Lamb on the Northeast Atlantic

Mann and Jones have a new post at realclimate discussing warmth at Svalbard, an island archipelago north of Norway. May temperatures are very warm there. They say that the differences are 5 standard deviations based in i.i.d. We’ve talked about the inappropriateness of i.i.d. assumptions in the context of Rasmus (BTW what ever happened to Rasmus over there? – I predicted that Gavin would ice him and this seems to have happened.) Terry drew this article to my attention. He attempted to post a comment asking about autocorrelation, but this seems to have been censored by realclimate. After all, that would be a “serious” discussion.

Mann and Jones say that “models show” "polar amplification” (although I didn’t notice any discussion of declining temperatures in Antarctica.) I’m not going to go through the statistics, but I thought that I’d post up some interesting comments from Hubert Lamb [1979] on the northeast Atlantic.

Lamb’s abstract says:

“Variations must take place in the ocean circulation when the general wind circulation varies. There are hints even within recent years that variations in the ocean between Iceland and Scotland and Norway can be big. The area has been regarded as the main path of the warm saline North Atlantic Drift water heading towards the Arctic; but when the polar water occasionally intrudes from the north, sea surface temperature is liable to fall by 3 to 5 deg C and presumably by more than this when, as in 1888, the ice advanced to near the Faeroe Islands.

The reconstruction of the situation between AD1675 and 1705 resulting from this study suggests a probable mean departure from modern values between the Faeroes and southeast Iceland amounting to about –5 deg C and at the climax in 1695, the polar water seems to have spread all around Iceland; across the entire surface of the Norwegian Sea to Norway and south to near Shetland. Support for this diagnosis is found in a considerable variety of reports of environmental conditions existing at the time in Scotland, south of Norway and elsewhere. The enhanced thermal gradient between approximately latitudes 55 and 65 N during the Little Ice Age, which this result indicates, offers an explanation for the occurrence in that period of a number of windstorms which changed the coasts in various places and seem to have surpassed in intensity the worst experienced in the region in more recent times.

Lamb’s article itself is an interesting discussion of wind circulation regimes in the north — a continuing theme in Lamb’s work. Lamb points out that there is a far larger flow of water poleward through the Atlantic than the Pacific (about 10 times more according to the estimates of the time). Well before Broecker’s Conveyor Belt, he pointed out interestingly that the coast of Brazil is like a wedge in the course of the South Equatorial Current and a shift of only 1-2 deg of latitude of this current to the nose of Brazil would affect the warmth and volume of the Gulf Stream (citing Brooks 1949), reporting that shifts of 1-2 degree latitude in wind circulation do occur.

He states that the volume of water transported southward by the East Greenland Current can vary by a factor of 10 (a normal distribution– anyone?). The East Iceland Current, the branch which heads southward around the east side of Iceland, is said to vary greatly both in volume.
Lamb reports that the Faeroe Islands (62N, 6.8W), which are to the southeast of Svalbard (73N, 20E) have a long series of SST temperatures, which “seems to be an interplay between the contrasting water masses in this sensitive area” and that the differences in 5-year means was twice as great as in C England air temperatures.

Lamb suggests that the cod fishery can serve as a proxy for SST since the “abundance of cod seems to be limited by the 2 deg C isotherm”. He then describes failures of the cod fishery in the Little Ice Age, relating this to changes in ocean circulation. He mentions that the worst years in the region seem to have been the 1690s, particularly 1695 when the sea ice surrounded Iceland.
He says that “between 1675 and 1700, the water temperatures prevailing about the Faeroe Islands presumably were on overall average 4 to 5 deg C below the average of the last 100 years, an anomaly 4 or 5 times as great as that shown by the thermometer observations in central England where the coldest decade (1690s) averaged about 1.5 deg C below the warmest decades in the earlier part of this century”

In passing, some of you may remember the peculiar truncation in MBH98 of the Central England temperature series to exclude the late 17th century – the very period which Lamb ascribes as being the worst of the Little Ice Age. In the Mann Corrigendum, he justified this on the basis that he used data from Bradley and Jones 1993. But this simply passes the question on: what business did Bradley and Jones 1993 have in truncating the late 17th century portion of this record - in an article purporting to show the non-existence of the Little Age?

Lamb also reported interesting information from travellers about depressed snowlines in Scotland in the 17th and 18th centuries, saying that “the apparent lowering of the snow line by 300 to 400 m in the Little Ice Age would imply an average temperature level 2 to 2.5 deg C below that of the mid-20th century, an anomaly rather more than twice as great as central England, but one that could be explained by the lower temperatures which we derive for the ocean within 500 km to the north”. He cited tree ring estimates from Norway.

He notes that Stolle 1975 had deduced that already in 1577 the Gulf Stream was turning east away from the American cost on a more southern track than has been normal in this century and suggested that movements continued even further south. For Greenland, he suggests (Lamb 1977) that the water temperatures in the fjords of southwest Greenland around 1000 AD at least sometimes reached values not less than 4 deg C warmer than the in the warmest part of the present century, concluding that this must have characterized the open seas occupied by the southernmost extension of the East Greenland Current. He then dates the advance of sea ice and decline of Greenland, with difficulty commencing by 1250 and regular communication ceasing by 1410. In contrast, he says that in the eastern part of the Norwegian Sea, there was little adverse impact until about 1550.

So what do we have here: in the Faeroe Islands, Lamb says that the fluctuations are far more marked than in Central England. Svalbard (73N, 20E) is to the northeast of Faeroe (62N, 6.8 W). So there’s every reason to believe that it would have fluctuations on the same or greater scale as Faeroe. So do high May temperatures in Svalbard prove that temperatures in this region were warmer than in the MWP? If so, then why haven’t the treelines in Finland and Siberia reached MWP levels? OK, maybe there’s a delay and maybe the treelines based on present temperatures will reach MWP levels. But they haven’t so far. So there must have been a sustained period of warmth in the past which permitted trees to thrive at higher altitudes in the MWP "locally" in Siberia, Finland and California and go further north in the Canadian Northwest Territories.

Lamb is a terrific read. He’s a fine writer with an eye for detail. The only reason his work fell out of favor was because it supposedly became obsolete with pseudo-quantitative Mannian multiproxy studies, which claimed the ability to make annual reconstructions. Now they seem to be resiling from that and Wahl and Ammann say that they are only trying for “low-frequency”. In any event, if you believe, as I do, that the entire Hockey Team corpus is of little merit, then it’s time to re-read Lamb and see what he actually says.

Reference: Hubert Lamb, 1979. Climatic variation and changes in the wind and ocean circulation: the Little Ice Age in the Northeast Atlantic, Quat. Res 11, 1-20.


198 Comments

  1. ET SidViscous
    Posted May 23, 2006 at 9:48 AM | Permalink

    " I predicted that Gavin would ice him and this seems to have happened."

    Sorry Steve, but you seem to be doing a bit of Mixed Metaphor here.

    This is of course the Hockey Team, therefore to "Ice" him would be to imply that he is in play and out there on the Ice.

    I don’t think PEnalty box is a good analogy as it implies a referee has put him there.

    In baseball the term is "to bench" I’m not sure what the Hockey equivalent is.

    Steve: In this case, he’s either been cut from the roster or sent to the minor leagues.

  2. John Hekman
    Posted May 23, 2006 at 10:11 AM | Permalink

    Not only are temps in Antartica lower; the temp on Greenland’s inland ice sheet is lower. And 90% of the world’s ice happens to be in those areas. Artic temps are up, but are not as high as they were in the 1930’s. How many s.d.’s were they from the mean in the Thirties?

    The hockey team is not scoring any goals. They think they scored a lot in the past. It’s as if the Maple Leafs didn’t show up in the NHL but only had press conferences to show off all the trophies they have won in the past (when there were only six teams in the league–just kidding, Steve).

  3. Steve Bloom
    Posted May 23, 2006 at 11:39 AM | Permalink

    From the RC article: “The numbers are fairly remarkable. April’06 was warmer than any previously recorded May, and January ’06 was warmer than any previously recorded April.” Exactly how many standard deviations are involved here is perhaps interesting, but the figures seem to speak for themselves. BTW, Steve M., that “Antarctic cooling” info is a little out of date.

    Elsewhere, it seems scientists have been… auditing the climate.

  4. Spence_UK
    Posted May 23, 2006 at 11:51 AM | Permalink

    I like that first BBC link you’ve got there Steve:

    To try to resolve the conundrum, the BAS team compared the data with 20 simulations of the climate over the last century.

    The team found that in all cases, the models failed to simulate the rise.

    Imagine that! Climate models not actually representing how the climate behaves today, let alone in 100 years time. Who’d have thunk it?

    Your second link is nothing more than an extrapolation of recent temperature… which is even more ludicrous than using a GCM. I hope I don’t have to explain why.

  5. Steve Bloom
    Posted May 23, 2006 at 11:57 AM | Permalink

    Yes, Spence, the models have never been thought to be anywhere close to perfect when it comes to the Antarctic. Apparently this is news to you.

    Just an “extrapolation of recent temperature”? Try looking a bit more carefully.

  6. John Lish
    Posted May 23, 2006 at 12:06 PM | Permalink

    #1 – Steve I thought I saw a VOG comment by Rasmus on the Al Gore thread in response to Roger Pielke?

  7. Steve Bloom
    Posted May 23, 2006 at 12:12 PM | Permalink

    Oh, and Steve M., on the subject of your strange snark regarding Rasmus: I know last Wednesday was a very long time ago, but notice the byline here. Rasmus has also remained an active participant within threads. But doubtless you have some statistical proof showing that he is in fact entirely absent.

  8. Spence_UK
    Posted May 23, 2006 at 12:13 PM | Permalink

    Yes, Spence, the models have never been thought to be anywhere close to perfect when it comes to the Antarctic. Apparently this is news to you.

    No, it isn’t news to me. I know that when models are taken outside of their comfort zone of incestuous hindcasting on the data their physical behaviour is trained upon, they give completely wrong results. So any spatio-temporal move outside this comfort zone results in wrong answers. I understood that much years ago, and it hasn’t been proven wrong yet.

    Just an “extrapolation of recent temperature”? Try looking a bit more carefully.

    Sorry, “just an extrapolation of recent temperature which has been dressed up as if it isn’t”. Hope that helps.

  9. Michael Jankowski
    Posted May 23, 2006 at 12:30 PM | Permalink

    Re #3: First link headline is “Winter air temperatures over Antarctica have risen by more than 2C in the last 30 years, a new study shows.”

    So that puts the starting point back in the cool middle of the already cool 70s (well, it was at least as cool and often cooler than the 30s, 40s, 50s and 60s!), when some (not all, of course) feared were were headed to an ice age. What a great bookend.

  10. Steve Bloom
    Posted May 23, 2006 at 1:01 PM | Permalink

    Well, Spence, I guess I’ll just have to adjust to the concept of the last 400,000 years as being “recent.” But I do like “incestuous hindcasting”! Not only does it roll trippingly off the tongue, it shows once again how little point there is in having these discussions with someone who is a priori convinced that the entirety of climate science is one giant conspiracy theory. You don’t like the models, you don’t like the theory, and now it turns out you don’t even like the observations.

  11. jae
    Posted May 23, 2006 at 1:07 PM | Permalink

    Rasmus has also remained an active participant within threads. But doubtless you have some statistical proof showing that he is in fact entirely absent.

    Hey, Bloom: it’s sad that you take so much pride in this kind of smart-assed commenting. What good does it do, anyway?

  12. Steve Bloom
    Posted May 23, 2006 at 1:23 PM | Permalink

    Er, jae, you might want to ask our host that question first. I’ll answer after he does.

  13. JerryB
    Posted May 23, 2006 at 1:45 PM | Permalink

    If temps at Svalbard keep climbing, they may eventually catch up to some of Isfjord Radio’s highs from the 1950s. :-)

    see: http://data.giss.nasa.gov/cgi-bin/gistemp/findstation.py?lat=78.25&lon=15.47&datatype=gistemp&data_set=0

    BTW, the article at http://news.bbc.co.uk/2/hi/science/nature/4857832.stm refers to winter temps only, at a particular altitude only, over only a 32 year period. More than a little bit of reaching going on there.

  14. Ken Robinson
    Posted May 23, 2006 at 2:16 PM | Permalink

    Might it be possible to hear someone comment on the subject of the thread please? I’ve never heard of Lamb’s work and would be interested in learning more about it from anyone with an informed opinion. If you’ll pardon the expression, we need less heat and more light.

    Regards;

  15. PHEaston
    Posted May 23, 2006 at 2:20 PM | Permalink

    Climate Chaos! – More from the BBC – a season on ‘climate change':

    http://www.bbc.co.uk/sn/hottopics/climatechange/programmes1.shtml

    The first programme is presented by the great David Attenborough, so it will be very interesting to see what he has to say. No doubt, he has observed climate change during his decades as a wildlife broadcaster. The reality is that for any given period of time, the climate is far more likely to be changing than not changing. I just hope he will provide a balanced, scientific judgement.

    Are We Changing Planet Earth?
    Wednesday 24 May, 9pm, BBC One
    David Attenborough draws on his life-long insights into our planet and presents his personal take on climate change. Part two follows next week.

    Test the Nation – Know Your Planet
    Sunday 28 May, 8pm, BBC One
    Are you aware of climate and environmental issues? We put the country to the test in the popular quiz show.

    Can We Save Planet Earth?
    Thursday 1 June, 9pm, BBC One
    Part two of David Attenborough’s investigation.

    Five Disasters Waiting to Happen
    Tuesday 6 June, 9pm, BBC Two
    We examine five global locations and scenarios: London, Shanghai, Mumbai, Paris and Tuvalu. All have been identified by experts as vulnerable to the effects of climate change.

    The Money Programme
    Friday 2 June, 7pm, BBC Two
    The Money Programme spends a week with a family in Teesdale – the area with the UK’s highest CO2 emissions per capita.

    Panorama
    Date and time TBC, BBC One
    The Bush administration has resisted calls to engage in Kyoto, and has been accused of a systematic campaign of disinformation and harassment against the scientific community – gagging scientists, re-writing major reports, and allowing the oil and coal industries to drive policy. Panorama investigates these claims.

  16. PHEaston
    Posted May 23, 2006 at 2:32 PM | Permalink

    The UK’s Independent newspaper had a very interesting article on icebergs in the North Atlantic. (The Independent – ironically – is as convinced of the case for man-made global warming as they were convinced that the case for WMD in Iraq was hogwash derived from the political distortion of expert opinion).

    http://news.independent.co.uk/environment/article548824.ece

    A very seasoned observer of icebergs in the North Atlantic explains how the number of icebergs can vary significantly, but with no-one knowing whether global warming will mean more icebergs or less.

    The most telling quote from this article is:
    “Back in the mid-1990s, when we had thousands of icebergs, I got a call from Japanese TV who wanted to do a story on us because they believed the large number of icebergs was indicative of global warming,” he says. “Then, in 1999, we had only 22 icebergs and I got a call from a European TV company who wanted to do a story because they were certain that the fact that there were only 22 bergs in the shipping lanes was a clear indication of global warming.”

  17. Steve McIntyre
    Posted May 23, 2006 at 2:42 PM | Permalink

    #14. Hubert Lamb is one of the most famous climatologists, if not the most famous, in the period that we primarily discuss here. His work in the 1950s through the 1980s argued that small changes in climate on geological terms (i.e. not an Ice Age) had measureable impacts on humans and that the changes could be analyzed. He wrote about the Little Ice Age and the Medieval Warm Period without calling them “so-called”. His viewpoint was the dominant paradigm before the Hockey Team took up the project of “getting rid” of the Medieval Warm Period.

  18. Ken Robinson
    Posted May 23, 2006 at 2:49 PM | Permalink

    Re: 17

    Thank you Steve. I’ll look into his work further.

    Regards;

  19. Peter Hearnden
    Posted May 23, 2006 at 3:22 PM | Permalink

    Hubert Lamb also said the following (from ‘Weather, climate and human affairs, a book of essays and other papers, Routledge 1988):

    page 345 “The future increase of CO2 expected and volcanic varition

    The atmosphereic carbon dioxide is widely expected to double during the course of the twenty-first century, though the time taken would be affected by possible moderation of the rate at whcih humanity burns coal and oil. Expectations of what the climate effect should be vary, though a resulting rise of world temperature oby 1.5-3C – much more than this in polar regions…”

    and (last page of book (351))

    “It must be understood in connection with future climate that there is no necessary contradiction between forecasts of 1) continued or renewed cooling over the next few decades due to either volcanic activity or solar output changes (or both*); 2) a rather strong warming, lasting some centuries, due to increase of CO2 and other pollution due to human activity…

    The knowledge that we now have as a result of climate studies contributes strongly to the dawning of appreciation of our responsibility to care for the future of the environment which the Earth provids for mankind, for the animal kingdowm, and for the planet”

    book ends.

    I agree with him!

    ( though *this clearly didn’t happen)

  20. Spence_UK
    Posted May 23, 2006 at 3:49 PM | Permalink

    Re #14

    My apologies Ken – you are quite right, sometimes I’m too easily dragged in to these type of arguments and they serve no value. I need to remember this isn’t sci.environment!

    Lamb’s work dates back to a time when the science of climatology was less polarised and provides a great deal of insight into recent climate. His book, “Climate, History and the Modern World” is well worth a read if you can get hold of a copy.

  21. Francois Ouellette
    Posted May 23, 2006 at 5:13 PM | Permalink

    Maybe off topic, but I got this article from Benny Peiser:

    Quaternary Research. Volume 65, Issue 3 , May 2006, Pages 431-442

    (http://tinyurl.com/fu3ou)

    A multi-proxy lacustrine record of Holocene climate change on northeastern Baffin Island, Arctic Canada

    Jason P. Briner, Neal Michelutti, Donna R. Francis, Gifford H. Miller, Yarrow Axford, Matthew J. Wooller and Alexander P. Wolfe

    Abstract

    Reconstructions of past environmental changes are critical for understanding the natural variability of Earth’s climate system and for providing a context for present and future global change. Radiocarbon-dated lake sediments from Lake CF3, northeastern Baffin Island, Arctic Canada, are used to reconstruct past environmental conditions over the last 11,200 years. Numerous proxies, including chironomid-inferred July air temperatures, diatom-inferred lakewater pH, and sediment organic matter, reveal a pronounced Holocene thermal maximum as much as 5°C warmer than historic summer temperatures from 10,000 to 8500 cal yr B.P. Following rapid cooling 8500 cal yr B.P., Lake CF3 proxies indicate cooling through the late Holocene. At many sites in northeastern Canada, the Holocene thermal maximum occurred later than at Lake CF3; this late onset of Holocene warmth is generally attributed to the impacts of the decaying Laurentide Ice Sheet on early holocene temperatures in northeastern Canada. However, the lacustrine proxies in Lake CF3 apparently responded to insolation-driven warmth, despite the proximity of Lake CF3 to the Laurentide Ice Sheet and its meltwater. The magnitude and timing of the Holocene thermal maximum at Lake CF3 indicate that temperatures and environmental conditions at this site are highly sensitive to changes in radiative forcing.

  22. Francois Ouellette
    Posted May 23, 2006 at 5:17 PM | Permalink

    #3: Wasn’t last winter abnormally cold in north-eastern Europe (especially Russia)?

    (though I don’t see the point in discussing anecdotal weather events. e.g the link in post #13 shows that temperatures vary wildly (as much as 8 C)from year to year in a particular station)

  23. Tim Ball
    Posted May 23, 2006 at 6:22 PM | Permalink

    I had two geat privileges in my climate career. One was meeting with Lamb at East Anglia to discuss my doctoral research. The second was having him as a reviewer for one of my articles. The editor sent me the review with Lamb’s permission because he knew it would be important to anyone working in historical climate.
    I often wonder what Lamb would make of the direction, actions and behaviour of the Cliamtic Research Unit (CRU) that he founded.
    With regard to water temperatures and cod the collapse of the fisheries on the west coast of Europe is well documented. The social impact is less well known. Many people emigrated to Canada inclduing the Skinner family of Dropmore Manitoba. The family owned a fish wholesaling business in Glasgow that foundered when the fishing collapsed.
    Thirty years ago I was communicating with Roger Pocklington a research oceanographer based at the Dartmouth Insitute. He was the first to remind me our careers had spanned at least two climate cycles. Sadly Roger did not last to see another one, but the solar cycle forecasts suggest if I can just hang on a few more years I will.
    Roger was measuring water temperatures in a transect from Newfoundland to Bermuda. When he began he recorded his findings of declining water temperatures and was the toast of the town because global cooling was the prevailing wisdom. Then the wisdom changed to warming but Roger kept reporting the cooling was continuing. Now nobody wanted to hear what he had to say. We were trying to communicate the potential collapse of the cod fishery on the east coast to anyone who would listen. Nobody would and instead blamed the cod collapse solely on the fishermen. It is now 14 years since cod fishing was effectively banned on the east coast but the cod have not recovered because the water temperatures remain too cool. The cod moved to warmer internaitonal waters where Europeans continued to fish. They aslo moved inshore of Newfondland but were off limits to inshore fisheries.

  24. Posted May 23, 2006 at 10:28 PM | Permalink

    Re: Message #3, “that “Antarctic cooling” info is a little out of date.”

    The study in question found a significant temperature increase only in the Antarctic mid-troposphere in winter, with no corresponding temperature change at the surface, which is where the ice is. The full article is here:

    http://taylor-et-al.notlong.com

    Pete

  25. Steve Bloom
    Posted May 24, 2006 at 12:11 AM | Permalink

    From the paper:

    “We report an undocumented major warming of the Antarctic winter troposphere that is larger than any previously identified regional tropospheric warming on Earth.”

    “These records indicate that the western side of the Antarctic Peninsula has experienced the largest measured annual
    near-surface warming (0.55-C per decade at Faraday/Vernadsky station) on Earth over the past 50 years (1). However, there have been few statistically significant temperature changes at the surface across the rest of the continent (2, 3), and some studies have suggested a slight cooling in recent decades (4).”

    This makes it clear enough that an unambiguous statement that “the Antarctic is cooling” is not supportable.

  26. Posted May 24, 2006 at 2:30 AM | Permalink

    One again Real Climate behaves like Goebbels; twice I have attempted to point out that temperatures at Svalberg this year so far are no higher than in 1939, despite the triumphalism of of Mann and Jones (shurely not that Mann and that Jones?) over at Real (sic) climate. I am now convinced that ALL “comments” on that site are by said Mann and Jones, known here in Australia as “Dorothy Dixers” (i.e. soft questions to the Prime Minister in parliament by his own side).

  27. JerryB
    Posted May 24, 2006 at 3:59 AM | Permalink

    Re #25,

    The allegedly unambiguous statement is, a: not unambiguous, and b: not made by anyone else in this thread.

    Re #26,

    RC is like RC; the reference to Goebbels is inappropriate. Also, the suggested attribution of the comments there seems quite bizarre.

  28. tom brogle
    Posted May 24, 2006 at 4:26 AM | Permalink

    Steve Bloom
    You are at it again.You cannot take data from one station or a small area and apply it to an area 20 times the size.
    The bulk of the Antarctic is cooling slowly.
    The extent of the winter sea ice is also expanding having now reached as far as it was when explorers such as Shackleton,Bruce,Filchner,Scott etc… sailed south in the early 1900s
    I read somewhere that an undersea volcano is warming the Antarctic Peninsula

  29. Peter Hearnden
    Posted May 24, 2006 at 6:01 AM | Permalink

    Re #26 and 1939 – this is worth a look , 1939 looks colder than this year to me.

    Btw, comparing people to nazis puts you in trouble with prominant people here.

    Re #28, Tom

    You cannot take data from one station or a small area and apply it to an area 20 times the size.

    surely this, in principle, applies more to the interior than the penninsula, there being more weather stations on the coast than the interior? So, while there are straws in the wind, being clear what the temperature trend is (certainly in the interior) is difficult.

    Wrt to undersea volcanoes, for a start you’d see the ocean depths warming more than the surface – is that the case?

  30. kim
    Posted May 24, 2006 at 6:10 AM | Permalink

    Does ozone variability there affect temperature metering?
    ===================================

  31. kim
    Posted May 24, 2006 at 7:55 AM | Permalink

    I’m struggling with an old map which doesn’t specifically name the Antarctic Peninsula, but does it underlie the general area of the magnetic south pole?
    ================================================

  32. fFreddy
    Posted May 24, 2006 at 8:03 AM | Permalink

    Re #31, Kim
    It’s the thin bit pointing up towards South America.

  33. kim
    Posted May 24, 2006 at 8:06 AM | Permalink

    Ah, Palmer’s. Now, let’s get out the compass. A bear, not necessarily native, exits his den, and walks south for 10 miles, east for 10 miles, then north for 10 miles, and is back at his den. Need his fur be white? Can you count the number of places his den could be?
    ===========================================

  34. fFreddy
    Posted May 24, 2006 at 8:32 AM | Permalink

    No and no. You’ve lost me on the Palmers.

  35. Tom Brogle
    Posted May 24, 2006 at 8:57 AM | Permalink

    Peter

    Hot water rises.
    Expanding sea ice proves my point.

  36. kim
    Posted May 24, 2006 at 8:57 AM | Permalink

    You are correct on no and no. My map calls that Palmer Peninsula. Rand-McNally-Cosmopolitan World Atlas, MCMLIII. What about the effect of the magnetosphere focussing radiative changes to account for the area that seems to have warming?
    ==================================

  37. Peter Hearnden
    Posted May 24, 2006 at 9:12 AM | Permalink

    Re #35, “hot water rises” but the amounts are tiny in comparison with the volume of the oceans. Is there any evidence of the oceans near active vents being enough warmer to change climate? No. Indeed, is there ANY evidence whatsoever of warm plumes eminating from the depths (you claim this is happening “hot water rises”)? Even of places where the stratification of the oceans breaks down a little due to volcanic activity? I think not.

    Re #36, what indeed.

  38. Tom Brogle
    Posted May 24, 2006 at 9:46 AM | Permalink

    But the Antartic is still cooling not warming as your AGW conjecture indicates.

  39. Posted May 24, 2006 at 9:49 AM | Permalink

    >the Antartic is still cooling

    which must be why huge sections are breaking off and melting! ;-)

  40. tom brogle
    Posted May 24, 2006 at 10:03 AM | Permalink

    Yes, the more ice that builds up the more there is to break off.The only ice shelves to have collapsed are near the peninsula the others are in much the same position as they were 100 years ago.

  41. Mark
    Posted May 24, 2006 at 10:08 AM | Permalink

    re #39:

    which must be why huge sections are breaking off and melting!

    The only parts “breaking off and melting” are those touching the south pacific waters, which are warmer. AS tom noted, the more the merrier, too. This is elementary science. Also, anyone that has ever looked at the GISS temperature data knows that this part of the continent is warmer whereas the majority the continent is actually cooler, a trend that has been in place since at least the 1940s.

    Mark

  42. Posted May 24, 2006 at 10:21 AM | Permalink

    what a brilliant deduction; as if parts in the middle of the continent, which is usually at -37C would be breaking off and melting! ;-)

  43. Steve Sadlov
    Posted May 24, 2006 at 10:25 AM | Permalink

    The most eye opening thing about this is that it implies levels of innate variation that ought to give us serious pause. Even if we had never had the Industrial Revolution, the data which are increasingly aligning tell us that we can expect drastic sorts of things, such as massive variation in the Gulf Stream, massive variation in where the ice edge can reach during Spring, and massive variation in things in Europe and North America. Personally, I would not be surprised to see the current Warm Period (and that is what it is, no matter how one wants to debate what is driving it) reach its apex soon. Then, the true test will begin. By the way, I tried to interject some things about salinity and the history of ice free seas in the NE Atlantic and low longitude Arctic since the LIA, onto that Real Climate thread. Consider, for example, the possible impact on salinity of the massive diversions of north flowing rivers done by the USSR in Asia, in order to irrigate the desert and Eastern steppes and the more southerly parts of Siberia. What has that done to sea ice? Consider the many ships that have sailed well north ever since the mid 1800s. Etc.

  44. Willis Eschenbach
    Posted May 24, 2006 at 10:35 AM | Permalink

    Steve, please see my posts 103 and 105 on the thread New Online Resources. There is no “5 sigma event”, it’s a fantasy.

    w.

  45. Steve Sadlov
    Posted May 24, 2006 at 10:52 AM | Permalink

    RE: #45. Indeed, it is highly unlikely that this Spring in Svalbard is all that much different than most springs there since the late 1800s. I’ve personally done aerial recon of both sea ice and snow cover in the area of the North Atlantic and Low Longitude Arctic on multiple occasions over the past several years. The pattern we see today is not at all abnormal within the context of the current multidecadal cycle, in the big picture sense. I hope I am still alive when it becomes obvious that we’ve reached the peak of the current WP and the next LIA is beginning. Who knows, with some of the breakthroughs we are now seeing, I might live that long!

  46. agn
    Posted May 24, 2006 at 11:16 AM | Permalink

    Another funny thing about Svalbard is that, despite its (to be exact, we are talking about Spitzbergen here, the main island in the Svalbard group) mid-point being at about 78deg N, its East coast stays ice-free for a very large part of the year. This is because of the Gulf Stream, which ends up just about here.
    Of course, the “scientific consensus” says that, with global warming, the Gulf Stream is going to shut down (because the increasing meltwater reduces the salinity of the Arctic Ocean – it’s kind of a long story), which will cause a mini Ice Age in Europe. However, the “scientific consensus” does not appear to have thought about what such an event would do to Svalbard and the rest of the Arctic. But without the Gulf Stream, one would imagine it will get pretty cold up there. So then the salinity increases again and… negative feedback mechanism, anyone?

  47. Posted May 24, 2006 at 11:18 AM | Permalink

    a) Willis can’t even understand the concept of a temperature anomaly plot
    b) thermohaline shutdown is not the “scientific consensus” at all

    two l’il strawmen, sittin’ in a tree…

  48. agn
    Posted May 24, 2006 at 11:41 AM | Permalink

    Oh good. Is there somewhere, on a website perhaps, that actually specifies what the “scientific consensus” is? Or even perhaps a definition of what constitutes “scientific consensus”? It would be very helpful as it is easy for us men of straw to get confused about this…

  49. Posted May 24, 2006 at 12:04 PM | Permalink

    ummm, well you’re the one claiming a thermohaline shutdown is a “scientific consensus” — so don’t you think you should provide the basis for this? It’s a possible scenario that’s bandied about, that’s about all I’ve seen. Unless you are talking about the movie “The Day After Tomorrow!” ;-)

  50. Robert
    Posted May 24, 2006 at 12:11 PM | Permalink

    re:35

    Hot Water rises is a bit of a myth. Rising and falling is due to density, basically less dense substances “float” or rise in liquids.

    Water has highest density at 4 degrees C, above and below this temperature point the density rises, hence Ice floats, and hot water rises. Water with a temperature of 1 degree C, very close to freezing, will “rise” if surrounded by 4 deg C water.

    It’s a bit of a nit pick, but it’s important to understand how peculiar water really is.

    cheers,
    Robert.

  51. Posted May 24, 2006 at 12:17 PM | Permalink

    The ‘Gulf Stream’ won’t be ‘shutting down’ unless the Earth stops rotating and the winds from the South West stop blowing.

  52. jae
    Posted May 24, 2006 at 12:23 PM | Permalink

    re: 50. You mean above and below 4 C the density decreases. Yes, water is really a “strange” substance in many ways. Thank God. If it behaved like nearly all other substances, ice would sink, and very little would ever melt! It’s interesting to picture how a body of water freezes–it appears that there can be no freezing, until all the water in the column is cooled to 4 C; only then can some of it get cooler and freeze.

  53. PHEaston
    Posted May 24, 2006 at 12:30 PM | Permalink

    Going back to Mr Hearnden’s comment no. 19, in which he quotes Lamb’s concerns (stated in 1988) that CO2 emissions may cause global warming. Most participants in this site would probably agree these are legitimate concerns (particulaly at the time they were stated).
    My understanding of this website is that it does not strive to disprove global warming or that CO2 emissions have an influence. It accepts that the potential influence of CO2 emissions on climate is a perfectly legitimate concern and one that justifies scientific study.
    The principal aim of this site (led by Steve McIntyre)is to promote good and open SCIENCE and root out crap or politicised ‘science’.
    The point is that whatever the concerns regarding global warming, there is as yet no convincing evidence that (i) current temperatures or trends are unprecendented (ii) that manmande CO2 emissions are having a significant (ie. measurable) impact on global temperatures (iii) that climate models provide a reliable prediction of future temperatures, and most importantly (iv) that there is no further room for debate.

  54. agn
    Posted May 24, 2006 at 12:45 PM | Permalink

    Re #53: exactly. So the point I was trying to make is that the words “scientific consensus” probably always need to go in quotation marks!

  55. Steve Sadlov
    Posted May 24, 2006 at 1:58 PM | Permalink

    Always a fan of the big picture, I simply cannot resist the obligatory almost real time cryo shot:

    It helps folks to visualize the places and things we are talking about here.

  56. Ken Robinson
    Posted May 24, 2006 at 2:14 PM | Permalink

    Re: 19 and 20

    Thanks, Peter and Spence respectively, for the references. I’ll look them up.

    Best regards;

  57. Steve Bloom
    Posted May 24, 2006 at 2:26 PM | Permalink

    Re #45: “The pattern we see today is not at all abnormal within the context of the current multidecadal cycle, in the big picture sense.” This is a funny way of describing a record low, don’t you think? Er, also, where’s the peer-reviewed work on the effects of that multi-decadal cycle on sea ice?

    Re #48: agn, for statements about the scientific consensus on AGW you might try the web sites of, oh, I don’t know, the major scientific organizations: AAAS, NAS, AGU, IPCC, the list goes on.

  58. jae
    Posted May 24, 2006 at 2:30 PM | Permalink

    Bloom: you would really be lost without that appeal to authority, wouldn’t you? Try some reasoning, for a change, eh? The “scientific consensus” often changes, you know….

  59. Willis Eschenbach
    Posted May 24, 2006 at 2:58 PM | Permalink

    Re 47, Carl, you say:

    a) Willis can’t even understand the concept of a temperature anomaly plot

    I understand an anomaly plot quite well … but you don’t seem to understand what they’ve done.

    They’ve taken the mean and standard deviation of a 30 year subset of a 94 year record. They then apply those numbers to determine if numbers outside of the subset are anomalous. This is ridiculous on the face of it. Bad mathematicians … no cookies.

    Think about it, Carl. Would it be valid to do the same thing with say a 10 year subset? Of course not. The sigma statistics are only valid if you use the mean and standard deviation of the entire dataset.

    w.

    PS – If you want to be taken seriously here, please dial down on the abuse …

  60. kim
    Posted May 24, 2006 at 3:19 PM | Permalink

    #55, that little green cryo three quarter circle looks about where the magnetic south pole might be. Or am I disoriented?
    =============================================

  61. PHEaston
    Posted May 24, 2006 at 4:18 PM | Permalink

    RE agn 48 & 54. The ‘scientific consensus’ is deemed to be represented by the IPPC TAR (third assessment report, 2001) which includes the famous statement that “there is new and stronger evidence that most of the observed warming observed over the last 50 years is attributable to human activities”. However, a careful read of the IPPC documments demonstrates that this is an ‘opionion’ (or perhaps gut feeling) rather than a conclusion based on reliable scientfic evidence. The IPPC unfortunately appears to be politicised. It includes a broad collection of scientists, but inevitably, those who have a common belief: that manmade CO2 emissions are causing climate change. There is little room for disent, to the degree that those who disagaree are simply excluded or end up resigning. A SCIENIST does not base his/her belief on ‘consensus’ or democracy. Beliefs are dervived from understanding the evidence and drawing an independent conclusion. We regularly hear ‘manmade global warming’ beleivers claiming that we should believe because it is the consensus. The ‘consensus’ is irrelevant if in your heart you are not satisfied by the evidence presented. What we demand is open and honest scientific debate,

  62. jae
    Posted May 24, 2006 at 4:52 PM | Permalink

    Too many of the “warmers” are relying on this non-scientific crap of “scientific consensus.” I maintain that the “consensus” definitely no longer sides with certain portions of the AGW theory. For example, I doubt that there is now a consensus among climate scientists that the infamous hockey stick studies are defensible, thanks to the work of Steve M and others. I also see more and more questioning of the surface temperature “reconstruction,” because Jones and the boys are playing the same games that the hockey stick team are playing. There is something really wierd and questionable about climate science, IMO (refusal to provide data and methods, lack of cooperation, etc). Also, there are many famous climate modelers who maintain (and have all along) that the models are not (yet) cabable of proving anything. I see the “consensus” changing to either “who knows” or “just another normal cycle.” Check out the HUNDREDS of references on CO2Science.org. (most in peer reviewed journals, BTW, Dano). There is getting to be a smaller and smaller basis for any so-called “concensus.” We will see how the IPCC handles all this doubt.

  63. joel Hammer
    Posted May 24, 2006 at 5:44 PM | Permalink

    I always thought you never were able to prove a scientific hypothesis. All you do, with all your hard work, is to say your data fails do disprove it.

    This being so, how can anyone claim to be sure that CO2 is causing global warming?

    What experiment could you run to disprove this?

    This is in the end an untestable hypothesis. Thus, the heated arguments.

  64. jae
    Posted May 24, 2006 at 5:51 PM | Permalink

    Joel: spot on, man!

  65. John Cross
    Posted May 24, 2006 at 7:13 PM | Permalink

    Re 50 and 52: Just to correct some of the science that is floating around here – seawater density continues to increase as the water cools (until it freezes at about -1.9C). It is only freshwater that has its maximum density at slightly less than 4C.

    Regards,
    John

  66. Steve Sadlov
    Posted May 24, 2006 at 7:19 PM | Permalink

    #65. That is a really great point. That is why I would like to see a much more in depth look at the influence which salinity (or more correctly, total molarity of all ions in solution) of not only the Arctic, but also naturally the Atlantic. It somewhat amazes me that it has not been more carefully considered, especially vis a vis sea ice dynamics. Without knowing the molarity (let alone, of course, winds, SSTs, air temperatures, solar flux, etc) how can one truly understand, let alone feign to predict, what sea ice is doing and why it is doing it? Sea ice may be the ultimate in tricky proxies.

  67. Steve Sadlov
    Posted May 24, 2006 at 7:22 PM | Permalink

    RE: #57. Please review the image I posted of sea ice coverage. What would be your own assessment of it? I am very curious about your analysis.

  68. Al Gore
    Posted May 24, 2006 at 7:29 PM | Permalink

    Whoops! I got the wrong website. Can anyone tell me where the real climate website is? The longer I stay at this site, the more disoriented I feel. Anybody got smelling salts?

  69. Posted May 24, 2006 at 7:41 PM | Permalink

    Re #62 jae

    The IPCC is all about politics, not science. They are not going to change their position base on science facts, we will only get the political correct solution.

  70. Tom Brogle
    Posted May 24, 2006 at 11:29 PM | Permalink

    Steve Bloom
    Carl Christiansen
    Peter Hearnden
    Hydrothermal vents are in deep ocean and do not heat the surface.However if there is a volcano on the Continental Shelf that is adjacent to the Antarctic Penninsula it would have been the cause of the collapse of the Larsen ice shelf.
    I repeat the Antarctic is still cooling the winter sea ice around it is still expanding.
    How does that stack up against your AGW conjecture?

  71. Steve McIntyre
    Posted May 24, 2006 at 11:40 PM | Permalink

    #59. Willis, you might be interested in looking at this old post on standard deviations from autocorrelated series. http://www.climateaudit.org/?p=342

  72. Steve Bloom
    Posted May 25, 2006 at 12:23 AM | Permalink

    Re #67: I can’t find anything on the UIUC site with daily anomalies, so I can’t really say for sure about “today” as such. If you look on the front page of the site, though, you’ll find an NH anomaly graph (I think by month, but unfortunately it’s not gridded such that one can tell for sure which month is which) that makes the current level (probably for April) look like it’s either at or very close to a record low. Anyway, March most definitely was a record low.

    Looking over at the NSIDC site (way more information than UIUC), here is a monthly anomaly graph that shows, yes indeed, April too was a record low. Here is some daily data that’s apparently pretty close to real time, but there’s no graph that I can find. I’m sure it’s fascinating to spend time with this data, but I’m content to wait for the monthly graph. Note that the documentation says the fresh daily data hasn’t had time to be run through quality control and will change slightly after it has been.

    The daily data aside, I’d say two record low months in a row is notable enough.

    BTW, don’t make the Mistake of Willis (MoW) and compare sea ice extent data between these two sites. They use the same data but apply different metrics, which among other things means that occasionally one may show a record month and the other not. (To be fair to Willis, Roger Pielke Sr., a climate scientist of thirty or so years experience, made the same mistake. So it could easily have been the MoRPSr, but Willis’ snark level was so much greater that I had to give him the prize.)

  73. Steve Bloom
    Posted May 25, 2006 at 12:48 AM | Permalink

    Re #70: While I am very pleased to be associated with Carl and Peter, had I said anything at all about this issue? I don’t think so. But while we’re on the subject, why don’t you have a look around the web and see if there is or isn’t an undersea volcano at the location you mentioned? Let us know what you find. Here‘s a start. Remember, Google Scholar is your friend.

  74. Tom Brogle
    Posted May 25, 2006 at 1:14 AM | Permalink

    Steve Bloom 73
    I refer you to your post 25
    “This makes it clear enough that an unambiguous statement that “the Antarctic is cooling” is not supportable.”
    “had I said anything at all about this issue?”
    Yes you have.
    If you don’t believe the Antarctic is cooling look at Real Climate there is a posting

    http://www.realclimate.org/index.php/archives/2004/12/antarctic-cooling-global-warming/

    saying just that and I’m sure you should accept that as gospel.

  75. UC
    Posted May 25, 2006 at 1:16 AM | Permalink

    When we look at the famous Chebyshev inequality, P(|x-m| >= ks)

  76. Willis Eschenbach
    Posted May 25, 2006 at 1:24 AM | Permalink

    Looking further at the effect of autocorrelation on time series has been quite fascinating. One way to calculate the effect of autocorrelation is to reduce the effective number of data points N. According to Wigley, a reasonable estimate of the reduction of N is that

    N(effective) = N (1-R1) / (1+R1)

    where R1 is the lag-1 autocorrelation. The Wigley paper regarding trends is available at http://www.climatescience.gov/Library/sap/sap1-1/finalreport/sap1-1-final-appA.pdf

    One of the effects of this reduction in N is to increase the standard error of trends. To quote from Wigley (emphasis mine):

    If the effective sample size is noticeably smaller than n, then, from equations (7) and (8) it can be seen that the standard error of the trend estimate may be much larger than one would otherwise expect. Since the width of any confidence interval depends directly on this standard error (larger SE leading to wider confidence intervals), then the effect of autocorrelation is to produce wider confidence intervals and greater uncertainty in the trend estimate. A corollary of this is that results that may show a significant trend if autocorrelation is ignored are frequently found to be non-significant when autocorrelation is accounted for.

    The increase in the standard error of the trend is given by

    Trend se (effective) = trend se * sqrt((1+R1) / (1-R1))

    This goes up fairly fast. Many temperature series have R1 values of .7-.8. Here are the numbers by which the trend error estimate needs to be multiplied to allow for autocorrelation:

    R1_____Multiplier
    0.5_____1.7
    0.6_____2
    0.7_____2.4
    0.8_____3
    0.9_____4.4

    Including autocorrelation, the general warming of the earth still has a significant trend. I used the TAVEGL2v Hadcrut dataset for global temperatures. To allow comparison with Svalbard, I used the period 1912-2005, which is the overlap between the two datasets. The global temperature has a lag-1 autocorrelation R1 of 0.89.

    The global trend (including the adjustment of the standard error of the trend for autocorrelation) is highly significant, at 0.07°/decade ± 0.014° (2 std. dev.).

    Svalbard, on the other hand, has a higher trend over the period, but the trend is not significant. Despite having a slightly lower autocorrelation (R1 = 0.79, the Svalbard trend for the same period is 0.25°/decade ± 0.54° (2 std. dev.).

    As Steve M. has pointed out, the lack of adjustment for autocorrelation has been a significant oversight in the general assessment of temperature trends. Un this case, it looks like Svalbard has a significant trend. Without the proper adjustment for autocorrelation, the Svalbard 1912-2005 trend is 0.25°/decade ± 0.18° (2 std. dev.). This would be considered as highly significant, but in fact, it is not.

    Once it is properly adjusted, as Wigley comments, “results that may show a significant trend if autocorrelation is ignored are frequently found to be non-significant when autocorrelation is accounted for.” Svalbard’s temperature trend is not significantly different from zero.

    Which is about my chance of getting this comment accepted at RealClimate, not significantly different from zero … they still haven’t accepted my very bland submission saying that their so called “Svalbard” dataset is actually the splicing of two datasets, Svalbard Luft and Istfjord Radio, from sites 60 km apart. See http://www.unaami.noaa.gov/analyses/sat/#table for details. The notes imply that there are three years of overlap, but in fact there is a year of no data from Istfjord Radio during the overlap, followed by two years that only contain three months! of scattered, non-contiguous data.

    w.

  77. Peter Hearnden
    Posted May 25, 2006 at 1:37 AM | Permalink

    Tom Brogle #70

    “However if there is a volcano on the Continental Shelf that is adjacent to the Antarctic Penninsula it would have been the cause of the collapse of the Larsen ice shelf.”

    But….there isn’t one!

    The Larsen B shelf has been there for at least 10,000 years.

  78. Tom Brogle
    Posted May 25, 2006 at 1:57 AM | Permalink

    Peter Hearnden
    There is a volcano
    “The volcano is located on the continental shelf of the northeastern Antarctic Peninsula, where recent changes in surface temperature and ice shelf stability have been observed.” http://adsabs.harvard.edu/abs/2005AGUFMPP41A0636Q

  79. fFreddy
    Posted May 25, 2006 at 2:02 AM | Permalink

    Re #77, Peter Hearnden

    The Larsen B shelf has been there for at least 10,000 years.

    Peter, what is the evidence for this ?

  80. Peter Hearnden
    Posted May 25, 2006 at 2:15 AM | Permalink

    Re #78, well, well, good find that man! I’d like to know precisely where it is, but, nevertheless, good find. Now, show me how it melted the ice shelf (remember, Mt Erebrus isnt ice and snow free…).

    fF, yup, in a paper I ‘ll try to dig out.

  81. agn
    Posted May 25, 2006 at 2:26 AM | Permalink

    Sorry, don’t seem to be able to post anymore. Pity, I like the site!

  82. Tom Brogle
    Posted May 25, 2006 at 2:29 AM | Permalink

    Peter
    This one is NEW and underwater.
    Explain why it wouldn’t melt an ice shelf
    The authors of the article imply that that
    this volcano is the cause of the temperature
    rise of the Antarctic penninsula.
    Can you find anymore ways of wriggling out of this one.

  83. agn
    Posted May 25, 2006 at 2:30 AM | Permalink

    I have found what the "scientific consensus" at IPCC says about thermohaline shutdown but the spam filter won’t let me post it. John A, can we discuss ways of improving this system, please?

    Steve: I’ll try to retrieve this. We are inundated with spam. I ask people to email me if they have a problem as, by and large, the Spam filter works very well and we could not cope without it.

  84. Peter Hearnden
    Posted May 25, 2006 at 2:39 AM | Permalink

    Re #81, yes indeed. Show me the evidence the sea under the shelf has been warmed by the volcano enough to melt an ice shelf that’s been there for 10, 000 years. You will have to show the volcano has increased its activity as well I suspect.

    Oh, and if you look where ‘Antarctic sound’ is you’ll find you have another rather major problem…it’s nowhere near Larson B…

  85. Tom Brogle
    Posted May 25, 2006 at 2:56 AM | Permalink

    I put your suggestion to my wife,
    Her comment was that if a new volcano started in relatively shallow seas it would
    warm them and cause the ice shelf to break.
    She thinks you are being unreasonable.
    Stop wriggling you know you’ve lost the argument
    I repeat, the Antarctic is still cooling the winter sea ice around it is still expanding and an undersea volcano caused the collapse of the Larsen ice shelf.

  86. Steve Bloom
    Posted May 25, 2006 at 3:00 AM | Permalink

    Re #81: How new? (“New” has a habit of being relative with geologic features.) Where (and how deep) is it relative to the ice shelves that collapsed (not melted, BTW)? When has it erupted? How much did it heat the water when it did? Does the surface water (where the ice shelves are located) show any sign of warming from this volcano? Are there other volcanos or geothermal activity in the area?

    As for the volcano being “the cause of the temperature rise of the Antarctic penninsula,” the authors implied no such thing. Note that their purpose was “to examine the benthic assemblages and biodiversity present on a young volcano.” Reduced ice and increased surface temps could have affected the life they were studying, and so were important to mention.

  87. Peter Hearnden
    Posted May 25, 2006 at 3:03 AM | Permalink

    Re #84.

    Tom (and your Mrs), just look where Antarctic Sound IS will you. You really think one volcano, miles away from an ice shelf can melt that said ice shelf? It just not a credible idea.

  88. Steve Bloom
    Posted May 25, 2006 at 3:14 AM | Permalink

    Re #79: Larsen A is about that old, as is (was) Larsen-B. The latter article seems to be a thorough discussion of the history of all of the ice sheets in that area, based on examining the contents of the sea floor beneath them. I saw no mention of volcanos.

  89. Tom Brogle
    Posted May 25, 2006 at 3:15 AM | Permalink

    Peter and Steve
    You are simply Nitpicking.The Antlantic sound extends to near to where the Larsen ice shelf collapsed.
    If the volcano warmed the sea and most people would expect it to.Sea currents would do the rest.
    Why don’t you give up ?

  90. Steve Bloom
    Posted May 25, 2006 at 3:20 AM | Permalink

    Re #88: Tom, I’m in a semi-foul mood from earlier in the day, but despite that I’m going to refrain from making fun of you.

  91. Tom Brogle
    Posted May 25, 2006 at 3:22 AM | Permalink

    Seve
    Are you giving up then ?

  92. Willis Eschenbach
    Posted May 25, 2006 at 12:48 PM | Permalink

    Whoa, climate change, ice in the place of eternal perdition … I resubmitted my first post to RC, and IT WAS ACCEPTED … yikes.

    Buoyed by my success, and in a state of unjustifiable optimism, I have submitted the following:

    Thank you, Mike. Much appreciated. However, the “source of the data” (CRU) doesn’t make individual station records available, so I had to make do with what I had. Upon further research, I found that the complete spliced dataset is available at http://www.unaami.noaa.gov/analyses/sat/#table

    In any case, the record is a spliced record, with a one year period between the two records without any data. I’m curious about the justification for the splice, since there is no overlap, and I’m curious that a spliced record with no overlap between the two station records would be used for this type of statistical procedure.

    There is a larger problem with the claim, however, This is that you have taken the mean and standard deviation of a subset of the data (1961-1990), and are using these figures to derive sigma figures for a time period which is out of the subset (Dec ’05 to April ’06).

    Mathematecially, this is simply not correct. For example, could we take the mean and SD of a ten year subset, say 1983-1992, and compare that to April ’06? If we do, we will get a very different answer. By that measure, the odds of April ’06 are only about one in a hundred.

    The only way to do a sigma calculation is to include all of the data when calculating the mean and standard deviation. Otherwise, the answer depends on which subset of the data is chosen, which is clearly wrong.

    This is a very serious error, which materially affects the stated results. I expect that, in conformance with your stated policies, my message (being scientific in nature, not a troll, and not abusive in any way) will not be censored.

    Many thanks,

    w.

  93. Steve Sadlov
    Posted May 25, 2006 at 1:54 PM | Permalink

    RE: #92. Willis, you indeed found a very serious error there. Well, I am being nice in calling it an error …. I could have just as well called it intentional distortion. RC plays mostly to the non scientists, non engineers, non mathematicians and non statisticians who are most suceptible to regarding “Environmentalism” as a sort of religious orientation.

  94. UC
    Posted May 25, 2006 at 1:55 PM | Permalink

    Astronomically improbable 7-sigma deviation in Reykjavik, June 1982..

    (www.unaami.noaa.gov/analyses/sat/sat04030.d)

    This didn’t get through for some reason:

    When we look at the famous Chebyshev inequality, P(|x-m| >= ks)

  95. Willis Eschenbach
    Posted May 25, 2006 at 3:21 PM | Permalink

    UC, thanks for the interesting pointer.

    I would bet very big money that it’s a typo, someone entered .9 for 9. Graph the June data and you’ll see what I mean. Reagan and Gorbachev met in Reykjavik during that month, and England played soccer with Iceland then, and the reports don’t mention the temperature.

    Which is why, when I see “astronomically improbable” or “6 sigma event”, my urban legend detector hits the peg …

    w.

  96. Posted May 25, 2006 at 5:49 PM | Permalink

    A little more on topic, there seems to be a correlation between sea ice extent around Svalbard, d18O in one of the (low lying) glaciers there and the temperture record of Jan Mayen, some 1100 km SW of Svalbard. According to E. Isaksson ea.:

    The Austfonna record correlates well with the temperature record from the more distant and southwesterly located Jan Mayen. A comparison of the ice-core and sea-ice records from this period suggests that sea-ice extent and Austfonna d18O are related over the past 400 years. This may reflect the position of the storm tracks and their direct influence on the relatively low-altitude Austfonna.

    The temperature record of Jan Mayen shows temperatures in the period 1930-1960 near as high as in recent years…

  97. Steve Bloom
    Posted May 25, 2006 at 6:16 PM | Permalink

    Speaking of storm tracks, have a look at this.

  98. Steve Bloom
    Posted May 25, 2006 at 6:34 PM | Permalink

    I forgot to add that like most aspects of the climate the jet stream shift must be a consequence of that pesky Antarctic undersea volcano. :)

  99. Willis Eschenbach
    Posted May 25, 2006 at 6:47 PM | Permalink

    Re 98, although said in fun, it is an interesting subject. I was pointed today to an new study in the May 25 issue of Nature:

    Embley, Robert W., William W. Chadwick, Jr, Edward T. Baker, David A. Butterfield, Joseph
    A. Resing, Cornel E.J. de Ronde, Verena Tunnicliffe, John E. Lupton, S. Kim Juniper,
    Kenneth H. Rubin, Robert J. Stern, Geoffrey T. Lebon, Ko-ichi Nakamura, Susan G. Merle,
    James R. Hein, Douglas A. Wiens and Yoshihiko Tamura, 2006. Long-term eruptive activity
    at a submarine arc volcano
    . Nature Vol. 441, No 7092, pp. 494-497, May 25, 2006 

    The abstract starts:

    “Three-quarters of the Earth’s volcanic activity is submarine, located mostly along the
    mid-ocean ridges, with the remainder along intraoceanic arcs and hotspots at depths
    varying from greater than 4,000 m to near the sea surface. …”

    I was suprised to find out that three quarters of the earths volcanic activity happens underwater. Made me curious about the overall effects, both thermal, and in terms of CO2. I know that some of the volcanoes emit liquid CO2 at depth … much to learn about this magnificent planet.

    w.

  100. Willis Eschenbach
    Posted May 25, 2006 at 7:09 PM | Permalink

    Re 97, Steve, thanks for the interesting article. However, this article illustrates the problem with news articles. It says that the study appears in todays issue of ScienceExpress. Only trouble is, it doesn’t …

    Thanks,

    w.

  101. Steve Bloom
    Posted May 25, 2006 at 9:47 PM | Permalink

    Re #100: Not Express, it turns out. Correct link here.

  102. Willis Eschenbach
    Posted May 25, 2006 at 11:09 PM | Permalink

    Thanks, Steve. I’ll take a look when I have a chance, may be a couple days.

    All the best,

    w.

  103. Tom Brogle
    Posted May 25, 2006 at 11:48 PM | Permalink

    Steve Bloom
    That ” pesky Antarctic undersea volcano” caused the demise of the Laraen Ice Shelf. It had nothing to do with the jet stream shift.

  104. Steve Bloom
    Posted May 26, 2006 at 12:16 AM | Permalink

    Re #103: It didn’t? Prove it.

  105. Tom Brogle
    Posted May 26, 2006 at 1:04 AM | Permalink

    Steve Bloom
    Itis said that data from ice cores e.g.the the Vostok ice core proves that CO2 causes GW.However the data show that CO2 rose after the temperature rose and fell after the the temperature fell. So the above “proof”
    is putting the effect before the cause. If after an ice age the earth warms over thousands of years there is enough time for the CO2 in the atmosphere to come to equilibrium with the CO2 in the ocean.
    As the ocean warms it loses CO2 to the atmosphere raising the atmospheric CO2 from 180 ppm to about 280 ppm
    In other words the change in CO2 is caused by the change in temperature.
    Ive done the mathematics involved and this theory works.Try it yourself
    And by the way the Antarctic is still getting cooler , the winter sea ice around it is still expanding and there is a recently active volcano on
    “the continental shelf of the northeastern Antarctic Peninsula, where recent changes in surface temperature and ice shelf stability have been observed.”

  106. Peter Hearnden
    Posted May 26, 2006 at 1:36 AM | Permalink

    Re #99, Willis for a start three quarters of the planet is covered by water. Then, spreading mid ocean ridges are, obviously, under water. I don’t find the figures at all surprising.

  107. Posted May 26, 2006 at 1:48 AM | Permalink

    Re #97

    Steve, there are direct correlations between the jet stream position and solar cycles (and explosive volcanic eruptions like the Pinatubo). It seems that the jet stream position is a stratospheric process that influences weather/climate in the troposphere, not the other way out (that means that GHGs and/or human induces aerosols have no/little influence on the jet stream position).

    For the mechanism in general: http://www.agu.org/pubs/crossref/2005/2005GL024393.shtml
    and http://www.nwra.com/resumes/dunkerton/pubs/jastp.xx.xx.xx.pdf
    For the USA: http://www.gsfc.nasa.gov/topstory/20010712cloudcover.html
    Specific for Boston: http://www.erh.noaa.gov/box/effects.htm

    Thus at three levels (world, USA, region) there is an effect measurable of solar activity via the stratosphere and jet stream position on weather/climate…

  108. Posted May 26, 2006 at 1:59 AM | Permalink

    In addition to what Tom said in #105, there is a period at the onset of the last glaciation (the end of the Eemian interglacial) where there was such a delay in CO2 change, that temperatures dropped almost to a minimum, before CO2 levels started to decline. So there was no overlap at all, and the 40 ppmv drop of CO2 didn’t cause an appreciable change in temperature…
    See the graphs at my web site here

  109. UC
    Posted May 26, 2006 at 2:00 AM | Permalink

    RE: #95

    You are right, it is very probable that it is a typo. But, we have a problem: if the measurements that do not agree with the ‘normal distribution assumption’ are always flagged, the resulting series is ‘forced’ to Gaussian. Thus, we can’t use that data to decide whether the original process is normally distributed or not. And if the process is not normally distributed, the 5-sigmas are not necessarily astronomically improbable. The probability can be anything up to 1/25.

    Another problem is the accuracy of the sd estimate, but that was already addressed.

  110. Tom Brogle
    Posted May 26, 2006 at 6:51 AM | Permalink

    Re 84
    Another one for Peter Hearnden:
    “the volcano enough to melt an ice shelf that’s been there for 10, 000 years.”
    I quote
    “First survey of Antarctic sub-ice shelf sediments reveals mid-Holocene ice shelf retreat
    C.J. Pudsey and J. Evans. British Antarctic Survey, High Cross, Madingley Road, Cambridge CB3 0ET, UK. Pages 787-790.
    Abstract
    The northern Antarctic Peninsula has warmed by 2.5 °C in the past 50 yr. This is significantly greater than the global mean temperature rise over the same period, and has been interpreted by some as an indication of anthropogenic climatic warming, which is expected to be amplified in the polar regions. As a result of the warming, several ice shelves have retreated or disintegrated. By retrieving marine sediments from locations that were previously covered by ice shelves, it can be determined whether the current retreat is unique within the Holocene, which spans the past 10 k.y., or whether it has earlier analogues. The authors collected cores from previously ice-covered areas of the northern Larsen ice shelf. Analysis of rock fragments and organic material found in the cores revealed that in the past, icebergs transported different rock types freely through the area. At such times the ice shelf could not have been present. Carbon-14 dating indicates that the ice shelf was absent from ca. 5000-2000 yr ago, but has probably been a stable feature for the past 1800 yr. Other evidence suggests the Antarctic Peninsula was relatively warm at the time the ice shelf was absent. Thus, the present loss of ice shelves cannot be taken as a “fingerprint” of anthropogenic climate change, since such events occur naturally.”
    You were wrong with the age of the Larsen ice shelf !

  111. Louis Hissink
    Posted May 26, 2006 at 7:15 AM | Permalink

    # 106

  112. Louis Hissink
    Posted May 26, 2006 at 7:16 AM | Permalink

    #111

    Grrrrr

    and #106

    Peter,

    I suspect you missed Willis’ subtlies.

  113. Peter Hearnden
    Posted May 26, 2006 at 7:18 AM | Permalink

    Well, if you can quote your paper as the facts, why can’t I? ;)

    That said, I will try to read both papers. I don’t dismiss either so I might indeed be wrong.

  114. Peter Hearnden
    Posted May 26, 2006 at 7:34 AM | Permalink

    Re #112, well, heck, it’s not impossible – few people are allways right, are you? But, let Willis speak for himself eh?

  115. JerryB
    Posted May 26, 2006 at 8:32 AM | Permalink

    Re #92 Willis,

    The “complete spliced dataset” may have more splices than you might expect, plus some “corrections” not mentioned in the amalgam, as well as WW II data from who knows where.

    See: http://www.john-daly.com/svalbard.mix for more numbers. and: http://press.telenor.com/PR/200109/833582_5.html for some Isfjord Radio history, from which the following excerpts are lifted.

    On May 3, 1911, the Norwegian parliament agreed that a radio telegraph
    station was to be built on Svalbard. The station was named Spitsbergen
    Radio and originally built at Green Harbour. It was later moved to
    Finneset just south of Barentsburg.

    In 1930, the station that had been named Svalbard Radio in 1925 was moved
    to Longyearbyen.

    In 1933, Isfjord Radio was established at Kapp Linne. The station was
    established to act as an intermediary for traffic between Svalbard Radio
    and ships in the waters around Svalbard.

    During the second world war, Isfjord Radio was decommissioned and
    destroyed by German occupying forces, but the station was rebuilt and set
    back into operation in 1946.

    Most of Isfjord Radio’s operations were moved to Longyearbyen when the
    airport was opened in 1975.

  116. Steve Sadlov
    Posted May 26, 2006 at 11:06 AM | Permalink

    RE: #97. There are several problems with that Murky News article and the study it describes. To depict the “Jet Stream” as a single, relatively straight line “river” is a gross oversimplication. The jet stream is only a concept to convey maxima in high altitude winds near the troposphere – stratosphere boundary. Sometimes, these maxima are well organized and coalesced into a single “stream” at other times the split, for eddys or even wash out into diffuse, very wide fields of turbulent flow. Furthermore, these maxima are not always as west to east as depicted in text books. For example, today, the “jet stream” is flowing from Siberia, up over part of the Arctic Ocean, then straight down from the Yukon across central British Columbia and down along the US West Coast, turning the corner near the mid point of the US – Mexican border. In other words, the stream can meander far and wide. Interestingly, a similar meander far the the south is also currently happening in China, bringing record cold temperatures to Guangdong Province. Desertification has myriad causitive factors, not least of which is land use. In summary, I would strongly challenge the claims of that paper. There is likely a specific political / ideological / spiritual agenda driving the theme of that paper.

  117. Willis Eschenbach
    Posted May 26, 2006 at 4:05 PM | Permalink

    Re 106, Peter, I’d never given it much thought. All you say is true, but when I think of volcanoes I think of … well, volcanoes. So I was surprised. I also never thought of the rift zones as volcanoes … although they exhibit “volcanic activity”.

    w.

  118. Willis Eschenbach
    Posted May 26, 2006 at 4:13 PM | Permalink

    Re 115, JerryB, thanks for the info.

    I actually was able to find some recent meteorological data from Isfjord Radio at at http://helios.hampshire.edu/~srNS/Svalbard/2005%20Data/Weather%20data/Isfjord%20Radio%20weather/Isfjord%20Radio%20Station%202004-2005%20Temp_RH%20044.xls. Surprise, surprise, it differs from Svalbard. The data covers from mid 1996 to 2004. During that time, Svalbard was consistently colder in winter and warmer in summer than Isfjord. The raw statistics are:

     
    ________________Svalbard_________Isfjord Radio  
     Average_________-5.1____________ -3.5  
     Std. Deviation______8.0_____________6.4

    This clearly shows why you shouldn’t use merged station data for mathematical analysis.

    w.

  119. Willis Eschenbach
    Posted May 26, 2006 at 4:17 PM | Permalink

    I have attempted to post again at realclimate, as follows:

    Mike [Mann], thank you for your answer to Hans Erren. You say:

    For any further details, you and any other interested readers should refer to the linked CRU website for information and references leading to an extensive body of literature that describes how CRU forms representative composite 5 degree latitude x longitude gridbox estimates (which is what is referred to here) that account for time-dependent sampling variations and potential inhomogeneities associated with the individual recording stations that fall within the same grid cell.

    Now I’m confused. You say that you are referring to the representative composite 5° x 5° gridbox temperatures … but in your original post, you say “Here are the recent Svalbard monthly surface temperature measurements …” and you compare them to the December ’05 to April ’06 Svalbard surface temperature measurements. You also refer to a check made with the Norwegian Met Service to make sure that your figures were correct, which clearly means that they are not gridbox temperatures.

    Could you clarify which dataset you used for your calculations, the merged Svalbard/Isfjord Radio surface temperature dataset, or the 5° x 5° gridbox temperatures?

    I ask in part because, despite an extensive search the CRU site, I cannot find any dataset for Svalbard or Isfjord Radio. Is the CRU the source of your dataset, and if so, where?

    Many thanks for your clarification,

    w.

    We’ll see … watch the pea under the shell, Michael Mann seems to be up to his old “hide the data” game again …

    w.

  120. Willis Eschenbach
    Posted May 26, 2006 at 4:24 PM | Permalink

    Block quoting got messed up on this one, so I’m re-posting …

    I have attempted to post again at realclimate, as follows:

    Mike [Mann], thank you for your answer to Hans Erren. You say:

    For any further details, you and any other interested readers should refer to the linked CRU website for information and references leading to an extensive body of literature that describes how CRU forms representative composite 5 degree latitude x longitude gridbox estimates (which is what is referred to here) that account for time-dependent sampling variations and potential inhomogeneities associated with the individual recording stations that fall within the same grid cell.

    Now I’m confused. You say that you are referring to the representative composite 5° x 5° gridbox temperatures … but in your original post, you say “Here are the recent Svalbard monthly surface temperature measurements …” and you compare them to the December ’05 to April ’06 Svalbard surface temperature measurements. You also refer to a check made with the Norwegian Met Service to make sure that your figures were correct, which clearly means that they are not gridbox temperatures.

    Could you clarify which dataset you used for your calculations, the merged Svalbard/Isfjord Radio surface temperature dataset, or the 5° x 5° gridbox temperatures?

    I ask in part because, despite an extensive search the CRU site, I cannot find any dataset for Svalbard or Isfjord Radio. Is the CRU the source of your dataset, and if so, where?

    Many thanks for your clarification,

    w.

    We’ll see … watch the pea under the shell, Michael Mann seems to be up to his old “hide the data” game again …

    w.

  121. Willis Eschenbach
    Posted May 26, 2006 at 4:26 PM | Permalink

    I give up … everything but the last line should be inside the outer block quote … showed fine in preview, didn’t post the same, go figure.

    w

  122. ET SidViscous
    Posted May 26, 2006 at 4:31 PM | Permalink

    “showed fine in preview, didn’t post the same, go figure.”

    Damn global warming effects everything.

  123. Steve Bloom
    Posted May 26, 2006 at 5:19 PM | Permalink

    Re #116: A resort to pure denialism, eh? Alternatively, you could try reading the paper itself (which I linked).

  124. Peter Hearnden
    Posted May 27, 2006 at 3:05 AM | Permalink

    Re #117 an honest admission and worthy for that imo. http://en.wikipedia.org/wiki/Mid_ocean_ridge , and ‘pillow lavas’ – http://en.wikipedia.org/wiki/Pillow_lava worth a look. We have Carboniferous pilow lavas not far away from here.

  125. JerryB
    Posted May 27, 2006 at 6:03 AM | Permalink

    Re #118 Willis,

    Thanks for that link, although it is to a different file. I found the 1996-2004 file in the directory at http://helios.hampshire.edu/~srNS/Svalbard/2004_data/Isfjord%20Temperature/ .

    Given the various weather station locations over the years, it may be necessary to be rather explicit in naming them. I would guess that your reference to Svalbard was to Svalbard Lufthavn (airport). (The whole archipelago is named Svalbard).

    For Istford Radio, I assume Kapp Linne.

  126. Willis Eschenbach
    Posted May 27, 2006 at 2:50 PM | Permalink

    Thanks, Jerry. The GISS files call it Isfjord Radio. Svalbard is indeed Svalbard Luft, which I understand means “Svalbard Airport”.

    w.

  127. Posted May 27, 2006 at 10:18 PM | Permalink

    Thanks Jerry for your 115 with details of station moves etc.
    In mid 2000 I put up a page commenting on the anomalous warmth in Isfjord Radio. http://www.warwickhughes.com/climate/isfjord.htm
    Comments by Phil Jones in those days too.
    There is a post on my Blog too.

  128. Steve McIntyre
    Posted May 27, 2006 at 10:43 PM | Permalink

    #127. I suspect that the reason why Jones won’t let people look at his data is because he’s embarrassed about it.

    I’ve spent about a morning on the temperature data and found errors in it right away – one African gridcell was out by a factor of 10 in some but not all months. You’d think that their famous quality control would have spotted it.

  129. Terry
    Posted May 28, 2006 at 8:31 AM | Permalink

    #127:

    There must be an easy (computerized) way to identify all outliers in the data similar to the African station you found. If so, it would be a very quick and powerful rebuttal of the RC Svalbard analysis: “RC claims that a 5-sigma event has only a 1 in xxx chance of happening, and so there should be essentially no such observations in the entire database. To test this, we looked to see how many 5-sigma evemts there actually are in the data and found y of them. RC, neglects to mention this fact and only reports the Svalbard outlier while making incorrect claims that such an event is highly unusual.”

  130. Steve McIntyre
    Posted May 28, 2006 at 10:30 AM | Permalink

    #129. Terry, if you click the satellite and gridcell category here. You’ll find some musings about temperature data that I did last summer. If you scroll down, you’ll see a graphic in which one site had monthly anomalies of up to 25 degrees! It was a very curious weather pattern because it happened anually for a period of years and then stopped.

    While it’s a large data set, I detected the gross clerical error literally within minutes. I simply ran autocorrelation coefficients for all the gridcells in the world and examined the outliers. Obviously they didn’t perform even this form of elementary quality control or they would have picked it up. The autocorrelation properties of the Southern Ocean data are completely implausible and this data looks particularly unreliable. In many cases, it looks like there were a few scattered measurements from presumably whaling expeditions in the 19th century separated by decades from the next data.

  131. JerryB
    Posted May 28, 2006 at 11:12 AM | Permalink

    Re #128 and #130 Steve,

    I recall you posting several times about CRU grid cell temperature files including some wild numbers in Africa. Was your mention of a station in 128 a typo for a reference to a grid cell?

    Re #129 Terry,

    Outliers may be problematic; some are real, some are errors. But a perhaps larger problem is a large number of small errors. See, for example, USHCN adjustments at http://www.ncdc.noaa.gov/img/climate/research/ushcn/ts.ushcn_anom25_diffs_pg.gif which cumulatively exceed the effects of any number of outliers. Are the adjustments correct? Place your bets.

  132. Posted May 28, 2006 at 8:11 PM | Permalink

    Re 128, 129 Thirty or so examples of Jones data gross errors are listed on my site, scroll down at;

    http://www.warwickhughes.com/climate/quality.htm

    to various links near the bottom of that page.
    The ~30 cases are at “Examples of aberrant monthly data entries and worse found throughout the Jones data” Remember these are just the standout examples. Who knows what would be turned up by more sophisticated searching.
    Then there is the link to the eyeopener, “Aberrant data from Peru”.

    In years past PDJ has told me in emails that all these errors are somehow caught by “Fortran”, presumably in the gridding stage. I am not aware this has been explained in the peer reviewed lit. I would have thought that regardless of that claim, you would want to go back and correct errors in your station data feeds.

    Re 120 I am just a bemused at anybody expecting to learn anything about CRU/Jones et al by asking MM at RC.
    The only online place I know of to begin a path of understanding is my 20th Anniversary Review site.

    http://www.warwickhughes.com/cru86/

    Reading the Fred Wood Comment plus the Wigley & Jones reply is the easiest first step.

  133. Steve Bloom
    Posted May 29, 2006 at 12:40 PM | Permalink

    Re #132: For anyone reading this who is not predisposed to global warming denialism, it should be pointed out in order for Warwick’s stuff to be true, many dozens of scientists must be either entirely wrong or engaged in some vast conspiracy. The thinking from most of the active posters on this site seems to tend toward the latter. The intent of the chief proponents of this line of reasoning seems to be to show that somehow the climate isn’t really in a long-term warming trend and so there is no global warming. Unfortunately for those folks (and in a broader sense for all of us), the physical evidence for warming is becoming very hard to ignore, e.g. the record-low Arctic sea ice and melting of the Tibetan glaciers (just to list two aspects that have made headlines within the last few weeks).

    Looking at site archives can be illuminating, e.g. this. Despite Steve M.’s show of confidence to the contrary, subsequently even Spencer and Christy admitted that any remaining discepancies between the surface record and their satellite data interpretation are likely due to errors in the latter. The comments contain some other gems. My favorite is TCO’s snark on comment #8 (“look at that big scary Powerpoint slide about increased storms this year on the RealClimate site”[here]), a remark made in early August of 2005. Irony, anyone?

  134. TCO
    Posted May 29, 2006 at 12:44 PM | Permalink

    the satelite data still has a definite difference (less) than the ground station data. same with the balloon data. The snark was completely justified.

  135. Steve McIntyre
    Posted May 29, 2006 at 12:48 PM | Permalink

    #133. Personally I have no doubt that the climate is warmer now than it was in 1850. Is it warmer than 1000? The evidence that is provided in the form of multiproxy studies is, in my opinion, not very solid. The fact that glaciers are retreating doesn’t prove that the Hockey Team reconstrucions are correct. And the Hockey Team reconstructions being incorrect doesn’t mean that CO2-forced climate change cannot be established on alternate grounds.

    I don’t think in terms of conspiracy. Does anyone think that all the brokers who recommended Enron were involved in some sort of conspiracy? Of course not. Were the wrong? Yes.

    Am I entitled to be interested in the climate of the past millennium? There’s no law against it.

  136. Terry
    Posted May 29, 2006 at 1:30 PM | Permalink

    Hey Steve, glad to have you back. Amongst the contra folks here, your posts are usually pretty high quality.

    I heard that you worked for an environmental outfit of some sort. What outfit? What do you do there?

    Thanks.

  137. Greg F
    Posted May 29, 2006 at 2:01 PM | Permalink

    …many dozens of scientists must be either entirely wrong or engaged in some vast conspiracy. The thinking from most of the active posters on this site seems to tend toward the latter.

    Come on Bloom, show us some proof that people here “tend toward” conspiracy theories. You made an assertion now put up the proof.

  138. Steve Bloom
    Posted May 29, 2006 at 3:44 PM | Permalink

    Re #136: Thanks, Terry, although I hadn’t realized I was gone. Periodically there’s not an active thread with anything of interest to me, or I get busy with something and don’t have time to comment for a few days. Regarding my environmental work, I don’t get paid for it. The principal (but not only) organization I do volunteer work for is the Sierra Club.

    Re #137: Greg F, if you really want a choice compilation of quotes from the regulars on this site that anount to conspiracy theorizing, I could spend maybe half an hour coming up with a dozen or so choice ones. I should hasten to add that while such rhetoric is not engaged in by all of the regulars, the regulars who don’t are notably uninterested in correcting even the less intellectually respectable assertions by those who do. I suppose we all must take out comfort where we can, but please note that over at RC similar such remarks on the other side of this issue are much more likely to get corrected if not deleted. Anyway, let me know if you’d like that list. Just having a quick look at this post, comment #s 69, 93 and 116 seem to fit the criteria.

  139. Spence_UK
    Posted May 29, 2006 at 4:28 PM | Permalink

    Re #138

    Unfortunately, Steve Bloom sees conspiracy theorists everywhere. Simply on the basis of my couple of posts near the top of this thread, he read my mind and determined I was a conspiracy theorist at heart. I didn’t bother defending myself because I was hoping to keep this post on topic by doing so… it seems the post is coming around full circle, and with over a hundred posts I don’t have such an issue defending myself now…

    Suffice to say, Steve’s conspiracy theorist radar is way out of whack here. I don’t subscribe to the conspiracy theory of climate science, but I do subscribe to the theory that in scientific fields with marginal quantities of accurate data, the resulting output is a better measure of the prevailing bias than the truth. There is no need for a conspiracy to exist for a bias to exist. The conspiracy theory hypothesis suits Steve’s ends better. (NB: note I said “Steve’s ends” here, not “the ends of the secret climate sect)

  140. Greg F
    Posted May 29, 2006 at 4:36 PM | Permalink

    Mr. Bloom,

    1) Do you understand the difference between proposing a conspiracy and pointing out bias?

    2) You said “the thinking from most of the active posters on this site”.

    You made a very specific accusation. Draw up a list of the most active posters and show how more then half are engaging in conspiracy theories.

    I should hasten to add that while such rhetoric is not engaged in by all of the regulars …

    You accused “most” of the regulars.

    …the regulars who don’t are notably uninterested in correcting even the less intellectually respectable assertions by those who do.

    So you are now adding a second accusation to provide proof for.

  141. Steve Bloom
    Posted May 29, 2006 at 5:33 PM | Permalink

    Re #139: Now you’re confusing me, Spence. None of the three posts I pointed to were yours.

    Re #140: OK, I’ll make a note and get to it, probably within the next few days. We’ll see what I can actually come up with in a half hour. My criterion, BTW, is simply that the content of the comment have the proper tone and seem to be seriously stated. Various defenses such as them not really being serious, not being representative of most of that person’s contributions, really just being an expression of frustration at the latest climate science crime, etc., don’t count.

    On the second point, that would be proving a negative and I am not going to spend the large amount of time necessary to do that. I would simply note that while comment #26 of this post did draw corrections from two people, the three I identified above did not, which is to say that if these are at all representative only the most extreme such comments tend to get rebuked.

  142. Dave Dardinger
    Posted May 29, 2006 at 7:13 PM | Permalink

    Steve B,

    It would appear from your choices of posts that you believe that believing a group has political motives is tantamount to believing in conspiracies. I strongly disagree as most organizations have a political orientation, in addition to any special interests they might have. To what extent the political orientation affects their activities and presentations is, of course, a matter of debate, but I’m not sure that claiming it in the case of the Chinese government, for instance, is much of a sign of conspiracy tendencies. Of course, if you think so, then I can imagine what a swoon you had when you read some of Carl’s posts!

  143. Doug L
    Posted May 29, 2006 at 8:08 PM | Permalink

    Re: Conspiracy theories

    Here is the Wikipedia entry on conspiracy theory:

    http://en.wikipedia.org/wiki/Conspiracy_theory

  144. Spence_UK
    Posted May 30, 2006 at 1:42 AM | Permalink

    Re #141

    As mentioned, I was referring to posts further up thread, specifically post #10, in which you reply to a statement I made as follows:

    it shows once again how little point there is in having these discussions with someone who is a priori convinced that the entirety of climate science is one giant conspiracy theory

    Perhaps this wasn’t supposed to be aimed at me, but it sure reads that way.

  145. Willis Eschenbach
    Posted May 30, 2006 at 2:36 AM | Permalink

    Moving from conspiracy theorists back to the subject of the thread, climate at Svalbard, I just tried to post on the thread over at RealClimate. I wanted to ask why my scientific questions were not being answered.

    Specifically, I wanted to know why they had not addressed the following:

    1) Is the use of a crudely spliced dataset for sigma calculations appropriate?

    2) What is the justification for using the 1961-1990 data (or any other subset of the full data) to calculate the sigma for a data point 26 years outside of the subset?

    Perhaps Steve Bloom or one of the conspiracy theorists would like to comment on this conspiracy to keep scientific questions off of RealClimate?

    It gets worse … or perhaps better. There is another dataset at:

    http://www.smhi.se/hfa_coord/nordklim/

    Analysis of this dataset and the NOAA dataset above shows that they both are a combined Isfjord Radio/Svalbard dataset, with the Isfjord Radio part of the dataset adjusted and the Svalbard dataset largely untouched, with the exception of a few single month corrections in the Nordklim dataset.

    The analysis of the difference between the two records shows that in the Nordklim datasets, different adjustments have been made to several separate periods of the Isfjord Radio record, viz: 1912-1931, 1931-1934, 1934-1940, 1940-1945, and post 1945. These time periods correspond very closely to the changes in station listed in comment #115 above, so I assume that they have taken great care to adjust for each time the station moved or was changed. Neither one is the same as the GISS dataset …

    The overall difference between the two datasets’ average and standard deviation was quite small. The change in the trend, on the other hand, was quite large — the trend in both the NOAA (and Mann/Jones) datasets were over .25°C per decade, while the trend of the Nordklim dataset is .15°C per decade.

    So we have three datasets in play. The first is the Mann/Jones RealClimate dataset, which is a crude splice of Svalbard and the Isfjord Radio data. The NOAA dataset is better, with at least a linear adjustment of the Isfjord Radio data. Finally, we have the Nordklim dataset, which has an adjustment for each of the moves, along with individual corrections for a number of months.

    Several things of interest. One is that the overall trend of all the datasets is not signifcantly different from zero. Despite being the poster child for global warming … it’s not significantly different from zero.

    Next, none of the datasets show the “six sigma event” April 2006 temperature event claimed on the RC site. The real numbers are 3.09 sigma (correctly calculated Mann/Jones figures), 2.87 sigma (Nordclim), and 2.47 sigma(NOAA).

    Finally, the Nordklim dataset is by far the most sophisticated splice. The Mann/Jones dataset is just two stations disguised as one. The NOAA adjustment is a straight adjustment of the mean and standard deviation of Istjord Radio, with no monthly corrections. The Nordklim dataset has actually adjusted for each move and change of the Isfjord station, as well as correcting a number of individual monthly records.

    And the trend of the best dataset? Remember, this is supposed to be the fastest warming station in the world, at over a quarter of a degree per decade … but the best data we have, the Nordklim data, shows a trend less than 60% of that, at .15°C per decade.

    The behaviour of Mann and Jones in this expisode has been nothing short of despicable. Having made a mistake, they refused to state the source of their data (surprise, surprise), refused to discuss an abysmal mathematical error, and now have closed the thread to comments. For shame.

    w.

    Tim Lambert, if there ever was a reason for a “Mann screws up again” blog thread, this is it. The mathematical error is both obvious and egregious … comment on this one, Tim, or you’ve lost every shred of credibility.

  146. Willis Eschenbach
    Posted May 30, 2006 at 2:42 AM | Permalink

    Correction — I wrote “six sigma event” above, when I meant “five sigma event” … it’s 1:45 AM, I’m working at being a sport fishing guide on the Kenai River in Alaska, days are long, sun sets at 11 and wakes up at 5, sleep is short, internet is intermittent, and I’m moving fast …

    My best to everyone,

    w.

  147. JerryB
    Posted May 30, 2006 at 3:32 AM | Permalink

    Willis,

    Alaska? Not Fiji. Hmmm.

    Back to data: which Nordklim data set at the link you posted?

  148. Willis Eschenbach
    Posted May 30, 2006 at 8:53 AM | Permalink

    Yeah, Alaska, I’m in Anchorage today, back to the Kenai (no regular internet access) tomorrow … man needs a change every once in a while …

    Nordklim_data_set_v1_0_2002.xls is a 9.1 meg Excel file containing a variety of climate data for a variety of Norwegian sites. The Svalbard station has the Norwegian station ID 99840, and the monthly mean temperature data starts on line 4639 of the sheet called “101”.

    Now that wasn’t hard … so why are Michael Mann and Phil Jones incapable of answering me when I asked them where their data came from? …

    w.

  149. Steve Sadlov
    Posted May 30, 2006 at 10:02 AM | Permalink

    RE: #133. Yeah, it’s a conspiracy alright – a conspiracy of laziness and sloppiness.

  150. Posted May 30, 2006 at 10:05 AM | Permalink

    Speaking of laziness, when are you dopes going to put up or shut up, and get a paper out? I’m sure “Mad” magazine can hardly wait! ;-)

  151. JerryB
    Posted May 30, 2006 at 11:56 AM | Permalink

    Willis,

    Thanks, I had misread part of your comment and surmised that you had found a different data set.

  152. Steve Sadlov
    Posted May 30, 2006 at 1:28 PM | Permalink

    RE: #150. Keep on assuming that no one here is published or will published. Heheheh …

  153. JerryB
    Posted May 30, 2006 at 6:05 PM | Permalink

    Re #152 Steve S,

    Be careful; you may shatter his illusions. What would he do without them?

    His apparent inability to cope with reality suggests that he would be in desperate straights without his illusional props. It would seem to be kinder to let him wallow in his fantasies, however bizarre they may be.

  154. JerryB
    Posted May 30, 2006 at 6:18 PM | Permalink

    Aside from the considerations mentioned in my previous comment, it is always good to remember not to feed the trolls. :-)

  155. Paul Linsay
    Posted May 31, 2006 at 6:06 AM | Permalink

    #155: Funny, when I looked at the data Reid is right, there’s been no warming in either satellite or ground based measurements since about 2000. Is that “conspiratorial fancy? “

  156. Greg F
    Posted May 31, 2006 at 4:50 PM | Permalink

    Re #157:

    Two things you may wish to consider about that graph: 1) NCDC actually showed 2005 as warmer than 1998 …

    The data set of which you speak is the GISS data set which just happens to be the hottest of all the surface data sets and the only one that shows 2005 as hotter then 1998. Also note that Hansen says:

    Our ranking of 2005 as warmer than 1998 is a result mainly of the large positive Arctic anomaly. Excluding the region north of 75N, 1998 is warmer than 2005. If the entire Arctic Ocean were excluded, the ranking of 2005 may be even lower.

    Our analysis differs from others by including estimated temperatures up to 1200 km from the nearest measurement station (7). The resulting spatial extrapolations and interpolations are accurate for temperature anomalies at seasonal and longer time scales at middle and high latitudes, where the spatial scale of anomalies is set by Rossby waves (7). Thus we believe that the remarkable Arctic warmth of 2005 is real, and the inclusion of estimated arctic temperatures is the primary reason for our rank of 2005 as the warmest year.

    For some perspective 1200 km is about the distance from Montgomery Alabama to Detroit Michigan or San Francisco California to Phoenix Arizona.

    Spencer and Christy have admitted) that any remaining minor discrepancies between their data and the surface record are likely due to errors in the former.

    Please provide a proper reference including page number.

  157. J. Sperry
    Posted Jun 1, 2006 at 9:50 AM | Permalink

    Re 158 (Greg F):

    Following Steve B.’s link to the April 2006 report, I found relevant material on the first two pages of the Ex. Summ. (for which Christy is one of the lead authors) and the first two pages of Ch. 4 (for which Spencer is one of the lead authors). While Steve B. says “Spencer and Christy have admitted that any remaining minor discrepancies between [S&C's satellite] data and the surface record are likely due to errors in the former”, a more precise reading of Ch. 4 indicates that there are likely errors in both data sets, and it is “generally agreed” that errors in surface data are small compared to errors in satellite data.

    Also interesting to note is this from the first page of Ch. 2 (for which Christy is the convening lead author): “Upper-air data sets have been subjected to less scrutiny than surface data sets.”

  158. Steve McIntyre
    Posted Jun 1, 2006 at 10:23 AM | Permalink

    #145. Willis,

    imagine that. They closed the thread on May 28 after only posting it on May 22. They sure don’t have much courage.

  159. Greg F
    Posted Jun 1, 2006 at 3:25 PM | Permalink

    RE:157 … hmm …

    Thanks J. Sperry.

    …a more precise reading of Ch. 4 indicates that there are likely errors in both data sets, and it is “generally agreed” that errors in surface data are small compared to errors in satellite data.

    Well “generally agreed” does not mean everybody agreed. To state “Spencer and Christy have admitted” is simply not supported by something written by committee.

  160. Steve Bloom
    Posted Jun 1, 2006 at 4:17 PM | Permalink

    Re #159: If they signed off on it, it sure does. As noted, Christy was the lead author of Chapter 4.

    Now I find that two of my prior comments (former #s 155 and 157) have been deleted from this thread. That nakes it awfully hard to participate.

  161. Steve McIntyre
    Posted Jun 1, 2006 at 5:24 PM | Permalink

    THey have have fallen prey to Spam KArma but they weren’t “deleted” – you know that, so why say it. Please don’t respond as I don’t want to waste bandwidth on this. Spencer and Christy have a recent discussion here

  162. Greg F
    Posted Jun 1, 2006 at 7:06 PM | Permalink

    Steve M,

    Thanks for the link to Spencer and Christy discussion.

    Re #159: If they signed off on it, it sure does. As noted, Christy was the lead author of Chapter 4.

    If there was ever evidence of your bias blinding you this is it. Christy was not “the” lead author. He wasn’t even one of the lead authors. The authors for chapter 4 were:

    Convening Lead Author: Carl A. Mears, Remote Sensing Systems
    Lead Authors: C.E. Forest, MIT; R.W. Spencer, Univ. of AL in Huntsville; R. S. Vose, NOAA; R.W. Reynolds, NOAA

    Contributing Authors: P.W. Thorne, U. K. Met. Office; J.R. Christy, Univ. of AL in Huntsville

    Now where you get this “signed off on it” concept I have no idea. The report is going to reflect the views of the majority. Nobody is going to have a veto.

  163. John Sully
    Posted Jun 1, 2006 at 8:19 PM | Permalink

    Jeez Greg F, it sure looks like Spencer was one of the lead authors. Or is Spencer from UAH someone else?

  164. Greg F
    Posted Jun 1, 2006 at 8:54 PM | Permalink

    Jeez Greg F, it sure looks like Spencer was one of the lead authors.

    Jeez John Sully, did I say he wasn’t?

  165. John Sully
    Posted Jun 1, 2006 at 10:20 PM | Permalink

    Actually, yes. It was fairly clear from the reference that “they”, meaning Spencer and Christy, signed off on it. A minor mistake is not that big a deal and certainly does not discredit his comment. He merely mixed up the two authors and their listing on the chapter credits.

    Of course, this is the same tactic used by most of the posters on this blog, find a minor mistake and try and discredit and entire study by pointing it out.

  166. Dave Dardinger
    Posted Jun 1, 2006 at 10:39 PM | Permalink

    Of course, this is the same tactic used by most of the posters on this blog, find a minor mistake and try and discredit and entire study by pointing it out.

    And just what do you consider a “minor” mistake? MBH98 has been shown to have major errors in every portion of the paper. Give us an example of nit-picking you’ve notice here where the poster wanted to discredit an entire paper for a small error.

    Actually if you’d wanted to find a “major” error, it was Greg addressing his post to Steve M for the link but then quoting other things from Steve B without indicating he’s switching correspondents.

  167. Greg F
    Posted Jun 1, 2006 at 11:02 PM | Permalink

    Mr. Sully you need to read what was written since you clearly have not comprehended. Nobody has a veto. You would understand that had you ever served on a committee. You don’t get to “sign off” on it. IOW, a report issued by a committee is the majority opinion. There is no reason to believe that every person on the committee will agree with every point in the report. Steve not only mixed up chapter credits but, more seriously, implied that there was only one lead author which there obviously wasn’t.

  168. Greg F
    Posted Jun 1, 2006 at 11:09 PM | Permalink

    [Dave Dardinger wrote]

    Actually if you’d wanted to find a “major” error, it was Greg addressing his post to Steve M for the link but then quoting other things from Steve B without indicating he’s switching correspondents.

    [/Dave Dardinger wrote]

    [hangs head in shame] I plead guilty! [/hangs head in shame]

  169. John Sully
    Posted Jun 1, 2006 at 11:18 PM | Permalink

    Greg, I have served on plenty of committees. On the ones I have worked on it is generally a consensus, unless you are serving in a legislature and even then many legislatures work by consensus (the student government I served in for 3 years generally worked this way).

    As for your other points: give me a break.

    And as for “major errors” in MBH 98: convince me. Steve has repeatedly misstated the main talking point of the paper. He claims that Mann said late 20th century warmth is the greatest in a millenium. The title of the paper is “Global Scale Temperature Patterns and Climate Forcing Over the Last Six Centuries” (btw, capitalization in the original title is incorrect, oh wait! That’s another error!) He clearly is not making any strong claims about the “so called” MWP which by all accounts ended in the 14th century, otherwise he would have said “Over the last Ten Centuries” or whatever Steve is saying about the study today. If you go by the anecdotal evidence from Greenland the MWP had clearly ended by the beginning of the 15th century.

  170. Posted Jun 2, 2006 at 12:16 AM | Permalink

    MBH98 was not millennial, however MBH99 was, correct?

    I’m not sure what Mann, Bradley and Hughes concluded in MBH99 but I wouldn’t be surprised if they said this was the warmest climate of the last thousand years, since I’m pretty sure nothing major changed about the shape of the reconstruction, the “tail” just got longer.

  171. Greg F
    Posted Jun 2, 2006 at 12:39 AM | Permalink

    I’m not sure what Mann, Bradley and Hughes concluded in MBH99 but I wouldn’t be surprised if they said this was the warmest climate of the last thousand years…

    This is what they said in MBH 99:

    Abstract
    Though expanded uncertainties prevent decisive conclusions for the period prior to AD 1400, our results suggest that the latter 20th century is anomalous in the context of at least the past millennium. The 1990’s was the warmest decade, and 1998 the warmest year, at moderately high levels of confidence.

    Conclusions
    While warmth early in the millennium approaches mean 20th century levels, the late 20th century still appears anomalous: the 1990’s are likely the warmest decade, and 1998 the warmest year, in at least a millennium.

  172. Posted Jun 2, 2006 at 1:04 AM | Permalink

    Thanks Greg, your facts sure beat my conjecture.

    As for the capitalisation in the title, “Global Scale Temperature Patterns and Climate Forcing Over the Last Six Centurie” looks entirely correct to me. What’s wrong with it? You could argue that “Over” should not be capitalized but that’s a bit iffy. The rule I was taught is that important words should be capitalized and unimportant ones should not be. I usually don’t capitalise “and”, “the” and “of” but do capitalise everything else. That seems to be what they’ve done too.

  173. John Sully
    Posted Jun 2, 2006 at 11:39 AM | Permalink

    Actually, I was the one who did the correct capitalization.

    BTW, MBH99 seems to be all about the stuff that Steve is always whining about: how it depends on the PC1, sparseness of data before 1400, lower confidence intervals the earlier you go… It seems to me as though MBH99 clearly acknowledged all of the problems which Steve has criticised.

  174. jae
    Posted Jun 2, 2006 at 4:26 PM | Permalink

    173

    BTW, MBH99 seems to be all about the stuff that Steve is always whining about: how it depends on the PC1, sparseness of data before 1400, lower confidence intervals the earlier you go… It seems to me as though MBH99 clearly acknowledged all of the problems which Steve has criticised.

    Man, you better do some reading in this blog! Steve has not dwelled at all on MBH’s statements about warming; his focus is on the lousy statistics and the failure to release data and methods. Regardless of Mann’s conclusions, he has absolutely no significant data to back them up!

  175. Willis Eschenbach
    Posted Jun 2, 2006 at 5:38 PM | Permalink

    #145, Steve, 6 days of discussion, might be an RC record … and I never even wrote in about autocorrelation, they shut it off before we’d even started discussing ordinary statistics.

    w.

  176. Brian Forbes
    Posted Jun 3, 2006 at 9:11 AM | Permalink

    I came across this:

    http://www.tcsdaily.com/article.aspx?id=030705C

    which relates to this thread.

  177. Jean S
    Posted Jun 3, 2006 at 10:12 AM | Permalink

    [So here is my full post. I hit "submit" accidentally earlier, so please delete (contest of) my #176 if it's still there. Sorry about that, maybe I should start editing my posts in real editors.]

    Willis asked (#99 in “MBH98 Figure 7 Unveiled”) my comments about this Svalbard issue, so I put my few cents here. I really do not have too much to add about the math part: it is just SO typical Mann. The most despicable (as Willis put in #136) thing I find in this issue is the obvious and canculated special pleading (aka propaganda via cherry picking) these guys are doing.

    First, anyone with any interest in climate knows that the basic weather patterns around here are dominated by the Golf stream. What I mean by this in this context, is that the temperature in the locations “above” Golf stream (Norwegian arctic islands, Island, eastern Greenland) seems to be similar whereas areas “below” Golf-stream (mainland Scandinavia) again seem to have similar weather conditions. To put it bluntly, temperature anomality in Reykjavik (Island; 64N 21W) does not predict well the temperature anomality in Trondheim (Norway; 63N 13E) although they are basicly on the same latitude and they are “relatively close” (1500km/1000miles). Why I have this here? Well, it has been relatively cold winter/spring this year on “the other side” of Golf stream. For instance, Jyvàƒ⣳kylàƒ⢠(Finland; 62N 25E) had the 4th coldest March on record…

    It is also telling that in their “update”

    Correction and update: (1 June) Micheal Shouler points out that we misread the previous April record (corrected above). And now that the May 06 data has come in at a record-breaking +0.9 C, our statement that April 06 was warmer than all previously recorded Mays is still true – but only just! Things move fast in this field…

    to the post they “forget” to mention when the previous records were done. But don’t take just my word for the cherry picking and twisting, actually plot some data (from, e.g., the NordKlim data set (see #145)). I plotted the mean April temperatures for the three Norwegian arctic locations. Also, see the temperature anomalities in Norway (the left picture) this spring (March-May) just kindly provided by the “Norwegian Met Service” (as they call it over RC). Notice that the Golf stream ends basicly to the north-eastern end of Norway… you’ll get the idea.

    Second, not only Mann&Jones cherry-picked, their math is terrible. Addition to what Willis & co have stated here, these northern temperature anomalities are not even close to “stationary ‘normal’ statistics” as they put it over RC. Just plot histograms of those mean monthly temperatures. BTW, the report of NordKlim is a rather good overview of the climate in Scandinavia for the past century. I think it is worth pointing out that, according to the report, the 20th century warming in the Nordic region was around 0.6C (about the SAME as the Jones (1999) global) and that 1930’s/1940’s were as warm as (or usually warmer than ) 1990’s.

    So Mann&Jones cherry-picked the (exceptional) warm weather conditions in Svalbard this year, twisted the math, and connected this to supposed “polar amplification”. Even this they managed to screw-up. Tells a lot about these “gentlemen”.

    Their post ends:

    But in a statistical sense, large outliers like this make it more probable that the underlying distributions are shifting and give us a glimpse into the types of anomalies we might expect to become more common in the decades ahead.

    I wonder what they would derive from the facts that all time high temperature record in Finland was set in 1914 and the all time low in 1999. Hel(l)sinki is freezing in the near future?!?

  178. Jean S
    Posted Jun 4, 2006 at 10:30 AM | Permalink

    Some additional links. There is an interesting April 2006 temperature graph (CRUTEM3) prepared by Prof. Jones’ own team. I wonder why Mann&Jones did not choose a station from Russia to talk about, e.g., the future of the permafrost in Siberia ;) Also the temperatures for the mainland Norway for the past 100 years can be seen from here.

  179. Jo Calder
    Posted Jun 5, 2006 at 5:29 AM | Permalink

    In another place, there’s a discussion of a On the Role of Global Warming on the Statistics of Record-Breaking Temperatures by Redner and Petersen, dated 17 Jan 2006, and marked as submitted to Journal of Climate. To my untrained eye, it looks like it assumes iid, but I’d welcome more informed comment.

  180. Willis Eschenbach
    Posted Jun 5, 2006 at 3:53 PM | Permalink

    Re 177, Jean S, thank you for your comments.

    I was particularly interested in getting a comment about their procedure of taking a 30 year subset of the data (1961-1990), figuring the mean and SD (incorrectly, as it turns out), and then using that to derive a probability (1 in 10^6) for the April 2006 temperature.

    I say this is not a valid procedure, that you have to use the mean and sd of the entire period in question, including the data point being examined.

    People have replied that they did it that way because the WHO uses a set 30 year period to derive their global temperature comparisons.

    I say WHO approval of that period does not make the procedure correct, as the laws of statistics are set by nature, not the WHO.

    Your comments?

    Thanks again,

    w.

  181. JSM
    Posted Jun 5, 2006 at 9:15 PM | Permalink

    It seems that hadCRU uses 1961-1990 as their baseline. That is why they cite 1961-1990 as the basis for sd and mean — it has already been calculated. Since the post was a quickie, just noting a rather large deviation from the mean, with no evidence that any major statistical analysis was done, why are you complaining so much about their method? It really seems rather pointless.

  182. Jean S
    Posted Jun 6, 2006 at 4:57 AM | Permalink

    re #180: If you assume your data is stationary, then estimating the mean and variance from a shorter period gives you a worse estimate than using the entire time period. I can not figure out any mathematical reason NOT to estimate them from the entire series. Maybe Mann & Jones do…

    I would not put my money even on the stationarity of the temperature series. Additional flaws in MJ analysis include:

    1) the right question is not what is the probability of a “5 sigma event” happening in Svalbard record,
    but what is the probability of a such thing happening in some record (such things are bound to happen somewhere even under stationarity and normality assumptions)
    2) the series in question is not even close to normal (it seems actually rather “heavy-tailed”), so picking the probability from normal distribution is meaningless (in this case it does not give even a decent approximation).

    re #181: IMO, it is pointless to put some misleading “statistical analysis” just to highlight the fact that the mean April temperature in Svalbard this year deviated from the 61-90 mean unless you want to mislead not-so-mathematically-oriented readers. And you have to be really stupid not to understand what they are implying with their post. I’m also rather sure this “5 sigma event in Svalbard” is going to appear in popular press pretty soon…

  183. Steve Sadlov
    Posted Jun 6, 2006 at 11:24 AM | Permalink

    RE: #178 – Both the intensity and peristence of winter conditions in far NE Europe and far Northern Asia this past season were indeed truly remarkable. Warmers hate that.

  184. Willis Eschenbach
    Posted Jun 6, 2006 at 4:12 PM | Permalink

    182, thank you very much, Jean.

    181, JSM, thanks for your post. They have assuredly done a statistical analysis, wherein they breathlessly announce that it was a “5 sigma event” with odds of less than one in a million of happening.

    However, different 30 year baseline periods give different results. I calculated all the 30 year periods, and you can get odds ranging from one in 287 to one in over 3 million … pick your period, and you can prove anythinbg.

    To use such a bogus analysis as a scare tactic to promote the idea of “arctic enhancement” of global warming is scurrilous and despicable. The warming in Svalbard occurred prior to 1920. Since then, the warming there has been 0.07° per decade, not statistically different from zero.

    That’s what I object to, using phoney “statistical analysis” to try to prove what isn’t true … and to make it worse, they closed the thread immediately when I tried to point that out to them. That’s not science, that’s science fiction.

    w.
    w.

  185. jae
    Posted Jun 6, 2006 at 5:01 PM | Permalink

    Come on, Willis, you are insulting the science fiction genre. This stuff is far too bad for science fiction; it is just sleazyscience. No wonder they closed the thread; they have to be pretty embarassed by now.

  186. UC
    Posted Jun 7, 2006 at 11:06 AM | Permalink

    Re 182:
    Short sample sd of positively correlated random process tends to underestimate the variability of the process (at least in the case of stationary process )

    One more 5 sigma event:

    Use http://fox.rwu.edu/~rutherfo/supplements/jclim2003a/ combinedannfullnh.txt, compute sample mean and sample deviation of years 1461-1490. Compare the result with year 1439 value.

  187. Steve McIntyre
    Posted Jun 7, 2006 at 11:47 AM | Permalink

    Take a look at the example in this post http://www.climateaudit.org/?p=342 for the effect of short samples. Also look at the post on Trenberth 1984 – see Right Category statistics and scroll down.

  188. UC
    Posted Jun 7, 2006 at 2:25 PM | Permalink

    Re 187:

    I see, you have gone through this already ;) Anyway, all this reminds me of ‘wishful data interpretation’
    ( http://en.wikipedia.org/wiki/Pathological_science )

  189. Steve Sadlov
    Posted Aug 24, 2006 at 4:56 PM | Permalink

    On the topic of Arctic Sea ice….

    I will make somewhat of a forecast here.

    At present we are at, or very near, the Northern Hemisphere annual extent minimum. Depending on which measures you use, we are either equal to or slightly higher than last year in terms of extent on the same date. I predict that we will reach this year’s minimum at an earlier date than we did last year. I predict that 2006 will not be an “all time low sea ice extent” at the minimum.

    There are some areas where barring a major reveral, extent is already clearly in “fall” mode.

    The lower extent in the Eastern Arctic (as a result of the known positive SST anomaly in the NE Atlantic) is balanced by the extent in the Western Arctic. Beaufort Sea and Chuckchi Sea both currently have a zero anomaly according the Cryosphere Today.

  190. David Smith
    Posted Aug 25, 2006 at 5:33 AM | Permalink

    Re #189 I agree with your forecast. My impression, not quantified, is that is has been cloudier/foggier than normal along the Arctic shores and maybe over the ice, too.

    To me, cloud cover plays a major role in whether ice melts or freezes. If the Arctic sun shines in summer, ice melts. If the sun is blocked by summer clouds, the ice melts only slowly. If Arctic skies are clear in the winter at night, temperatures drop. If wintertime night skies are cloudy, there’s less cooling.

    I’d like to see some kind of Arctic cloud cover index, to see if there are patterns. I know there are weather oscillations, like the NAO, but I have not seen data on how things like that affect cloud cover.

  191. Dave Dardinger
    Posted Aug 25, 2006 at 8:19 AM | Permalink

    re: 190

    I’d seem to me that after nightfall in the arctic, i.e. for six months, there will be little evaporation due to no energy (from the sun) to overcome the latent heat of evaporation, though there will be some from wind. This probably means there are few clouds except around the edges where there’s snowfall from “incoming” air. This would mean that mostly there will be clear skies. Where there is ice, the surface temperature will drop rapidly and then the ice will act as a barrier for passing heat through to the surface. Where there’s clear sea, huge amounts of energy will be lost to space, even with added CO2 to slow things a mite. So, the upper waters will quickly be reduced to the freezing temperature and then ice will form. I’d be nice to have a reference to heat loss from ice according to thickness if anyone knows of one. (I’ll go looking after I post this and report back if I find anything useful.) I suspect that even if all ice were melted in the arctic, the time to complete icing over of the arctic circle wouldn’t change too much.

  192. Dave Dardinger
    Posted Aug 25, 2006 at 8:37 AM | Permalink

    Here’s a link to a table containing the thermal conductivity of ice. As you can see the conductivity doesn’t change too much till you get to quite cold temperatures, so we could assume about 2.5 watts per meter-deg K. So if someone could calculate what the IR loss from water at about sea water freezing would be, we could compare that to the loss after a meter or two of ice forms. Presumably the loss of heat via IR radiation will match (i.e. be limited by) the heat conducted through the ice. To be even more explicit, the surface temperature of the ice will decline until the amount coming through the ice is balanced by the amount radiated away. The loss from open water, OTOH, will be limited by the radiation since convection processes will maintain the surface at a relatively uniform temperature until ice forms.

  193. David Smith
    Posted Sep 7, 2006 at 7:41 PM | Permalink

    Miscellaneous notes:

    Looks like the Arctic sea ice is close to its September minimum. 2006 should not be a record low. September southern hemisphere (winter) sea ice is a few percent from its 30-year record high.

    The NOAA Arctic watchers agree that Arctic cloud cover was heavier than normal in August, and actually cooled the Arctic a few degrees below (30-year mean) normal.

    Forecasted upper-air patterns will make it tough for any major hurricane to threaten the US in the next 10 days. The historical peak of the season is September 10 (Sunday) and, historically, the season winds downs relatively fast thereafter. Steve M’s musings about persistence look to be on the mark.

  194. Steve Sadlov
    Posted Sep 7, 2006 at 7:47 PM | Permalink

    Looking on the Cryosphere Today site, a number of basins actually have a slightly positive anomaly at present and have incurred at least one week of growth. Most notably, the abnormalities in the NE Atlantic and the Arctic near Russia appear to have largely abated. We may have already hit the technical minimum. Unless there is a late reversal, this year’s technical minimum may register as having been in August.

  195. David Smith
    Posted Sep 19, 2006 at 8:58 PM | Permalink

    The US ice folks indicate that the 2006 Arctic sea ice extent will not set a record low (about 30 years of records, I believe).

    In fact, it may end up about 4’th lowest, assuming no late-season warm fronts.

    Ice area (the area covered if all the ice was squished together) should also not set a record low this year.

    This, of course, is weather, not climate, and is not a trend so far as I know.

    What’s interesting is that a lot of ice was preserved this summer when Arctic weather patterns shifted to a cloudier, stormier pattern in late July. The power of clouds and water vapor!

    This coincided with a strengthening of 500mb winds (the ones that help shove weather around). Here in the 30N western hemisphere we are still seeing stronger-than-normal upper winds, which have brought us our first real cool front of the season.

  196. David Smith
    Posted Sep 28, 2006 at 9:01 PM | Permalink

    Here is a link to the September 27 Arctic sea ice outlook.

    It looks like 2006 will not be a record low ice extent. In fact, it may end up third or fourth from the lowest. That is quite a change from 2005.

    To me, the most interesting plot is the one of ice extent versus time (months of the year). Note how we were headed towards a record low as late as July, but then conditions shifted and the Arctic was able to “keep its cool”.

    This is about the time that the hurricane season sort of collapsed and the weather charts started showing faster movement of systems in the mid-atmosphere.

    I wish we knew why.

  197. David Smith
    Posted Oct 3, 2006 at 7:52 PM | Permalink

    The final special report on Arctic sea ice is here.

    Sea ice extent this year was the fourth-lowest on record, which is greater than had been expected following 2005’s record low.

  198. David Smith
    Posted Oct 22, 2006 at 7:26 AM | Permalink

    Update on Northern Hemisphere ice and snow cover –

    October 21, 2005 ice and snow cover is shown here.

    October 21, 2006 ice and snow cover is shown here.

    It looks like global warming is continuing to take a slight pause. Both snow cover and Arctic ice extent are up.

    As always, I mention that this is weather, not climate, and is likely just part of the normal year-to-year variation. The Arctic ice cover long-term trend remains downward.

    The thing that has impressed me about this years ice recovery is the aggressive ice growth in Atlantic waters. In a few isolated Atlantic spots, current ice extent appears to slightly exceed the long-term (1970-2000) average.

Follow

Get every new post delivered to your Inbox.

Join 3,246 other followers

%d bloggers like this: