Open Mind

Arctic Stations

September 10, 2009 · 63 Comments

I almost didn’t think it was possible. Anthony Watts has yet another post attempting to support the totally false claim that the arctic isn’t warming. This claim definitely fits into the “bag of hammers” category; the post is courtesy “Lucy Skywalker Green World Trust,” at whose site you can find the thrust of the post, this map. It links to graphs of temperature from northern stations.

No such graphs would be complete without a caveat, courtesy of Paul Vaughan, which tells us a lot about how they’ve given such a completely wrong impression:

Paul Vaughan notes at WUWT that he “spent a fair amount of time updating these graphs (& others of Daly’s for other regions)” using and adds a cautionary note: The time-frame and aspect-ratio of the timeplots can be manipulated to create the illusion of a steep trend in recent years.

Other than plotting such a short time span that there’s nothing but noise on the graph, how can you set the time frame and aspect ratio to create the “illusion” of a steep trend? But it’s easy — ridiculously so — to set the time frame, axis scale, and aspect ratio to give the illusion of no steep trend. That’s exactly what’s been done, but it’s only one of the ways in which these graphs are manipulated to say the thing which is not.

Consider the data for Eureka, Northwest Territories. Here is the graph they give you:


Note that the y-axis extends a full 55 degrees C (that’s 99 deg.F). How can you expect to see a significant change, even a very large change, say 2.5 deg.C (which climatologically speaking is huge), when the y-axis extends 55 deg.C? Don’t forget to put 5 time series on the same graph so it’ll be more cluttered and even harder to see the details. While you’re at it, chop off the data at the year 2003, even though the available GISS data continue up to the present. Here’s a better graph of temperature in Eureka:


I’ve made the y-axis extent small enough that you can actually see what’s happening, and used all the available data. Is that what Paul Vaughan means by “manipulated to create the illusion of a steep trend in recent years”?

Ending at 2003, rather than showing all the data up to the present day (which is the subject at hand, isn’t it?) is a form of cherry-picking — the 2nd method used to deceive. It’s also used for the graph of data from Talkeetna, but this time they don’t even go up to 2003, the plot ends in 1999:


The GISS data for Talkeetna go up to 2004, are oh so easy to get, and those last five years tell a very interesting story:


It turns out that most of the stations plotted through this arctic map will show a recent upward trend, if the plots aren’t manipulated to suppress it by excluding recent data, or by making the y-axis extend so far that everything looks flat.

Imagine that.

It’s also ironic that in a previous post, Paul Vaughan criticized me for implying that smoothing data is always a bad idea, in spite of the fact that I had implied no such thing. In fact smoothing is a tremendously useful process; when done properly it’s a terrific way to suppress noise, and is especially good for viewing graphs to get a much better idea of what the trend might be. Just graphing the data with no noise reduction, and performing no analysis, makes it easy to miss the long-term behavior. Drawing conclusions about trends still requires analysis — but looking for them on graphs is greatly aided by noise reduction.

Yet in none of these graphs is any noise reduction attempted, not by sophisticated smoothing methods (lowess or wavelets) or by brute force smoothing (averages or moving averages). That’s a pity, because smoothing reveals the trends with tremendous visual impact. Here’s a lowess smooth of the temperature anomaly from Eureka:


And here’s the smoothed temperature anomaly from Talkeetna:

The author of the no-arctic-warming post begins by saying “What sudden recent warming? What Hockey Stick? I don’t see any.” With well-known and robust noise reduction, even Anthony Watts can see it.

As for cherry-picking, leaving out recent data is only one way to do that. Another is to leave out data records.

If you wanted to know about the far north, how would you select your data? Here’s a good start: go to the GISS station data page and click your mouse as near the north pole as you can, to find the stations nearest the pole. Then you might even limit your results to stations with data extending at least partway into the 2000s. That would give you 14 stations, the ones nearest the pole with some data in this millenium. It’s certainly a reasonable way to begin, and every one of those stations should surely be on your list to examine.

Yet only 5 of those 14 stations are included on the arctic “map.” Some of them should be omitted; although data extend into the 2000s there are large gaps which make interpretation and analysis difficult. But some of them are so obviously relevant, one has to wonder why they’re not included. For instance, the author includes 6 stations from Alaska but omits the northernmost of all, Barrow. Maybe that’s because the Barrow data look like this:


Smoothing the Barrow data makes the trend even clearer:


Another station which should certainly be on the list is Resolute, in the Northwest Territories:


Smoothing the data shows quite a “hockey stick”:


Yet another curious omission is Svalbard, which has only 30 years of data but is surely relevant to the recent trend in the arctic:


The smoothed version shows another “steep trend in recent years.” For those who are interested, actual analysis confirms it without doubt.


What’s the bottom line? Most of the stations selected for the “map” actually contradict the author’s claim when they’re plotted with all the data, and at the right scale for people actually to see the details. Noise reduction makes the predominant pattern so obvious, nobody in his right mind can deny it. Analysis confirms those conclusions. And, including many stations which definitely should be on the list but are left out, strengthens the conclusion: Yes, Virginia, the Arctic is warming. Fast.

Yet we have to endure yet another post claiming that which is not. This is supported by several techniques, including:

  • Manipulating the time-frame and scale of the plots to create the illusion of little or no trend in recent years.
  • Avoiding noise reduction which would make trends blatantly obvious.
  • Cherry-picking, both time frames and station records.

Some may consider it a waste of time to dwell on what’s so obvious. But in their attempts to discredit arctic warming, Watts & Co. are trying to undermine the truth. I’m drawing the line — at the arctic circle.

Categories: Global Warming

63 responses so far ↓

  • Paul Tonita // September 10, 2009 at 2:50 am | Reply

    Nicely done Tamino. A very nicely drawn line.

    On that topic, I’ve been wondering lately, why do Arctic tamperature profiles usually extend from 60° North to the pole? I ask because someone tried to tell me recently that there were dozens of stations that showed the 1930’s were warmer than the 90’s. They didn’t offer any ref’s for that, and the only figure I could find online to support that was a graph which defined the Arctic as everything from 70°North to the pole.

    I don’t know if that distinction turns out to be very important, except for maybe number of stations perhaps.

    [Response: Traditionally the arctic includes areas north of the arctic circle, i.e., those which are in all-day sunlight at the summer solstice. As for the climatological use of the term, I can only guess that it includes areas north of 60deg.N latitude because that's where most of the cryosphere is.]

  • barry // September 10, 2009 at 3:11 am | Reply

    Thanks, Tamino. I had wondered why stations were missed and why no analysis was done. Didn’t even think about the Y axis shemozzle. I don’t have the skill to do what you do, and really appreciate you spelling it out for us non-statisticians.

    Question, slightly off-topic:

    Thinking of the 2007/2008 Arctic sea ice minima, where 2008 was greater extent than 2007 – but the declining trend steepened, is it the case that each subsequent anonaly that falls below the trend line will steepen it, and the opposite if the next anomaly falls above it?

    My hunch is that the trend has already steepened from this year’s current extent.

    [Response: Generally yes, if a new data point is below the existing trend line then the updated trend will be even more negative.]

  • Deep Climate // September 10, 2009 at 3:56 am | Reply


    I think the trend line of most interest is arguably that for average sea ice extent in September, which represents the annual summer minimum.

    It looks like August 2009 was pretty much right on the trend line (for August), so chances are September 2009 will be too (and the trend line of minimum extent will be more or less unchanged).

    Of course, you can bet there will be a lot of braying over at WUWT about how the sea ice continues to “recover” since 2007.

  • barry // September 10, 2009 at 5:46 am | Reply

    Thanks, DC. Yes, I was thinking of the September minimum. I asked a WUWT participant to calculate that trend including current data and he said it had steepened.

    Tamino, the reference to ‘hockey stick’ – is Lucy Skywalker referring to anything you (or anyone else) previously posted? As I understand it, ‘hockey stick’ refers to millenial temperature reconstructions. I thought it was odd (propagandistic) to make that reference in a post about the ~100 year instrumental record, and suggested that no one claims the instrumental record has a hockey stick shape.

    [Response: I was just poking a needle in the opposition.]

  • Aslak // September 10, 2009 at 6:28 am | Reply

    There is a splice of svalbard instrumental records called “svalbard airport reconstructed” which starts 1911. It can be found here

  • Slioch // September 10, 2009 at 10:10 am | Reply

    Thanks Tamino

    Very timely. Lucy Skywalker had just come up on the UK Daily Telegraph, Christopher Booker, blog:

    on September 09, 2009
    at 08:10 PM]

    so it is useful to be able to link to your demolition of it. [Hopefully to be posted soon by the DT]

    BTW, I don’t at all ” consider it a waste of time to dwell on what’s so obvious”. Every single person who is misled and manipulated at WUWT and similar sites has a vote and can lobby their political representatives. We are in the middle of a political battle to make changes, and people and their views, however misguided, influence politicians.

  • Barton Paul Levenson // September 10, 2009 at 10:26 am | Reply

    Paul Tonita: someone tried to tell me recently that there were dozens of stations that showed the 1930’s were warmer than the 90’s.

    BPL: This is a major denialist line recently, and is a nearly complete lie. Here’s the truth:

  • Mark // September 10, 2009 at 10:54 am | Reply

    I would like to see the smoothed graphs at the same scale as the unsmoothed data.

    You can easily see the shape is right, but the zooming in that is the result of removing the wild variations (noise) around that mean (signal) makes it require a few looks at both graphs to confirm.

  • Kelly // September 10, 2009 at 11:06 am | Reply


    Excellent post. Clearly demonstrates the importance of effective analysis rather than just eyeballing the data.

    Proper charting techniques are critical in evaluating climate data. Here’s an example of poor charting at

  • TrueSceptic // September 10, 2009 at 11:21 am | Reply


    Thanks again for the expert analysis.

    We can only guess whether incompetence or dishonesty drives Watts & co.

    BTW it is amusing to see Girma Scaling being used at Watts. You see, using anomalies or an appropriate Y-axis is scary whereas using the full °C range is comforting and means the changes can be ignored.

  • Mark // September 10, 2009 at 12:05 pm | Reply

    “Of course, you can bet there will be a lot of braying over at WUWT about how the sea ice continues to “recover” since 2007.”

    Mind you, if they have to, they’d bray about that every winter.

    Until even the winter extent is less than the 2007 summer extent…

  • Dappled Water // September 10, 2009 at 12:38 pm | Reply

    Although sea ice extent is a useful measure, the actual volume of sea ice is an indicator too. According to NSIDC 2008 represents the lowest volume on record:

    Most of the Arctic sea ice is now 1st year and will melt quickly if conditions anywhere near 2007 occur. Don’t think we will have to wait long to see the 2007 record broken.

    Thanks for the tireless work debunking these denier lies Tamino. I certainly learn a thing or three.

    Whilst the talk is about the Arctic, have you seen the James Balog talk at TED?. The footage of the Ilulissat Glacier in Greenland is incredible.

  • Mark // September 10, 2009 at 1:44 pm | Reply

    “Although sea ice extent is a useful measure, the actual volume of sea ice is an indicator too.”

    Like the BMI, it’s a good measure if it’s all you can get.

    BMI *tends* to indicate you’re overweight. But not all that reliably.

    Sea Ice extent *tends* to indicate sea ice volum, but not all that reliably.

    BMI: Muscle doesn’t count, nor body shape

    Sea ice extent: depth doesn’t count, nor does ice survivability

  • Paul Tonita // September 10, 2009 at 2:28 pm | Reply


    Yes, the fellow has trotted that canard out before. He parrots the stuff that doesn’t matter, and frankly isn’t interesting. He didn’t have anything to say about the reconstruction, and the import of this new contribution. Like everyone here probably knows, it’s that dang signal to noise ratio…

  • Philippe Chantreau // September 10, 2009 at 3:12 pm | Reply

    Mark @ 10.54, keep in mind the smooth is of temp anomaly, whereas the data is temp.

  • Juliette // September 10, 2009 at 3:20 pm | Reply

    I know this doesn’t come from a major blog, so no need to spend time debunking, but I found the most amazingly inappropriate trend line in a graph here:
    I had to check several times whether or not this was irony brought to a new level. Look at it, if only for posterity.

  • Eric L // September 10, 2009 at 3:45 pm | Reply

    You know, we were having a similar discussion in the most recent thread over at Denial Depot. He should graph the temp in Kelvins so we can see how much temps have really changed and not get the visual distortions above.

    • TrueSceptic // September 10, 2009 at 6:21 pm | Reply


      Quite. I confess I find it fun trying to guess who’s who at DD (I don’t want to be told, of course). I’m sure some must frequent Open Mind and Deltoid.

  • Hank Roberts // September 10, 2009 at 6:18 pm | Reply

    Tamino, you have the beginnings of a book here.

    Pity the title “How to Lie With Statistics” has been used already. But these examples are so much better ….

  • Hank Roberts // September 10, 2009 at 6:31 pm | Reply

    From NSIDC:

    near the top:

    September 8, 2009
    Winds cause sea ice to spread in August
    … Atmospheric circulation patterns in August helped spread out sea ice, slowing ice loss in most regions of the Arctic. … this year’s minimum ice extent will probably not reach the record low of 2007 …

    further down the page:

    “Studies by Mark Serreze, Masayo Ogi, and other researchers have shown that low-pressure patterns promote spreading of the ice pack, a process known as divergence. While ice divergence increases extent, it can also accelerate melt because there are more dark, open-water areas between the floes to absorb the sun’s energy, promoting melt on all sides of the floes.”


    My conclusion, just as a reader:

    Extent is not as low; ice has been spread out due to wind. Left unsaid yet — how much ice remains? How much first year ice will we see, assuming the ice stays spread out this way? If so it will forming with little chunks of 2nd-year ice mixed through it.

    Need more higher resolution imagery from the defense satellites!

  • Paul Tonita // September 10, 2009 at 7:08 pm | Reply

    How would graphing temperature in Kelvins be any different than in Celcius? There is no difference between a change of +1.4 °K from 273°K and +1.4°C from 0°C. The distortions would be the same. It’s the choice of window that makes the overall trend difficult to eyeball, plus the addition of other series which appear to justify the window.

  • Derecho64 // September 10, 2009 at 7:41 pm | Reply

    I had an old boss who once said “A carefully-chosen contour interval can hide a great many sins.”

    Watts and his minions over at WTFWT have discovered that principle. I know that for my work, I choose a y-axis that goes from -∞ to +∞, but strangely, all my plots are perfectly horizontal lines.

  • dhogaza // September 10, 2009 at 7:45 pm | Reply

    How would graphing temperature in Kelvins be any different than in Celcius? There is no difference between a change of +1.4 °K from 273°K and +1.4°C from 0°C. The distortions would be the same

    Spend some time over at Denial Depot and I think you’ll figure it out :)

  • Paul Tonita // September 10, 2009 at 8:31 pm | Reply

    Spend some time over at Denial Depot and I think you’ll figure it out :)

    Lol…well now don’t I feel silly. That’s pretty funny. If it was in video form, he could have a regular segment on The Colbert Report!

  • Kelly // September 10, 2009 at 9:22 pm | Reply


    Since “How to Lie With Statistics” has been used already, what about Lying Climate Charts!

    Here’s a link to my post about an Australian Senator’s Lying Climate Chart.

    Should we start collecting these?

  • David B. Benson // September 10, 2009 at 9:36 pm | Reply

    Meteorlogically, there are three zones in each hemisphere. Starting at the euqator (approximately) the northern Hadley cell extends to about 30N. From 30N to 60N is the temperate zone, From 60N up is the polar cell. All approximate.

    From 60S going south there is essentially nothing but Antarctica, just a little icy ocean. BC and the prarie provinces end at 60N, Yukon and Northwest Territoriees above that. Orkney Islands, Scotland, and St. Peterburg are about 59N. Helsinki, Finland, is 60N.

  • Ray Ladbury // September 11, 2009 at 12:28 am | Reply

    Kelly, How about ” lying sacks of a certain messy substance that is the product of normal biological function”.

  • David B. Benson // September 11, 2009 at 1:40 am | Reply

    Oops. Antarctica not that big. It mostly fits within the Antarctic Circle:

  • barry // September 11, 2009 at 4:21 am | Reply

    Extent is not as low; ice has been spread out due to wind. Left unsaid yet — how much ice remains?

    Would I be right in thinking that you could get that answer by looking up sea ice area?

    From this information:

  • barry // September 11, 2009 at 4:23 am | Reply

    First para above should have been blockquoted – replying to Hank above.

  • Marion Delgado // September 11, 2009 at 4:55 am | Reply

    The Climate Pages (2 sections – more than is devoted to any other subject) at that site are a definitive list of all the corporate shills and anti-environmentalist cranks in the business.

    I call fake site, fake group, and all to push the meme, as “Spiked” and “Real Marxism” do, that people who are actually in corporate pay are just wanting to help the Third World and puppies.

    Just like all the tobacco lobbyists still saving Africa from Rachel Carson.

  • Marion Delgado // September 11, 2009 at 5:04 am | Reply

    And even if the GWT is for real (all you can find from outside sources is that they own about a third of Neah Power Systems [NPWZ]), then I think “Lucy Skywalker” and crew are keeping their friends and colleagues in the dark and feeding them nightsoil.

  • Juliette // September 11, 2009 at 11:06 am | Reply

    OK, We’ve come to expect nothing else from Fox News, but this is beyond belief:

    (disclaimer: this video is partly about Greenpeace, where I work)

  • Mark // September 11, 2009 at 12:26 pm | Reply

    Phillipe: “Mark @ 10.54, keep in mind the smooth is of temp anomaly, whereas the data is temp.”

    I know. And smoothed. But if you print both graphs out for one dataset (smoothed and raw) on thin paper (or trace out on tracing paper), then place one on top of the other, they don’t fit by eye.

    Purely because the auto-zoom is decreasing the axis since it doesn’t diverge as much from the “mean”.

    And temperature anomaly is just “how much different from a reference point is this” and so the raw temps can be done against the same reference point.

    Just imagine both graphs overlaid on each other.

    The smoothed graph should follow the rough path of the raw graph.

    Heck, try it yourself. Use photoshop/Gimp/PSP/…

    You’ll find that you can get them to match up better if you zoom out on the smoothed graph.

  • Mark // September 11, 2009 at 12:34 pm | Reply

    E.g. take the Svalbad (the one that needs less work to fit together) images.

    On the anomaly graph, I would put the -2 line somewhere around -7 on the raw data graph and the +2 anomaly line somewhere around the +3 on the raw data.

    With that scale change, it’s easier to line up the two dips at around 1993 and 1999 to the dips on the other graph.

    And laying them over helps show why the sharp drop in the raw data at about 1988 gets very nearly nuked.

  • Kelly // September 11, 2009 at 1:20 pm | Reply


    Your link to Fox- O’Reilly video really surprised me. I understand O’Reilly, my surprise is with AccuWeather’s representative.

    Is Bastardi’s position AccuWeatether’s position?

    Is AccuWeather claiming global cooling?

  • Kelly // September 11, 2009 at 1:38 pm | Reply


    I just checked AccuWeather’s climate change blog writen by Bret Anderson. Bret has a post about Joe Bastardi’s Fox video and a statement by Bastardi that you should read.


    Looks like some ( at least one) Accuweather meteorologist(s) are not convinced about the role of CO2.

    Accuweather is a major player in the weather industry. I know where Fox stands on climate change, I’d like to know where Accuweather stands on climate change.

  • dhogaza // September 11, 2009 at 1:45 pm | Reply

    Would I be right in thinking that you could get that answer by looking up sea ice area?

    Yes, except for the fact that apparently the measurement uncertainty for the satellite sea ice area calculation is somewhat higher than for sea ice extent.

    Apparently that’s why NSIDC and IJIS/JAXA concentrate on sea ice extent rather than area.

  • dhogaza // September 11, 2009 at 1:54 pm | Reply

    Is Bastardi’s position AccuWeatether’s position?

    Is AccuWeather claiming global cooling?

    AFAIK it’s not official, but Bastardi’s their most senior and best-known forecaster and AccuWeather makes no effort to put any distance between themselves and Bastardi’s outspoken denialism.

    Here in PDX at least AccuWeather is far less reliable than the NWS weather forecasts which Bastardi also loaths. A few years ago one of PA’s senators entered a bill that would’ve made it illegal for NWS to provide free public forecasts, to “even the playing field” for AccuWeather.

    Fortunately it didn’t pass.

    I just checked AccuWeather’s climate change blog writen by Bret Anderson. Bret has a post about Joe Bastardi’s Fox video and a statement by Bastardi that you should read.

    Bastardi shows a firm grasp on the science, doesn’t he? He’s almost as well-informed as Watts.

  • Mark // September 11, 2009 at 2:11 pm | Reply

    Gross errors with Bastardi:

    1) To be sure, the cooling now is happening faster than it should because it got so warm in the first place via the super nino of 97-98.

    Nope, it’s not even cooling.

    Colder doesn’t mean cooling when you’re talking about climate.

    2) The last warm cycle reached its peak in the late 50s when we actually surfaced submarines at the north pole in 1959… IN MARCH!

    Cycle? What’s causing that cycle and why aren’t we at the same temperature now as we were in the last repeat of that “cycle”?

    And the north pole isn’t changed by the extent or thickness of ice. A strawman. Might as well say since man has been to the North Pole before, it must have been as warm then too.

    3) We can’t know till after the period that is coming up through 2030 whether co2 is really a player or not.

    Why? What will we learn by then that we don’t know now?

    After all, there’s no cycle evident so far because despite having a cooling cycle period from the ocean and a cooling cycle from the sunspot activity, we are still warmer by quite a lot than last time we had such a confluence from these two.

    And you would only know “by the end of this period in 2030″ if you had some idea of what the cycle was.

    He doesn’t say what it is, though.

    Probably because concrete predictions can be tested where handwaving arguments can’t.

    4) Common sense dictates that a trace gas needed for life on the planet would not be the cause for destroying life on the planet.

    Why does common sense say that? H2O is needed for life on the planet, but we still drown in it.

    Iron is a trace element: without it you die. But you can still be poisoned by it.

    Same with cyanide: needed to kill off infected cells, but take a teaspoon, and you’re dead.

    Problem with common sense is he doesn’t seem to have it.

    5) In the end, for me its all about getting the weather right.

    And for a weatherman, this is right.

    But if this is all his needs cover, he’s not a climatologist.

    6) What I opine about global warming is not because this is my goal…to be in this debate, it’s a by product of the work I put in to do what I was made to do..forecast the weather

    So he’s not a climatologist. He’s a weatherman.

    But the climate in 2100 doesn’t change his weather forecast, so he doesn’t care or even need to know what it’s doing.

    So it’s easy to say whatever will get you on the lecture circuit rather than spend time and effort trying to find the truth.

  • Mark // September 11, 2009 at 2:13 pm | Reply

    This is also revealing:

    “Please remember my opinion is my own, but perhaps the reason for my successes in this field are because of the countless hours I spend that no one but me and the good lord above sees in preparing to answer the call.”

    So AGW can’t be right because only the Lord God can do this?

    And the successes in his field may exist, but his field isn’t climatology.

    It’s a very PHB attitude, isn’t it: I’m great at people management skills so I must be a great nuclear physicist.

  • dhogaza // September 11, 2009 at 3:39 pm | Reply

    The last warm cycle reached its peak in the late 50s when we actually surfaced submarines at the north pole in 1959… IN MARCH!

    The real problem with this statement is that leads open up in the arctic sea at all seasons, and that submarines surfacing up there find open leads or areas of thin ice they can surface through (up to about 3 feet thickness IIRC) if that’s part of the mission.

    It’s totally unrelated to climate conditions, currents break up sea ice and they drift apart, together, forming open leads, pressure ridges, etc.

    If leads didn’t open in winter over much of the arctic sea in winter, seals wouldn’t be finding them to breath and polar bears wouldn’t be finding them in order to hunt the seals using them!

  • Philippe Chantreau // September 11, 2009 at 5:05 pm | Reply

    ” the raw temps can be done against the same reference point.”

    Ok but isn’t anomaly computed to the monthly average? (I think that’s the way it is at least in GISS). That would make for a changing “reference point” no?

    Anyway, just details, I get your point.

    But I am NOT trying to do it myself. I have come to realize lately that climate blogging for a layman like me is a rather pointless exercise.

    It occurred to me that the time spent arguing with bozos saying that carbonic snow is “broadly feasible” on Earth is sadly wasted. I’d rather practice trumpet, swim, ride my bike or look for cool new music on you tube. Just better for my own sanity.

    And for all in search of a nice, refreshing, totally not climate related recreation, I find this guy quite amazing:

    [edit: please no embedded video in comments.]

  • Philippe Chantreau // September 11, 2009 at 5:06 pm | Reply

    Not implying you’re one of the bozos, Mark, just that I have cut drastically on blogging in general.

  • luminous beauty // September 11, 2009 at 5:17 pm | Reply

    From what I can tell by mere appearances , Joe Bastardi believes 10 years = 91 months.

  • Kelly // September 11, 2009 at 9:54 pm | Reply

    Here’s a video by Joe Bastardi discussing solar cycles and climate. He also discussed polar ice conditions by showing a US submarine at the NP in March, 1957 (5 minute point in video)


    [Response: I see Bastardi has joined the Dalton gang. How foolish to think that climate scientists haven't thoroughly investigated the influence of the sun on climate.]

  • dhogaza // September 11, 2009 at 11:08 pm | Reply

    How foolish to think that climate scientists haven’t thoroughly investigated the influence of the sun on climate.

    Remember the operative phrase is “AGW is a fraud”. So it doesn’t matter if climate scientists know the sun impacts climate, they lie about it in order to push their political agenda, which is to enslave the world under a worldwide socialist government.

    Wow, in black-and-white, that sounds really crazy, doesn’t it?

    Right up there with “death panels pulling the plug on grandma”, which I’d guess Watts and Bastardi believe to be true, too.

  • Ray Ladbury // September 12, 2009 at 1:42 am | Reply

    I suppose that believing the first impossible thing is the hard part. Once you’ve convinced yourself of that, I suppost you can convince yourself of all sorts of absurdities–like the proposition that uneducated twits will stumble upon the truth the professionals miss, or that the entire scientific community is conspiring to defraud the entire world.

    “Those who can make you believe absurdities can make you commit attrocities.”–Voltaire

  • Philippe Chantreau // September 12, 2009 at 4:19 am | Reply

    Sorry Tamino, didn’t know how to prevent the thing from automatically embedding and just leave a link. The dude’s name is Edmar Castaneda, I highly recommend.

  • Lucy Skywalker // September 12, 2009 at 5:19 pm | Reply

    Tamino, I’m honoured that you take notice of me. However, it would have been nice if you’d invited me to respond earlier. Because since you did not, I was naturally a bit peeved and did a whole page on our website – just for the record that you have misrepresented my position and I can respond. However, if you think I’m misrepresenting anything in my response to you, I’d be glad for your response. I try to get things right in the end – engineering quality often starts out flawed and becomes workable by checking out and correcting, in my experience.

    [Response: Exactly where is that page?]

  • george // September 12, 2009 at 8:37 pm | Reply

    “It took me a solid month of study to be sure which side was correct [on AGW]” — from the My Story page on the Green World Trust site linked to by Lucy Skywalker above.

    Makes the efforts of Anthony Watts look amateurish in comparison, for sure.

  • Lucy Skywalker // September 13, 2009 at 7:02 am | Reply

    Tamino, thanks for response. My responses page was indicated at the bottom of my original page, as people’s comments started to come in – still work in progress. Then I realized I needed a whole page to respond adequately to you. Page is here. I’m open to corrections re how I’ve represented you, please let me know.



  • Kelly // September 13, 2009 at 6:05 pm | Reply

    Some more information on the USS Skate and Joe Bastardi’s claim about open sea in March, 1959.

    1. Comment by Oakden Wolf on Accuweather’s Joe post.

    “From the account of the USS Skate:
    “the Skate found open water both in the summer and following winter. We surfaced near the North Pole in the winter through thin ice less than 2 feet thick. The ice moves from Alaska to Iceland and the wind and tides causes open water as the ice breaks up. The Ice at the polar ice cap is an average of 6-8 feet thick, [not anymore!!] but with the wind and tides the ice will crack and open into large polynyas (areas of open water), these areas will refreeze over with thin ice. We had sonar equipment that would find these open or thin areas to come up through, thus limiting any damage to the submarine. The ice would also close in and cover these areas crushing together making large ice ridges both above and below the water. We came up through a very large opening in 1958 that was 1/2 mile long and 200 yards wide. The wind came up and closed the opening within 2 hours. On both trips we were able to find open water. We were not able to surface through ice thicker than 3 feet.”

    2. Link to USS Skate photos – scroll down to see 9 arctic region photos.

    3. Polynyas (areas of open water) – link to NSIDC discussion of them.

    4. My conclusion: the USS Skate, using sonar, found a polynyas in March, 1959. So Joe Bastardi’s sea ice anecdote doesn’t tell us anything about sea ice trends.

  • Ray Ladbury // September 13, 2009 at 9:06 pm | Reply

    Kelly says: “So Joe Bastardi’s sea ice anecdote doesn’t tell us anything about sea ice trends.”

    On the other hand, it tells us all we ever need to know about Bastardi, but then if you speak Italian, you already knew that.

  • Mark // September 14, 2009 at 3:05 pm | Reply

    “Ok but isn’t anomaly computed to the monthly average?”

    Not necessarily.

    Do you take the monthly average and compare each day with that (that would be the anomaly compared to the monthly average)?

    Or, do you take the monthly average for the last 30 years and compare that month’s average to that 30-year month’s average?

    Or, even, take the average temperature for the last 30 years and compare that month’s average to that 30-year average?

    Or take a period other than “the last 30 years”?

    What if you don’t even have 30 full years?

    What if you do, just, for one dataset, but not for another? Then your anomaly isn’t comparable. Or if you had more than 30 years for one dataset and took a historical average for one dataset and not for another?

    Again, you are missing out what you mean by anomaly and making your data say something that nobody can refute because nobody knows what you did.

    “That would make for a changing “reference point” no?”


    There’s the same anomaly reference point for GISS data. Or the same anomaly reference point for HADCrut.

    And they say what the reference point is (if it isn’t “the freezing point of pure water”).

    “I have come to realize lately that climate blogging for a layman like me is a rather pointless exercise.”

    As a layman, you CAN (and SHOULD) ask for definitions.

    E.g. when someone says “the anomaly here is…” ask “what is the reference point?” GISS does.

    When someone compares two different sources, ask “are the reference points the same”.

    You will find that denialists cherry-picking their way will not answer.

    And from that you can draw your own conclusions about who is the more trustworthy.

  • Philippe Chantreau // September 16, 2009 at 4:19 pm | Reply

    I was thinking about GISS Mark, as I said in my post. There is no need to try to convince me of the dishonest tactics used by denialists, I’ ve seen a wide variety of them and endured torrents of verbal abuse from some on John Cook’s site, especially on stratospheric cooling (a good bunch had to be edited and was plain foul language). I’ve spent more time than I should have here at Tamino’s to sometimes try to unweave the wordy rethoric of our friend Michel (who has gone missing in action?).

    My understanding of GISS anomaly is that an average temp is computed (not sure exactly how) for each month of the year on the 1951-1980 period. A given month average temp will be higher or lower than the 51-80 average for that month and that difference is what is called the anomaly, is that more or less correct?

    If yes, it means that the anomaly is truly a monthly value and that every month of the year has a different “reference” temp against which the anomaly is computed. Hence all the excitement at WUWWT when a month shows a negative anomaly, whether or not that has anything to do with a long term trend (which it obviously doesn’t).

    Furthermore, HadCRU, although they have a similar method, use a different reference period. That’s all I had in mind, really. Now I may be wrong in my understanding, if so I will gladly welcome clarification.

    Make it short and to the point though, this is already way more time than I want to devote to blogging.

    • TrueSceptic // September 16, 2009 at 5:21 pm | Reply


      That is my understanding. Why else would GISS say

      Best estimate for absolute global mean for 1951-1980 is 14C = 57.2F,
      so add that to the temperature change if you want to use an absolute scale
      (this note applies to global annual means only, J-D and D-N !)

      This came up at Deltoid re a different dataset using a different baseline (NOAA IIRC) but the same principle applies.

      (I was disappointed not to get a reply from Barton Paul Levinson.)

  • Barton Paul Levenson // September 18, 2009 at 9:46 pm | Reply


    I stopped going to Deltoid after they ran an anti-Christian editorial.

    • TrueSceptic // September 18, 2009 at 11:43 pm | Reply


      It was 5 days before you left Deltoid (a great loss IMO).

      Here is that post.

      617 BPL,
      That’s what I thought when I first started looking at this a few years ago but you need to explain the following:-
      Global Mean Monthly Surface Temperature Estimates for the Base Period 1901 to 2000
      J F M A M J J A S O N D Annual
      12.0 12.1 12.7 13.7 14.8 15.5 15.8 15.6 15.0 14.0 12.9 12.2 13.9
      Jul – Jan = 15.8 – 12.0 = 3.8
      (from here, where it says,
      The global monthly surface temperature averages in the table below can be added to a given month’s anomaly (departure from the 1901 to 2000 base period average) to obtain an absolute estimate of surface temperature for that month.
      It is clear that you must add the anomaly for each month to the base temperature for each month.
      2 examples of monthly anomalies from here
      1988 1 0.4590 1988 2 0.2942 1988 3 0.4029 1988 4 0.3672 1988 5 0.3046 1988 6 0.3239 1988 7 0.2847 1988 8 0.2479 1988 9 0.2499 1988 10 0.2321 1988 11 0.1616 1988 12 0.2789
      Jul – Jan = 0.2847 – 0.4590 = -0.1743 (negative!)
      1998 1 0.5657 1998 2 0.8288 1998 3 0.6059 1998 4 0.7107 1998 5 0.6309 1998 6 0.6404 1998 7 0.6980 1998 8 0.6697 1998 9 0.5007 1998 10 0.4393 1998 11 0.3604 1998 12 0.5124
      Jul – Jan = 0.6980 – 0.5657 = 0.1323
      If the monthly anomalies use the same base temperature for all months, how can the differences between, say, Jan and Jul anomalies be so small, when the absolute temperatures differ by several degrees?
      Lastly, why does GISS say
      Best estimate for absolute global mean for 1951-1980 is 14C = 57.2F, so add that to the temperature change if you want to use an absolute scale (this note applies to global annual means only, J-D and D-N !)
      That is, you cannot use this single figure for monthly anomalies as each month has its own base temperature.
      In short, Courtney was right: absolute monthly average temperatures vary by several degrees and monthly anomalies in themselves do not, and cannot, show this.
      I’d like to hear from Eli too, please. :D

      I hope the post isn’t mangled too much.

      Eli did respond, BTW.

  • Barton Paul Levenson // September 19, 2009 at 3:48 pm | Reply


    I think you’re right about that and I was wrong. You can tell the folks on Deltoid if you like.

    • TrueSceptic // September 19, 2009 at 4:19 pm | Reply

      Thanks. I hope you understand that it gave me no pleasure to support Courtney (in this one instance) but I had no option if I believed he was right.

      That thread has gone so far into loony-land that I doubt anyone remembers that we discussed technicalities such as the meaning of monthly anomalies. :)

  • Halldór Björnsson // October 2, 2009 at 12:32 am | Reply

    Regarding Svalbard,
    A quality controlled homogenised data series can be found in the Nordklim data set, which is related to the NARP set referenced above.

    The data in Svalbard goes back to early 20th century.

Leave a Comment