… Or It Could Mean Nothing

Kevin Drum wrote:

CLIMATE CHANGE UPDATE….The Wilkins ice shelf is collapsing:

A chunk of Antarctic ice about seven times the size of Manhattan suddenly collapsed, putting an even greater portion of glacial ice at risk, scientists said Tuesday.

….British Antarctic Survey scientist David Vaughan attributed the melting to rising sea temperature due to global warming.

….Vaughan had predicted the Wilkins shelf would collapse about 15 years from now.

All the usual caveats apply. However, this is one more data point suggesting that global warming may be happening faster than our current models predict, not slower.

I responded in the comments section:

A few observations.

  • Global temperatures have been flat for 8-10 years, after being up substantially the decade previously.
  • Recent ocean measurement work as reported on NPR show ocean temps. over last 5-6 years to be flat to slightly down
  • 98% of Antarctica has cooled over the last decades and has built up ice pack — 2% has warmed (in the Antarctic Peninsula). I will leave it to the reader to guess where Al Gore sent his cameras
  • In August 2007, or about a half year ago, sea ice extent around Antarctica was the largest ever recorded (since measured by satellites in 1979). So, within the last 6-8 months, Antarctica had record sea ice buildup.

Given this backdrop, it is astounding that one could interpret the collapse of an ice sheet that happened faster than one scientist predicted as "accelerating global warming." I can’t think of any mechanism where the behavior of an ice shelf would be a more sensitive measure of the pace of global temperature change than would be the direct measurement of air and sea temperatures themselves.

There are two ways to interpret this ice sheet collapse that are far more "reality-based"

One, the collapse is a result of the fairly well-known and relatively isolated local/regional warming in the Antarctic Peninsula (where I believe this shelf is located). In other words, it signals a local phenomenon rather than a global one, or

Two, the scientist who originally predicted the date of the ice shelf collapse made an incorrect prediction. There is no particular loss of face in this – after all, such events are part of cycles that last long enough that, in many cases, we have not been able to observe even one entire cycle with modern tools. It would be the height of hubris to say that we understand and forecast these decadal and even longer cycles and events well enough to declare that deviation from forecast must represent a change in nature rather than our own poor understanding.

Update:  More here via Q&O

50 thoughts on “… Or It Could Mean Nothing

  1. Alan Roberta

    The latest news about climate change is so alarming (the right wing would say alarmist) as to make many people want to plant their aching heads in the sand. Some scientists using advanced computer models now argue that if we want to stop the Earth from warming, the amount of carbon we should be emitting is … none. None? As in, zero? As in, shutting down the global industrial economy? After all, global energy demand is expected to accelerate until at least 2020. Yet attempts even to slow the rate of increase of carbon emissions have paralyzed world politics for more than a decade.

  2. Patrick Hadley

    The Antarctic Peninsula is between 63 and 73 degrees south – so it is as far from the south pole as northern Scandinavia is from the north pole. All the snow at sea level in Finland, Norway and Sweden always melts during the summer.

  3. Scientist

    And yet again, the erroneous claim that temperatures are ‘flat’ is trotted out. Here is the most recent set of ten five year running means of GISTEMP global temperature anomalies, in units of 0.01°C:

    35.4 45.8 48.6 47.8 51.4 57 56.2 59 65.8 67.4 68.2

    Do they look flat to you?

    And I can only assume you invented your figure of 98%. Doran et al (2002) found that 58% of the continent had cooled between 1966 and 2000. And have a look at the results of another study covering 1980-2007. Does that look like 98% to you? It’s obvious you don’t take the trouble to do actual research on the issues that you vent forth on so confidently.

    Patrick Hadley – what relevance does snow melt in Scandinavia have to a discussion about Antarctica? Are you trying to suggest that latitude alone can tell us something?

  4. bill-tb

    It’s also quite possible the under ice volcanoes in the area are more active that known, like with the Arctic Ocean. We knew little about the undersea activity on the Gakkel Ridge until the ice cleared somewhat in the last few years. Now it’s believed to be one of the most active undersea segments on the whole of the Atlantic mid ocean rift.

    As the Earth cools, the argument is going to heat up, no doubt about that. Too much invested in the hoax to quit now.

    Aren’t those ARGO diving buoys cute? Some real cool engineering went into that system.

  5. James Howe

    A couple months ago there was a report about how an undersea volcanic eruption 2000 years ago was probably contributing to the fast movement of the Pine Island Glacier (http://news.nationalgeographic.com/news/2008/01/080121-antarctica-volcano_2.html). Another article described research into buried lakes which move around and have contributed to rapid ice movement. Is the Wilkins ice shelf anywhere near the Pine Island Glacier, and could it have been affected by the volcanic eruption as well? Could this collapse be related to buried lakes (the number and location of which are just now being discovered because of new satellites?)

  6. Patrick Hadley

    The relevance of the comparative distance of the Antarctic Peninsula from the south pole should be obvious – as indeed should be the link between latitude and climate. To confuse conditions a very small and atypical part of Antarctica with that of the continent as a whole seems rather unscientific. The total ice in Antarctica is certainly growing steadily year on year – how does that fit in with global warming theory?

    I have yet to read much reaction from the alarmists to a series of rather more significant scientific results. It has been nine years since the first Argo floats were launched. They provide data from the whole of the oceans down to two miles in depth, collected by 3,000 highly accurate floats. NASA tell us, in the most low key way possible, that the oceans are cooling – yet 80% to 90% of global warming is supposed to occur in the oceans. Last year Gouretski and Koltermann proved that the XBT readings had exaggerated deep ocean warming over the fifty years before Argo by a factor of 0.62. The NASA Aqua satellite results show clouds having a negative feedback on warming. Yet “Scientist” wants to make a big deal out of erosion of an ice shelf 2,500 km from the south pole.

    That is the difference between sceptics and alarmists. Sceptics look at all the evidence that comes out and seek to put it in context and fully understand it. Alarmists cherry pick bits of news, no matter how meaningless, that suit their agenda, but completely ignore any research or evidence that does not fit in with their faith in Settled Science.

  7. Scientist

    Considering the bold claim made in your final paragraph, it’s ironic that you’ve made such a major error of thought in your first. Here’s a little exercise for you – here are several cities roughly 55° away from the equator: Edinburgh, Moscow, Petropavlovsk-Kamchatsky, Ushuaia, Stanley. Do you think they all have the same annual average temperature? Decide that, then go away and look up the answer.

    Now, for the next major misconception, do a little thought experiment. Imagine a place where the temperature never rises above freezing, and with very low precipitation. It is covered by ice. If the average temperature was -20°C, but rises to -15°C, how much of the ice will melt? At the same time, imagine that the amount of precipitation increases. What will the effect of that be on ice mass?

  8. Patrick Hadley

    Scientist, do you really think that latitude has no relevance to climate?

    But can you point out any alarmist site that has dealt with the recent Argo findings? I would love to read an explanation from the “consensus” about where all the extra heat caused by AGHG has gone – but I cannot find one.

    The Gouretski Koltermann paper was published over a year ago in the AGU (a proper journal) after peer review and seems to be fully accepted and used in later papers and conferences. It states very clearly that the supposed warming of the oceans – absorbing around 90% of the global warming – has been exaggerated by a factor of 62%. Where has the effect of this finding on climate models been discussed in alarmist media?

    Why is it only Roy Spencer who is talking about the data from the Aqua satellite that, he claims, shows that clouds are operating as a definite negative feedback – which if true knocks a huge hole in the alarmists’ climate models?

    You will find that every piece of news about the climate is given an airing on many sceptical sites – whether it seems helpful to their side or not , but that anything that is a bit awkward to the alarmists is completely ignored by them.

  9. Patrick Hadley

    Incidentally “Scientist” if you do not think that moving 2,500km away from the poles is relevant to climate consider that 2,500km south of Edinburgh takes you to Morocco, 2,500km south of Moscow and you are in Jordan, 2,500 km north of Stanley you are in the region of Argentina north of Uruguay. And of course your illustration served to show that close proximity to the sea – such as being a peninsula – will have a big effect on the climate of a region.

    As for your question about ice and precipitation, how does it help the alarmist position about the risks of the icecaps all melting?

  10. morganovich

    some perspective:

    the full Wilkins 6,000 square mile ice shelf is just 0.39% of the current ice sheet (just 0.1% of the extent last September). Only a small portion of it between 1/100th-1/200th of Wilkins has separated so far, like an icicle falling off a snow and ice covered house. And this winter is coming on quickly. In fact the ice is returning so fast, it is running an amazing 60% ahead (4.0 vs 2.5 million square km extent) of last year when it set a new record. The ice extent is already approaching the second highest level for extent since the measurements began by satellite in 1979 and just a few days into the Southern Hemisphere winter and 6 months ahead of the peak. Wilkins like all the others that temporarily broke up will refreeze soon. We are very likely going to exceed last year’s record. Yet the world is left with the false impression Antarctica’s ice sheet is also starting to disappear.

    ice extent graph here:

    http://icecap.us/images/uploads/current_anom_south0325.jpg

    ironically, it’s shaped like a hockey stick…

    there is no evidence of any “unprecedented” melt in the antarctic. this is just the same evergreen story these guys trot out every year. 0ver 70% of sea ice melt in the antarctic every year and has for every year since we have known anything about it. when it does, it’s dramatic. it makes impressive footage. but that doesn’t make it unusual or a sign of global warming.

    this is yet another of these massive leaps of faith the media is asking us to take. ice falls, global warming is the cause, but never is any evidence shown actually linking causality. we are just supposed to take it on faith, like the positive water vapor feedbacks that are looking more and more to be the opposite of what observably occurs.

    there are lots of other reasons an ice shelf can fracture. recent work on the much discussed larsen b shelf:

    http://www.igsoc.org/journal/54/184/j07J086.pdf

  11. morganovich

    oh, and “scientist”, nice try using a 5 year running mean to obscure a 6-7 year trend. the cooler tail falling off obscures the recent trend, but the trend is there and not ambiguous in spite of such attempts at statistical misdirection. is a five year average the way to measure a 7 year trend? no.

    look at the actual GISS data:

    http://www.junkscience.com/MSU_Temps/MSUvsGISTEMP.html

    see any warming since 2001?

    not even the exaggerated GISS data (no other global set has temps exceeding those of 1998 in later years. UAH, RSS, and HADCRUT3 all have 1998 as peak) shows warming since 2001.

    so what are you carrying on about?

  12. coveman

    Thanks to the Antarctic Circumpolar Current (ACC), the Wilkins Ice Shelf lies DIRECTLY DOWN-STREAM of the recently-discovered undersea volcanic activity mentioned by bill-tp. Just a thought.

  13. Mister Jones

    Warming doesn’t have to be a reason for the ice shelf breaking off. Even in periods of extreme cold a similar breaking off is just as possible. It’s all down to the mechanical stresses put on the ice from tidal surges, ice thickening, ocean current activity etc.

    Think of it like a pane of glass subject to varying pressures, including it’s own weight. I know the glass analogy is a poor one, as glass is many times more mechanically strong than ice, but sea ice like any material, can become very brittle and shatter under various mechanical conditions. Besides, in an Antarctic environment there is constant pressure from reforming ice on the main landward ice pack which will inevitably result in large masses breaking off periodically. There doesn’t have to be any warming at all for this to happen.

  14. Fred

    So this appears to be a simple case of glacial calving . . . sheets of ice pushed out over water where they float and eventually cause themselves to crack and drift away. It is especially common at the end of the summer season in glacial areas.

    Which means this glacier, as they are wont to do, is expanding.

    But Al Gore told me that glaciers retreating was the the signature piece de resistance evidence of global warming. Which must then mean that glacier’s advancing is a de facto sign of global cooling.

    I am not a scientist and I did not stay at a Holiday Inn Express last night.

  15. An Inquirer

    I have a question that does not seem to be addressed in the news stories or at Real Climate or elsewhere. What is the difference between an ice shelf and sea ice? Much sea ice melts each year, but the collapse of an ice shelf seems to be something differnt. Perhaps this portion of the Wilkins ice shelf will be refrozen in a couple of months, but I have a feeling that it would not be called an ice shelf.

  16. Scientologist

    Dear Inquirer,

    Sea ice floats on the sea. An ice shelf rests upon land.
    We at the scientology center, NASA Goddard, believe antarctica will ascend to the astral plane, so it is shedding its ice for that purpose.

  17. Scientist

    Patrick Hadley – looks like I really will have to spell it out. NOT EVERYWHERE EQUIDISTANT FROM A POLE WILL HAVE THE SAME CLIMATE! Comparing northern hemisphere latitudes to southern hemisphere ones is really stupid. Most 12 year olds know that the south pole is significantly colder than the north pole, and this means that a place at 65°S will be significantly cooler than a place at 65°N. The fact that you didn’t know this tells me a lot about your scientific abilities. The fact that you think the latitude of Scandinavia can tell us anything about glaciers in Antarctica is similarly informative.

    morganovich, joshv – amazed that I would choose to minimise the effect of internal variation to better discern the long term trend? That’s because you are wilfully blind to the data. Is a seven year trend meaningful in climate studies? No it isn’t. Does a five year running mean tell us more about climate than single year averages? Yes, yes it does. That’s why I quoted it. Do you understand? And how about actually fitting a line to the curve. Someone’s done that here. Tell me what you see.

  18. An Inquirer

    Dear Scientologist,
    Thanks for your reply. It prompted me to look at non-Climate websites to look for an answer. Your answer is somewhat confusing; according to Wikipedia (which admittedly has been known to contain mistakes) an ice shelf also floats on the sea. The difference appears to be that sea ice forms when sea water freezes, but an ice shelf comes from glaciers that spread out on the sea. Which leads me to my next question: If a glacier feeds an ice shelf, wouldn’t we expect occasional break-ups? — especially at 65 dgrees South? — especially when we know that the ocean current in the area has warmed a bit? And will the glacier refeed the ice shelf?
    Although my impression may disppoint the pro-AGW movement, I do not see this Wilkins break-up on the Penisula a big deal in the context of what else is happening in Antarctica.

  19. morganovich

    i see cooling from late 2001, just as i described.

    can you really not see that?

    i also see the GISS trend that makes new highs post 1998 when no other of the world data sets do.

    i also see that this chart is not up to date and omits the continued cooling through late 2007 and early 2008. so your trend is missing the most significant part of the new change. current temps, even from GISS, are lower than anything on that chart.

    an average will keep rising even after a series peaks. this is why it is a TRAILING indicator. it may be a good long term tool for trend analysis, but it is slow to register change. the statement “the globe has not warmed since 2001 based on this data” is factually accurate.

    given your measure, it would take a truly violent down move to disrupt the average trend in the near term. perhaps this is why the AGW folks like it. if the current temps stop showing warming, find me a series that does. but even the 5 year avg will begin to flip down of current temps persist. if that happens, what will you say?

    i am not, as you seem to feel, arguing that 6 years of temps disprove global warming. what i am saying is that since there has been no warming for 6 years, it’s unlikely that GW was the cause for this sudden ice shelf fracture. ice melts based on current temps, not temps 5 years ago, especially coastal ice which is so seasonally variable anyway and waxes and wanes annually to such a significant extent.

    much was made of the crack in the larsen b shelf a few years back. then ice went on to expand to record levels.

    most of an ice shelf floats. is some cases, thickening can make it crack off. so can tectonic activity on it’s land anchor. there are lots of potential explanations. but as the ocean has not warmed since 2003 and the atmosphere has not warmed since 2001, doesn’t it seem a bit off to you to attribute this event to warming? how do you justify that?

  20. Patrick Hadley

    Scientist I took a risk in challenging you to find all the alarmist websites that had dealt properly with Argo, Gouretski and Aqua – there must be some out there that I have not been able to find – please do try to locate one, I really would like to know what the “consensus” is saying. I suppose all they are all too busy making a big thing over some ice sheet calving to bother with important experimental findings. Considering 99% of scientists are all agreed about global warming they seem pretty thin on the ground when it comes to explaining away uncomfortable results from serious research.

    You still seem to miss the point that the Wilkin’s Ice Sheet (which is floating on the sea, but I won’t make too much of your error about that, although even the BBC did admit that in its blanket coverage of this story) is a very long way from the south pole, and since the peninsula is surrounded by water it is not exactly typical of the rest of Antarctica.

  21. dreamin

    By the way, there seems to be a meme going around that the temperature plateau over the past few years is not a plateau if you do a moving average and pull earlier temperature increases into the average.

    Well duh.

    Looks to me as though scientrollogist is getting a bit desperate.

  22. Papertiger

    Has anyone ever asked the scientist why he wants the world to end in a ball of fire?
    HESHEIT seems very determined, as if it isn’t just that he wants it to be true, HESHEIT seems to need it to be true.

  23. teqjack

    Gah. Or Gaia.

    Floes of this size happen once in a while, certainly within every five year period.

    Volcanoes or not, glaciers move! Thought experiment – picture glass 0.1inch thich by 3.0inch wide by 40feet long resting on a 40foot table: push it slowly off the table lengthwise: will all 40feet have to be pushed off before it breaks?

    Oh, and that Rhode Island (or Connecticut) thing, I am old enough to remember (vaguely) reports of such a floe (roughly 60miles by 40miles) off Greenland late Fifties or early Sixties, with speculation whether some of it would survive to strike the mainland. I don’t think it even reached major shipping lanes…

  24. Scientist

    morganovich – if you see cooling, you should get your eyes tested. Perhaps you are simply mentally drawing a line from the highest point you see to the lowest. Do you think that’s an appropriate treatment of the data?

    If you do linear fit the annual average temperatures from 2001 to 2007, you find an upward trend of 0.18°C per decade. If you do a linear fit to the monthly average temperatures from Jan 2001 to Feb 2008, you get an upward trend of 0.10°C per decade. If you do a linear fit to monthly averages from March 2002 (the month with the highest anomaly) to February 2008 (joint lowest anomaly), then you get an upward trend of 0.005°C per decade. So in other words, if you treat the data inappropriately by maximising the influence of internal variability, and cherry-picking your start and end dates, you can minimise the upward trend. Treat the data properly and the rise of roughly 0.2°C per decade which began in the mid 1970s is clear even when you just look at the last seven years of data. How can you disagree with this? The statement ‘the globe has not warmed since 2001 based on this data’ is not accurate.

    GISS temperatures include the poles. HADCrut, RSS and UAH don’t. El Niño is a tropical phenomenon. Therefore its magnitude as measured by GISS was lower than that measured by the other three.

    the crack in the larsen b shelf is a wonderful piece of misstatement. The Larsen B ice shelf, which had been there for 12,000 years, is not there any more. Its collapse was unprecedented during the Holocene.

    Patrick Hadley – so what if it’s a long way from the pole? That has no relevance whatsoever. What matters is how long it had been around before it collapsed. It’s 300km closer to the pole than Larsen B was though.

  25. morganovich

    the giss data also include all manner of adjustment not included in other data. claims that they are “more complete” are ludicrous. and how are you claiming they include the poles? the sites they use in the antarctic are predominantly on the penninsula.

    and don’t even get me started on linear fits. i used to write tech analysis programs for currency futures, so i know all these tricks. where is the data you are using to get this “linear fit” are you using the 5 year avg data?

    placing a spline curve (much better for fluctuating systems) in the data gets you this:

    http://icecap.us/images/uploads/0325bob-carter.jpg

    see the curve turn down?

    huh. cooling.

    and i did not say 2001-2007. i said since 2001. and yes, it is a selected bit of data. it says, from the recent maxima, look what has happened. we are talking about over 6 years here. if you can’t see that warming has stopped from jan 1 2002 to present, i don’t know what to tell you. it has. it’s not ambiguous in any of the data sets. you may argue it is not indicative of an end to AGW, but not that it has not happened. you simply don’t have a leg to stand on.

    and that period has had greatly reduced sunspot activity etc. the sun is driving it, just like it drove the warming. co2 is not a big factor. response lag means that a few year pass before the solar effects really turn up. but they are in the pipeline now to use a hansenism.

    so let me ask you a serious question:

    as a “scientist” surely you must accept that a theory can be invalidated by data. how much cooling do you need to see while co2 still rises before you give up on AGW. what metric would make you abandon this belief? surely there must be one. how much non-correlation of temps and co2 is needed? at what timescale? how much cooling would indicate the warming has stopped? how long would it need to go on?

    or are you wedded to this as a religion and not a scientific endeavor?

  26. Scientist

    Spline fits, well now there’s an opportunity for some trickery. Spline fits are totally ad hoc, by definition they cannot have any predictive value, you can make them look more or less however you want them to look by choosing the order and the number of knots, and they have absolutely no relation to physics. They are pure maths. You cannot derive a trend from a spline fit.

    Take the annual average temperature for 2001-2007 from any measure of it you care to choose. Do a linear fit. Tell me what you get. You talk in disbelief of a “linear fit”, apparently aghast that all you’ve been told about temperatures no longer rising is not true. All you have to do is check the data. You should do this whatever the claim you are faced with. If someone said temperatures were rising by 2°C per decade, you would check and you would find that they weren’t. Now someone says temperatures aren’t rising at all and you simply take that as fact. Check it and you’ll find it’s not true.

    Sunspot activity is irrelevant. If you can find an 11 year periodicity in the temperature record you’ll be finding something new. The solar cycle has a negligible effect on climate. Even a repeat of the Maunder minimum (estimated forcing -0.32W/m²) would not offset the effects of CO2 since 1750 (estimated forcing +1.7W/m²).

    All else being equal, the rate of rise of CO2 concentrations should lead to a warming of 0.2°C per decade or thereabouts. If all else remained equal, and the temperature trend became statistically significantly different from +0.2°C, then the effect of CO2 would need to be reconsidered.

    As a fool, is there any point at which the reality of human-induced warming will actually dawn on you? How much of a temperature rise are you waiting for? On all the evidence, it seems that no amount of evidence will persuade you.

  27. Vermont Weatherman

    Scientist-
    Since the data on tropospheric warming does not fit the GHG theory:

    Douglass, D.H., Pearson, B.D., Singer, S.F., 2004a. Altitude dependence of atmospheric temperature trends: Climate models versus observation. Geophysical Research Letters, 31, doi:10.1029/2004GL020103.

    Douglass, D.H., et al., 2004b. Disparity of tropospheric and surface temperature trends: new evidence. Geophysical Research Letters, 31, doi:10.1029/2004GL020212 .

    ….according to the scientific method, when observations and predictions do not match the hypothesis, then either the hypothesis needs changing, or the experiment needs to be redesigned. Considering that the data has been verified by HadCRU et al, it suggests it is time to re-work the hypothesis.

    Of course, we could consider the words of IPCC senior scientist Kevin Trenberth: “…there are no (climate) predictions by IPCC at all. And there never have been”; instead there are only “what if” projections of future climate that correspond to certain emissions scenarios. …..None of the models used by IPCC is initialized to the observed state and none of the climate states in the models corresponds even remotely to the current observed climate”.

    If that is the position of the IPCC, then instead of calling people “fool”, wouldn’t your time be better spent coming up with a hypothesis that fits the data, and in turn creates a computer model that resembles the observations?

  28. Vermont Weatherman

    Scientist-
    Since the data on tropospheric warming does not fit the GHG theory:

    Douglass, D.H., Pearson, B.D., Singer, S.F., 2004a. Altitude dependence of atmospheric temperature trends: Climate models versus observation. Geophysical Research Letters, 31, doi:10.1029/2004GL020103.

    Douglass, D.H., et al., 2004b. Disparity of tropospheric and surface temperature trends: new evidence. Geophysical Research Letters, 31, doi:10.1029/2004GL020212 .

    ….according to the scientific method, when observations and predictions do not match the hypothesis, then either the hypothesis needs changing, or the experiment needs to be redesigned. Considering that the data has been verified by HadCRU et al, it suggests it is time to re-work the hypothesis.

    Of course, we could consider the words of IPCC senior scientist Kevin Trenberth: “…there are no (climate) predictions by IPCC at all. And there never have been”; instead there are only “what if” projections of future climate that correspond to certain emissions scenarios. …..None of the models used by IPCC is initialized to the observed state and none of the climate states in the models corresponds even remotely to the current observed climate”.

    If that is the position of the IPCC, then instead of calling people “fool”, wouldn’t your time be better spent coming up with a hypothesis that fits the data, and in turn creates a computer model that resembles the observations?

  29. joshv

    Scientist,

    Has the rate of warming in the past decade decreased compared to the rate of warming in the previous decade?

    Also, I find it truly astonishing that you think straight lines are a better match for the underlying physics than something that looks like a sine curve.

    We know for a fact that climate oscillates, look at the last 100 years, it’s quite clear. You’d draw a straight line through all that and call it more predictive than morganovich’s spline?

  30. Scientist

    joshv – you can plot the trends yourself if you like – do one for the last thirty years, and one for the last seven years, and see whether their gradients differ. I’ll do this myself later this evening.

    Absolutely – a straight line is infinitely more predictive than a spline fit, because it’s mathematically meaningless to extend a spline curve beyond the available data. I’m guessing you don’t know exactly what a spline curve is, so here’s an explanation: a spline curve is made up of many polynomials which have the same y value and gradient at points called knots. The number of knots and the order of the polynomials can be chosen depending on how you want to fit the data. No knots means no fit. Extending individual polynomials beyond the knots between which they have been fitted is meaningless and will give you wrong results if you try it. A spline fit is a mathematical device for interpolating, not a scientific device for extrapolating. It is logically impossible to extrapolate a spline curve. It can’t be done.

    In contrast, when the only forcing that is changing significantly is changing in an approximately linear fashion, then a linear fit has a basis in physics and can usefully be extended beyond the available data.

  31. morganovich

    scientist, once more you reveal the woeful inadequacy of your reading comprehension. for the third time “since 2001″ means jan 2002 to present. you keep including 2001 in your numbers because it was a cold year. then you accuse me of cherrypicking. but i’m not falling into your trap.

    and what a spline does that a straight line cannot is show a change in trend. if, between time x and x+1 a trend changes, a straight line can NEVER capture this. even you must surely agree to that. i’m not going to get caught up in your ludicrous tactic of distorting everything everyone says, ignoring their caveats, and attempting to make straw man arguments against the resulting hyperbole. i’m just going to keep coming back to the data, and to what i said. do try to stay with the discussion.

    and where are you getting that sunspots don’t correlate to/lead climate nonsense? are you just taking radiative intensity into effect and ignoring solar wind/high level cloud relationships. low solar wind = high ionization of upper atmosphere = more high clouds = greater albedo = more cooling. so decreases in solar activity effect the earth climate in multiple ways.

    and where did you get the 1.7 w/m2 number. i’ll bet you used the old equations for forcing that assume an infinite atmosphere (or more likely, whoever you grabbed the figure from did).

    and why i am not surprised that you utterly duck the “what would change your mind” argument and start in with the mindless personal attacks. i have real doubts you have ever even considered what would invalidate your thesis.

    how do you explain the lesser warming compared to the surface observed in the atmosphere at 9000m (300 pascals) where the GH effect takes place if this warming is GH driven? how do you explain that temps dropped for a 50 year period in the last century in the US per the NOAA data while co2 rose all the while? 50 years of weather?

    i’ll even one up you and go first:

    show me conclusive empirical evidence that CO2 has caused significant warming by demonstrating that it is a leading variable to climate changes (eg co2 rises then temperature rises) with a strong correlation (.75 r2 or better, both of which can be exceeded by ocean currents and lagged solar) over any meaningful timeframe (techtonic, intergalcial, 100 year, whatever). then show me that positive feedbacks exist around such forcing that will cause it to accelerate and become worrying. i have never even seen convincing correlation between temperatures and co2 apart from the early vostok mistakes (from too large a timescale to see that co2 was a lagging, not a concurrent variable).

    show me that and i will start to listen seriously to you.

    i challenged you before to show me one conclusive piece of empirical evidence that CO2 has been a precipitator of climate change as opposed to a concurrent variable of very limited impact.

    you did not.

    so i challenge you again.

    i’m open to data.

    convince me.

    and convince me you are open to data too. how bad a fit to climate must co2 have to not be considered a primary variable? .3 r2? what?

    how much would the world need to cool and for how long before you give up on the warming trend? if you insist on anchoring warming in 1890, a straight line will have positive slope for a helluva long time. and we should be glad it does, at the late 1800′s were a very cold time with a much more difficult climate for humans than currently prevails.

    so what would end a warming trend? it’s a simple question. what data would cause you to say the warming trend has ended/abated?

    put up or shut up.

  32. Scientist

    Vermont Weatherman – the two papers you mention truncated their data at 1996. As they were published in 2004 there was no good reason to truncate the data there. Want to make a small bet as to how the same analysis would turn out if you included data since 1996?

  33. morganovich

    ps.

    here’s a graph of all 4 world temp sets since jan 1 2002.

    draw all the straight lines you want. 6 years. no warming.

    http://wattsupwiththat.files.wordpress.com/2008/03/ttttpart1figure3.png

    and you are dead wrong on the solar forcing change being less than 1.7w/m2.

    solar irrad in mid 1600′s was 1363.5 w/m2 it rose to 1366.5 w/m2 in 2001.

    now, i’m no “scientist” but i make that out to be 3w/m2 increase. and i’m pretty sure 3 >1.7…

    and if you really want to dig in, read this analysis subjecting recent temperatures to the chow test (sum of mean square error) to see if there was a trend breakpoint and see if you still think straight line is the way to go. looks like the old line broke to me.

    http://wattsupwiththat.wordpress.com/2008/03/13/to-tell-the-truth-will-the-real-global-average-temperature trend-please-rise-part-2/

  34. Scientist

    A spline curve is a mathematical convenience and has no basis in physics.

    In normal English usage, ‘since x‘ does not mean ‘starting from x+1‘. If you want to be understood, use language properly.

    Fitting a straight line to annual average temperatures from 2002-2007 yields an upward trend of +0.09°C per decade. It is obviously incredibly hard for you to get your head around the fact that even if you egregiously cherry pick the data, it doesn’t show what you want it to show.

    Why do I think the solar cycle doesn’t affect climate? Because there is no eleven year periodicity in global average temperatures. For crying out loud, do you never even try to understand, or look at the data?

    How to explain 50 years of cooling in the US when CO2 was rising.. well let me see.. firstly will you and all the other morons who frequently get confused on this point please get it into your heads that the US is not the world. We are talking about global warming, are we not? Now for another reason… maybe it’s because CO2 is not the only thing that affects climate? It’s slightly pathetic to think that temperatures should rise in absolute lock step with CO2. Where did you get the idea that they should? The obvious return question is how do you explain the observed temperatures over the last 100 years if CO2 has no influence on temperature?

    I’ve already told you what data would cause me to say the warming trend has ended, you dope. Go back and read. And I’m not going to spoonfeed you the facts about CO2 as a climate driver. Do a search for ‘paleocene-eocene thermal maximum’ and follow some links, if you’re capable of doing so. I see no evidence at all that you are open to data, as you claim. I mean, you can’t even comprehend a simple linear fit to six years of temperature data! I doubt you’re going to understand anything really scientific.

    Oh yes, as I was typing I see you found yourself another brick wall to run into. You are no scientist, clearly, because you didn’t realise that the important figure with solar forcing is the insolation at ground level averaged over the globe (you knew that Earth is spherical, right?), not the simple value of the solar constant at Earth’s orbit. Taking your numbers for the solar constant and an albedo of 0.39, the forcing is -0.5W/m². Now, is that more or less than 1.7?

  35. josh

    Scientist, follow your straight line out 500 years and tell me how predictive it is. There is no mathematical meaning to extending any curve beyond it’s endpoint – it’s just guesswork. All a fitted curve can tell you is something about the properties of the data you have.

    If you are lucky your curve encodes some basic physical principles of the system and wonder of wonders it applies to new observations as well. Perhaps you could clue me in to the physical principles of global climate encoded in your straight line.

    I am not particularly mathematically enamored with morganovich’s spline, thought it is a much better fit than your straight line, but you could certainly do a fourier transform of the temperature time series and get an even better fit. This curve would at least encode the frequencies components of past climate variations. Assuming those frequencies extend into the future is a much better bet than assuming a linear increase for the next 100 years.

  36. Scientist

    I never claimed that you could extrapolate over 500 years from a simple linear fit. What I said was that a linear fit has some basis in physics, given the current forcings, and that a spline fit is purely a tool for interpolation with no possibility of extrapolating it. A more rigorous approach obviously involves computer codes describing realistic models of the climate system.

  37. morganovich

    scientist-

    you are a joke.

    you go back and pick nits about wording to to duck the issue yet again.

    read the posts. i made very clear that i was talking about 1/2002 on several times, but you ignore it repeatedly.

    then provide no backup for your data.

    then you make the spurious comparison of albedo influnced solar forcing to an unsubstantiated (an likely incorrect) statement about co2 that takes no limits or feedback to that system into account. that’s apples and oranges you charlatan.

    but to make you look even stupider, those numbers ARE AT THE SURFACE. it’s forcing data. albedo is already accounted for. what, you think we had satellite monitoring in 1600… clown.

    remember the argument you lost about feedback (a mechanism which you clearly don’t understand)? CO2 forcing is system limited too.

    besides, if so much radiation is getting caught in the atmosphere, why is it warming less that the surface? where is the GH fingerprint? you sure dodge that question a lot. because you know there is no answer and it wrecks your whole argument.

    and how can you possibly substantiate your claim about spline curves and physics? your claim is akin to throwing a ball in the air watching it reach an apex, come halfway down, drawing a line between the thrower and that point and claiming the ball is still going up.

    spline fits show changes in trend. single lines do not. this is an unarguable fact. have you ever studied basic, newtonian physics and or statistics? if so, your teachers must be horribly embarrassed you are pathetic the way you cling to “you can’t draw it forward so it shows nothing”. we’re not even talking about predictive value, we’re talking about assessing things that have already happened. you don’t even make the rudiments of sense.

    so a straight line has a “basis in physics” but a curve does not? where do you get this drivel? a straight line is pure math too. not many of those in nature. do you even listen to yourself?

    you obviously understand absolutely nothing about trend analysis. the reason to use a curve and not a line is so you can see the trend change. the zero derivative is a dead giveaway…

    and for what it’s worth, i do in fact have a science degree (physics) among several others including a bunch of statistics work from an econ degree. i was just making fun of your obvious trouble grasping basic scientific principles. what kind of scientist do you claim to be anyhow? you’ve never revealed that.

    and if you think solar has no influence on temperature, then how can you possibly claim that co2 does? solar has over twice the r2 to temp over the last 100 years that co2 does. i’ve show that work to you enough times that i’m not going to bother linking it again as you obviously don’t/can’t read.

    but heres a chart of solar forcing. gee, wonder why it’s been warming…

    http://www.junkscience.com/Greenhouse/irradiance.gif

    and there’s no 11 year cycle because the earth takes about 8 years to adjust to solar changes. so it damps out such a short cycle. more benefits of negative feedback stabilizing climate.

    and where did you tell me what it would take for you to give up your AGW religion?

    nothing you posted since i asked even touches on that. liar.

    and again, no empirical data on co2 being a primary climate driver.

    keep in mind, that in the face of large and incontrovertible natural climate variability in the past, it is those claiming that humans are causing the current changes that have a case to prove here. assuming AGW as the base case unless someone can prove otherwise is unreasonable and unscientific. but i’m sure such an august “scientist” as yourself would already know that…

    you are a useless troll.

  38. Scientist

    Only time now to address two points. Firstly, I really would have thought someone with a degree in physics would know the difference between a flat area with no atmosphere facing the sun, and a rotating sphere with an atmosphere. The numbers you quoted are not the solar irradiance at the surface, and you look very, very stupid.

    And secondly, your chart of solar irradiance shows that if you average over the cycle it’s been constant since 1951. You claim that it takes eight years for the earth to adjust to solar changes. So, temperatures should have been flat since 1958, unless some other factor was at work. I wonder just what that factor could be?

    Oh and why not, one more thing: spline fits don’t show changes in trend. They are a simple interpolative tool, and exactly what they show depends on what order polynomial you choose, and how many knots. I could generate a spline fit to the temperature data that closely tracks the actual data by using high order polynomials and monthly knots. That would tell us nothing about trends. Your spline fit tells us nothing about trends.

  39. morganovich

    are you really that dumb?

    solar heat is flat, to temps stop?

    you must be being willfully ignorant here. no one is that dumb.

    put a pot of water on the stove. turn up the heat. if you can show that the water stops heating once you stop turning up the heat, go get your nobel prize. it will keep warming until it reaches a max potential. higher heat makes it warm faster. this is so elementary i can’t believe you don’t get it. you know less science than al gore or most school kids.

    and as i said, these are FORCING numbers. the w/m2 unit measure is a dead giveaway there. albedo is taken into effect. you are ridiculous.

    and for the last time: straight line interpolation is useless for determining a trend change. they are not differentiatable. if a trend is changing, a stright line cannot tell us anything about that. and as we are discussing a change in trend (as shown by the prior linear trend breaking with .05 or so avg p value of break by chow test) there is no basis for using linear analysis except to OBSCURE what is going on.

    which seems to be your intent…

  40. Scientist

    Fucking hell, you’re stupid. You said yourself you thought eight years was the lag between solar changes and temperature. How, then, have temperatures carried on rising since 1959 when they should have stopped?

    And yes, the units of forcing are W/m². Lots of other quantities also have the same units. The numbers you quoted are not forcing. They are values of the solar constant. Look it up.

    And you can’t differentiate a straight line? You really don’t know any maths!

    You really do epitomise the standard of debate I’ve come to expect here. You make incredibly basic errors that you just don’t have the capability to understand, you don’t do even the most basic checking of the things you are quoting, and you will not deviate from something you’ve chosen to think no matter how ridiculous it is.

  41. morganovich

    hahahahahaha. you are a such a moron. add heat to a system at a constant level, and it keeps warming until it reaches capacity. add more heat, it warms more quickly. did you try the pot on the stove yet? do you need to keep turning the gas up to make water boil.

    the climate system has response inertia. it takes a period of time to overcome that. but once you do, it keeps moving until a new equilibrium is reached reflecting the new inputs. water can warm for a very long time and hold a lot of heat if you give it slow, steady energy. i really can’t believe you can not understand this basic fact. did you get your degree at al gore university?

    a straight line has a constant derivative moron. therefore it tells us nothing. besides, i said “straight line interpolation”, again your reading comp fails you and your ludicrous straw man proves meritless.

    Linear interpolation on a set of data points (x_0, y_0),\, (x_1, y_1),\,\dots,\,(x_n, y_n) is defined as the concatenation of linear interpolants between each pair of data points. This results in a continuous curve, with a discontinuous derivative, thus of differentiability class C0.

    If a C0 function is insufficient, for example if the process that has produced the data points is known be smoother than C0, it is common to replace linear interpolation with spline interpolation, or even polynomial interpolation in some cases.

    you really do believe that a ball thrown up in the air that reaches an apex and comes halfway down is still in an uptrend because the straight line from origin to current position is angled up, don’t you?

    i give up with you. i had thought you were just ignorant. but it’s obviously a much deeper issue.

    unless, of course, you finally want to show me the CO2 evidence i keep asking for.

    thought not…

  42. DKN

    Maybe it is mentioned in the comments and I missed it, but the linked article says:

    “We believe the Wilkins has been in place for at least a few hundred years, but warm air and exposure to ocean waves are causing it to break up,” Scambos said:

    Doesn’t that kinda imply that within historical times the Wilkins was *not* “in place,” presumably because of “…warm air and exposure to ocean waves…” at least comparable to those of today? And, if so, then isn’t it plausible that current conditions are not drastically unusual, or indisputably determined by human actions?

  43. morganovich

    and because i know you will pick this nit to avoid substance once more, i should have said “tells us nothing about a change in trend”. given your track record, i doubt it is reasonable to expect you to recall antecedents from prior discussions…

  44. Keith

    Scientist, what data are you using to support your statement, “If you do a linear fit to the monthly average temperatures from Jan 2001 to Feb 2008, you get an upward trend of 0.10°C per decade.”

    If you go to http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts+dSST.txt, plot the monthly anomaly data from Jan. 2001 through Feb. 2008, and add a linear regression fit, you get a slope of .0024 deg C per year or .024 deg C per decade. The satellite data from RSS or IAH show an equally insignificant amount of cooling over the same time span.

  45. Dean Clark

    Hey Scientist… all I see you do is belittle people, call them morons and say they have no “scientific background”. I would like to read your thoughts but can’t get over your belittling of people. This is so typical of the GW crowd… belittle the opposition and discredit them by name calling. The folks who do rebuttal you are not belittling you and are discussing. Perhaps you should follow their lead… it would make your arguments at least bearable to read.

  46. Vermont Weatherman

    Scientist-
    The atmospheric physics changed after 1994? I’d love to see the paper(s) that refute the studies I mentioned.

  47. Scientist

    Keith – I was using this data. If you use the land and sea, but use annual means instead of monthly means (and this is sensible because it’s the trend we’re interested in, not the noisy internal variation), then you get an upward trend of +0.04°C per decade.

    Lest it need reiterating, we can tell almost nothing from just five years of data, but sensible treatments of the data show upward trends consistent with the previous trends. If you only choose the noisiest data, and cherry-pick the data so that it starts with a particularly warm month and ends with a particularly cold month, then you can derive a linear fit which is virtually flat, but such an exercise is basically meaningless.

    Dean Clark – plenty of people here are highly abusive towards anyone who points out basic flaws in their arguments. Anyone who is not abusive, I’ll have a perfectly civil conversation with. Anyone who is abusive, I respond in kind. Fair enough, no?

    Vermont weatherman – if data after 1996 is analysed, the conclusions derived from pre-1996 data only are not supported. Incidentally, model predictions of a troposphere warming faster than the surface are not specific to greenhouse gas forcing scenarios. Increased solar activity would have the same effect. The difference is that increasing greenhouse gases should cause the stratosphere to cool, while increasing solar insolation would cause the stratosphere to warm. The stratosphere is definitely cooling, and actually that makes it quite difficult to determine the actual temperature of the troposphere from satellites, because the channel which measures the temperature of the troposphere has some overlap with the stratosphere, introducing a cool bias.

    A couple of interesting papers: Amplification of Surface Temperature Trends and Variability in the Tropical Atmosphere, Contributions of Anthropogenic and Natural Forcing to Recent Tropopause Height Changes; and a discussion at realclimate.

  48. morganovich

    the first paper sounds just like the argo: the observations mostly disagree with the models, but again we’re sure it’s the observations that are incorrect.

    how many times can that argument be used before the conclusion becomes either: 1. the models are wrong or 2. we are terrible at measuring this stuff? (and i note the former is much more plausible)

    “On multidecadal time scales, tropospheric amplification of surface warming is a robust feature of model simulations, but it occurs in only one observational data set. Other observations show weak, or even negative, amplification. These results suggest either that different physical mechanisms control amplification processes on monthly and decadal time scales, and models fail to capture such behavior; or (more plausibly) that residual errors in several observational data sets used here affect their representation of long-term trends.”

    so it’s “robust” in models but absent from most of the data. this must be a new definition of “robust” that i am not familiar with.

    and how fortunate that radiosondes do not share your alleged issues with tropospheric monitoring and that they also show a lack of AGW fingerprint through direct measurement. or is there a problem with their thermometers as well?

    regarding the second, i have a question about how they make a jump of logic:

    “A model-predicted fingerprint of tropopause height changes is statistically detectable in two different observational (“reanalysis”) data sets. This positive detection result allows us to attribute overall tropopause height changes to a combination of anthropogenic and natural external forcings, with the anthropogenic component predominating.”

    ok, so they predicted a change, it seems to have happened, but how to they get from there to the attribution of the anthropogenic component dominating? this seems like the same bad assumption the IPCC uses that change is caused by humans until otherwise explained.

    where is the evidence that humans did this? i would be very interested to see that. how is this “predominant” attribution arrived at?

    and go back and read the comments, no one on this thread is abusive to you until you start it. (see your post 3/27 9.42am)

    still waiting on that CO2 evidence…

Comments are closed.