C for it to contribute to rising sea level. And subtract from those hours the number of hours the sea is below the freezing point, creating new ice. I did not see any of the averages above 0C. You could assume an increase in average temperatures means an increase in melting hours, but you would need proof for that assumption. Oops, used the “p” word, that is a bad word in Mannian climate science. And we all know about about “assume.”
David Middleton says: “Unless there is some reason to assume that a logarithmic, exponential or other type of function would better capture the trend, linear regressions are generally the best choice.”
Thanks for an interesting article but this last statement is just main-stream climate stupidity.
You don’t fit a linear “trend” or any model because you don’t know what else to use therefore it must be “best”. You make a model that you think may mean something and fit it.
If you can’t think a model that may mean something, you don’t fit one , it’s that simple.
If the term “linear trend” was banned from climate science 95% of them would be out of a job. The 5% that remained may be capable of producing some science.
Berényi Péter says,
“(I am fed up with bloated headlines in scientific journals that contribute nothing to science but are designed to be picked up by MSM zombies disinclined to check the background.)”
This got me thinking.
Partly for the fun of it, and partly as a teaching mechanism, WUWT could have a yearly program like the Academy Awards. However, the awards will be for mis-deeds, in the spirit of the Darwin awards. If there’s a budget for it, you’d want a big stage, emcees dressed in Tuxedos and formal gowns, lots of glitters, maybe even some singers (of parodies) and mime/dancers (featuring the Hot to Trotters; they could do stuff like rhythmically touch ice statues and pretend that they’ve been burnt; another variation, inspired by the opening scene of 2001, A Space Odyssey, is that the dancers could be dressed as scientists, and touch an ice slab, which them causes them to de-evolve, pronto. You’d know they de-evolved, because they’d start dancing like monkeys).
We could call these the Climate Catastrophe Awards. Categories would include:
1) The Dupe-ee (or Dupe-ease) Category, awarded to a newspaper or magazine, for most misleading headline. (I’d vote for NYT/”Scientists Report Faster Warming in Antarctica”)
2) Hot Air Category, for the most Hansen like (i.e., super catastrophic) prediction by a scientist
3) Blood and Gore Category, highest number of distinct errors (which remind us of wounds) within a given piece of work. You could also call this the Inconvenient Truths Category.
4) No Blood, Just Gore Category, for most disgusting statement by a climate catastrophist. The pedophilia insinuation will have stiff competition from the Parncutt’s Nazi-like death wishes.
5) Exxon-Mobil Big Blank Check Category, for researcher who deserved the biggest check from Exxon-Mobil for the past year, but actually got nothing from them. In addition to a statue, a big, 3 foot blank check from Exxon-Mobil will be printed up and filled out, awaiting only the CEO’s signature
Hey, if I could have got away with crap like this in high school and college in the late sixties and early seventies, I could have been a PhD and a Nobel Laureate.
Man Bearpig says:
December 27, 2012 at 2:39 am
Am I right in surmising that this is the only station showing such warming? If so wasn’t this picked up on by the reviewers? If not is the journal using suitably qualified reviewers?
Keith Guy says:
December 27, 2012 at 2:41 am
Can anyone please enlighten me? Is the NASA GISS data for Byrd station raw data, or has it already been adjusted in some way?
I do not have all of the raw numbers with me right now, but as I recall, only the West Antarctica area is “warming” – the rest of Antarctica is getting substantially colder over time – particularly the higher very cold central plateau well away from the water. The water, of course, closely surrounds the peninsula where this ONE thermometer (used to be) located.
NASA-GISS automatically (and without outside auditing nor manual oversite) changes EVERY past record every month based on today’s (most recent) light readings because Hansen is trying to “smooth” every available long-term record into a wide a spacing as possible . I can’t prove it, but I have no reason to think Hansen’s algorithm – which adjusts past temperature records based on today’s latest reading and today’s light values, while editing out past records that have intervals or gaps and smoothing the data between stations up to 1200 km – does anything different for the Antarctic than it does for central Tennessee.
Vuckevic: – Circumpolar current – by far the world’s strongest ocean current. Its waters and the atmosphere have an 8 year temperature cycle) ACW-Antarctic circumpolar This wave propagates westward against the current but ultimately ends up travelling eastward, due to the massive size of the ACC, at a slower rate than the mean flow. The wave circles the earth every eight to nine years (White and Peterson, 1996). It has a long wavelength (wavenumber=2) resulting in two crests and two troughs at any given time. The crests and troughs are associated with massive patches or pools of warm water and cold water respectively. The areas can be thousands of kilometers long. The warm patches are 2 to 3°C warmer than the mean sea surface temperature (SST) and the cold patches are 2 to 3°C cooler than the mean SST.
http://www.spacedaily.com/images/antarctic-circumpolar-wave-bg.jpg
That graphic is very interesting. The Antactic is the only place in the world where the tides propagate as more or less straight waves, like they are going somewhere instead of turning around in circles around an antidrome.
Secondly the fact that this is a double wave is also suggestive of a tidal phenomenon.
Something fairly regular seems to happen at the other end as well. Though seems a bit longer.
http://i49.tinypic.com/xudsy.png
David Middleton says:
December 27, 2012 at 6:43 am (responding to)
David Middleton: Please name one trend of any magnitude in climate science over any time frame that is linear. EVERY temperature record is cyclical – from the air under a passing solar eclipse over a 15 minute period, to the daily rise and fall of temperature overnight, to the seasons to rainfall to snow and ice extents, to the century-long little ice ages between the 110,000 year glacier cycles that repeat themselves over a 4,000,000 year cycle – the ONLY linear relationship in the earth’s history is Hansen’s
T(t) = Mann(CO2) + b
David M and P Solar – with respect: before thinking about how to ‘capture a trend’ should one not consider whether or not there really is any reason to suggest that there is any trend at all? If, as IPCC has hinted years ago, climate is chaotic – i.e. is not state determined (the coming state is not determined by its current state) why should we expect to find any trend at all? Imposing a trend on a heap of numbers is a private psychological exercise and little to do with understanding the ‘nature of things in the real world’.
Please forgive the compression of this argument; I am sure you get my drift? There’s a lot of this about – grabbing at trends without first trying to determine whether or not the system is chaotic. The ghastly truth is that much science is (inadvertently) fraudulent in this respect.
As Nurse said – science is all we have to hand – what he did not say was -‘ be prepared to be surprised!’
I agree that the application of linear regressions to cyclical functions is wrong. The problem is that you almost never have sufficient instrumental temperature data from which to build a model to demonstrate the millennial –scale cyclicity.
If you look at the non-Hockey Stick Northern Hemisphere reconstructions (Moberg, Loehle, Esper, Ljungqvist, etc.), it’s pretty easy to build a model because you can actually identify the millennial-scale cycle. Davis & Bohling (2000) identified numerous Holocene cycles in the GISP2 ice core using Fourier analysis. The dominant cycle had a period of approximately 1,000 years. The ~60-yr cycle was the most prominent sub-centennial oscillation.
The Antarctic instrumental records are too short to show even the 60-yr cycle. The global instrumental records (HadCRUT) clearly show the 60-yr cycle; but are way too short to show any of the longer period cycles. The apparent secular trend obtained by linear regression is the upward leg of the millennial cycle.
I agree with you. The problem is that almost all of the instrumental climate data are of too short a duration to “see” that cyclicity.
See this WUWT post of mine for a discussion on this topic.
These are as “raw” as you can easily download from NASA-GISS.
The station selection menu has three options:
1) GHCN v3 (adj) + SCAR data
2) After removing suspicious records
3) After GISS homogeneity adjustment
I selected the first option. These are not raw data; but they have not been “homogenized.”
oeman50 says: (December 27, 2012 at 7:16 am)
> I am with Adam, or am I missing something ? What about the fact that it does not matter what the average temperature is, it matters how many hours the ice sheet is above 0C for it to contribute to rising sea level.
Is the air temperature (that is being measured) a good proxy for the temperature of the ice?
I would have thought that the air temps being above 0C would melt ice very slowly, wouldn’t the rate of ice melting be more of a function of the temperature of the ice itself and the amount of sunlight energy it is absorbing?
” Carlyle says:
December 27, 2012 at 1:24 am
You just can not trust real data. It might give the wrong impression.”
That about sums it up. People with a cause can’t stand the truth.
While there is a great deal of stochastic variability, the climate of the Holocene has had very clear patterns of climatic oscillations (quasi-periodic fluctuations). In my opinion, the best, most succinct analysis of these patterns is:
Davis, J. C. and Bohling, G. C. (2000), The Search for Patterns in Ice Core Temperature Curves. Environmental Geosciences, 7: 213–214. doi: 10.1046/j.1526-0984.2000.74003-3.x
A version of this paper can be found in Lee Gerhard’s Geological Perspectives of Global Climate Change.
mitigatedsceptic says:
December 27, 2012 at 8:22 am
long term, we need to be able to “work it completely through” from first principles and with EVERY factor and EVERY relationship completely defined. At that time, we can run computer models that will predict the future weather, and over time, the future climate.
Short term? Why not use a physical model that WORKS over the period we do have some sort of data for?
In other words, build a “Stonehenge-quality analog calculator” that predicts the seasons and keeps track of solar and lunar positions over a period of centuries to within one morning per year. Granted, we would build that “calculator” with a spreadsheet and pixels rather than granite and hillsides, but that’s only because hillsides are not very mobile, and can prove difficult to edit”
But, Stonehenge is more accurate after 5000 years than the IPCC and 23 supercomputers are after 5 years!
you don’t have to know Einstein’s Special Theory of Relativity nor Newton’s law of Gravity nor Copernicus’ orbits nor Kepler’s rules nor even Aristotle’s epicycles and their useful fudge factors to “plot” the stars, moon, and sun rise positions to know the seasons and when to plant and when to harvest.
Right now, all we need is “good enough” …. Given time, the theory will follow.
Right now, we need to know if the long-term warming period and the 60 year short AMO-PDO-Arctic cycle that see us warming up from the Little Ice Age but cooling from the Medieval Warming period and Roman Optimum will end in 2000, 2060, or 2120.
mitigatedsceptic says: David M and P Solar – with respect: before thinking about how to ‘capture a trend’ should one not consider whether or not there really is any reason to suggest that there is any trend at all?
That was what I was saying , sorry if I was not clear enough.
Good – I did not make the correct inference from your posting – sorry!
Now if such systems as climate behaviour are so complex in structure and strictly chaotic in behaviour, common sense (perhaps not too common in the lab?) would dictate that here is a limit to human knowledge that cannot be overcome by appeal to the current causal paradigm and it will not yield until a new paradigm is on offer. Consider this problem in the same light as a flat-earther who cannot explain why an ideal cannon ball shot off in any direction returns to him from exactly the opposite direction. Everyday experience has taught the flat-earther that his paradigm is robust and produces predictable forecasts of most phenomena but the paradoxical cannon ball is simply beyond the limits of his knowledge and must remain so until he changes his viewpoint to accepting the round world hypothesis. From my limited experience, I know of no paradigm that could help unravel the paradox of chaotic systems and help us predict their future behaviour. Maybe chaotic systems are just florid examples of the long-standing problem of inductive inference that so troubled 18th C. but which has been side-tracked ever since then.
Anyway – to be brief – I am coming to the conclusion that no one can predict how climate will change and that politicians who listen to climate scientists are being conned. I do not want to judge whether or not the climate scientists are wanton fraudsters or simply ignorant or poor trying to earn a crust by riding a passing bandwagon; but whoever they are, whatever their motives, they are doing humanity a great disservice.
RACookPE1978 says: In other words, build a “Stonehenge-quality analog calculator” that predicts the seasons and keeps track of solar and lunar positions over a period of centuries to within one morning per year.
God damn, even Stone Henge doesn’t use linear trends. Just shows how far back climate science has regressed. (An it’s not a linear regression, it’s exponential !)
David Middleton
“…Unless there is some reason to assume that a logarithmic, exponential or other type of function would better capture the trend, linear regressions are generally the best choice.”
I guess my education as a Physical Chemist taught me to empirically derive or develop from first principles the “correct” functions. For interpolation purposes one can default to linearity if the data supports, but for understanding of the underlying mechanisms or any attempt of extrapolation one must use the proper function and it’s unlikely anything is strictly linear.
If this field of climatology wants to predict the future they simply have to quit with the linear extrapolations.
richard says: December 27, 2012 at 3:53 am
“….The paper overall does not say what McGrath reports. This is not about “warmist” or “skeptic”, but an obvious inability to take a science paper, consider it, and give a reasonable and accurate report of its contents and findings with some intelligence applied to them by the correspondent or journalist…..This is not journalism or reporting, but, deliberate or not, advocacy….”
This is simply ‘propaganda science’ – the message is decided and the headlines are written with little regard to the story.
One study, a single station, infilled and adjusted records, most warming occurred before 1988 …. and that is enough to justify headlines such as these?:
“Faster Warming in Antarctica..” ..” West Antarctic Ice Sheet warming twice earlier estimate….” “West Antarctica warming much faster than previously believed….” ” Western Antarctica is warming three times faster than the rest of the world…”
I’m really a bit surprised that intelligent CAGW proponents don’t cringe at this sort of thing, but seemingly they all see this as good science and good scientific proof.
I agree with you.
The problem is that there aren’t enough instrumental temperature data, particularly in places like Antarctica, of sufficient record length from which to derive the correct (cyclical) functions.
The Byrd Station data are what they are. The slope was steeper before 1975 than it was after. If you break it up into segments, as I did, the Gorebots will accuse you of “cherry-picking.” If you apply a polynomial or quadratic function, the Gorebots will accuse you of over-fitting. There’s enough data for a proper spectral analysis, to identify the long-term cyclicity.
Should be: “There’s not enough data for a proper spectral analysis, to identify the long-term cyclicity.
David L says:
December 27, 2012 at 9:37 am
Certainly true.
IF – and as the esteemed Dr Robert Brown at Duke has pointed out – if we knew everything and every deptail about every factor affecting the world’s climate over even a 100 year span (much less a 1500 year span over 1-1/2 long term cycles) then we could begin programming the real solution (the real analysis) from first principles. That IS the right way to do it.
But we cannot. We simply do not know enough to begin thinking we can define enough variables to define enough restraints, constraints, and relationships between the restraints to calculate “climate” from first principles.
Further, the world’s climate is so varied, so random that it varies by +/- 0.2 degrees even in the satellite record over intervals as short as 3 months! Therefore, real world, every “proxy” and thermometer readings for the past 2000 years need to be “plotted” with a +/- 0.25 “Gray Line” of “average trend before any decision can be made about long-term cycles.
A “real world” analog-equivalent as the CORRECT combination of the correct series of cyclic influences will begin the process – and note always that NOT all of these cycles will be a uniform sinusoid of mathematically perfect cosine waves of exactly even periods!!
This “almost good enough” or “Poor Robert’s Almanac” summation of two, three, or five simple cyclical approximations is all that is what is needed over the next 15 to 30 years. We are looking for a beginning, that first plot of the Atlantic rift zone and the Pacific rim of Fire and the Ethiopian fracture zone and the seafloor earthquakes before we can start projecting continental drift and calculate plate tectonic movements.
That, plus a better agreement of when the Medieval Warming Period, the Roman Optimum, and their previous cycles actually occurred.
I don’t think any automated weather station is going to last very long in this area. They quickly get buried in snow, blown-down by the wind or sink into the ice as the components warm up in the Sun and melt the ice below its connection stands.
Look at this large crane left at Byrd station.
http://www.flickr.com/photos/kool-kini/3290311282/
I’m in the Evil O & G industry myself. One of the things my wicked overlords insist is the sniff test. In this case they would have asked me if, as the raw data doesn’t show the same trend as the “fixed” data, if I hadn’t detected a hint of sulphur wafting around.
Of course, being Evil, afterwards they would not have terminated my services exactly. They would have redirected me into Marketing. Which appears not too dissimilar to Climatology.
David Middleton
“…Unless there is some reason to assume that a logarithmic, exponential or other type of function would better capture the trend, linear regressions are generally the best choice.”
Linear regressions are certainly the easiest to compute but as mitigatedsceptic and P. Solar indicate, in order to justify using linear regression an assumption as to the existence of a trend, let alone the nature of that trend, has to be made. Mr. Middleton, with whose work I am always impressed, refers to “other types of function” which would include such complex dynamical systems as fractals.
When addressing David L.’s suggestion that “EVERY temperature record is cyclical”, David Middleton says
“I agree with you. The problem is that almost all of the instrumental climate data are of too short a duration to “see” that cyclicity.”
The same argument surely applies to the assumption of linearity.
RB says:
December 27, 2012 at 1:30 am
“The fact that temperatures are rising in the summer means there’s a prospect of WAIS not only being melted from the bottom as we know it is today, but in future it looks probable that it will be melting from the top as well,” he said.
Can someone tell me if this is even physically possible? What is the ambient temperature in the Antarctic, even with “rising” summer temperatures?
Here are the monthly temperature averages for Byrd. Dec is the warmest months with -14.2C. I doubt air temperatures ever get above 0C.
http://www.levoyageur.net/weather-city-BYRD-STATION.html
Note the absence of a lag between the summer solstice and maximum temperatures. This tells us that solar insolation directly drives surface temperatures and we can infer the ‘warming’ resulted from decreased clouds.
The WAIS will lose ice from the surface by sublimation at temperatures well below zero, depending the amount of solar insolation it receives. To call this ‘melting’ is misleading and poor science. But necessary to put into peoples minds that increased atmospheric temperatures from increased GHGs is the cause of ice loss, when it clearly isn’t at these temperatures.