the Air Vent

Because the world needs another opinion

Area Weighted TPCA Check

Posted by Jeff Id on July 7, 2009

I put these images together from Ryan’s latest post and my own closest station post to show how truncated PCA is doing a good job with locating station information at the correct grid points. TPCA has no method for knowing exactly where each station is located. Ryan added eigenvector weights to each surface station during imputation which improves the odds that during convergence the station information is located at the correct grid area. We think of this like a sanity check to demonstrate that the method is working. My own opinion is that it’s doing an excellent job now and the Antarctic reconstruction by expectation maximization is repaired thanks to Ryan’s huge efforts.

fig_6[1]

Figure 1 Ryan Recon 28 pc

id-recon-spatial-trend-by-distance-weight-1956-2006[1]

Figure 2 Jeff Recon, Closest Station

fig_7[1]

Figure 3 Ryan Recon 28 PC 1967 -

id-recon-spatial-trend-by-distance-weight-1967-2006[1]

Figure 4 Id recon closest station 1967 -

The next graphs are the trends, even the shape of the linear trends is strikingly similar between TPCA and an area weighted calculation. Which is more correct? It doesn’t matter at this point because the differences are so small. I didn’t bother to rescale the graphs to match each other yet and before you say it, I know Using is spelled incorrectly in the last two plots.

Ryan Recon 28 pc

Ryan Recon 28 pc

id-recon-total-trend-by-distance[1]

Jeff Recon Closest Station 1957 - 2007

id-recon-trend-closest-station-1967-2007[1]

Jeff Recon closest station 1967 - Dec 2006


38 Responses to “Area Weighted TPCA Check”

  1. NukemHill said

    I didn’t bother to rescale the graphs to match each other yet and before you say it, I know Using is spelled incorrectly in the last two plots.

    Heh.

  2. Bob H. said

    Now that there is a much more accurate analysis, it is time to ask the $64,000 question…Why is the peninsula warming so much more than everywhere else on the continent?

  3. Jim said

    @Bob H. said July 7, 2009 at 1:09 pm:
    My guess is the peninsula is heated by the surrounding water. Just a guess, though.

  4. Because it sticks out into the ocean and a warmer current laps its shores?
    Note that it’s really only its tip that is exhibiting anything and that’s under half a degree per decade.

  5. page48 said

    RE: #4, “….Note that it’s really only its tip that is exhibiting anything and that’s under half a degree per decade.”

    Intentionally dramatic color choices (I realize that the colors used here were adapted because they are the same as the ones used in the Nature paper).

    I noticed some time ago that climate scientists have chosen colors to make relatively small differences in temp appear visually to be quite large. It’s the same principle as using bizarre scaling in graphs to grossly emphasize a small rise in temp over time.

    Even in ordinary weather broadcasting relatively cool temps are warmer colored, now. I can’t prove it, but I swear the colors were changed at some point in the last 20 years to make the visuals more dramatic.

    These are silly points, I suppose, but the marketing aspects of AGW have been really well thought out.

  6. Jim said

    Adam Gallon:

    Apparently the Antarctic Circumpolar Current bangs against the peninsula and is relatively warm.

    http://www.antarcticconnection.com/antarctic/news/2008/040308warmer.shtml

  7. Ryan O said

    Sometimes simple is just SOOOOOO good. 😉

  8. Jim said

    If you look at the blue-line delineation of the current, you can see how where it is close, the region is hotter. Not a perfect match, but pretty good.

  9. Bob H. said

    I can go with the Antarctic Circumpolar Current (ACC) as an explanation. Can anyone tell me why the ACC has been warming? If the current increased its temperature at the same rate, there wouldn’t be any “extra” warming…True?

  10. curious said

    Bob – maybe a dim question but can anyone tell me why it shouldn’t (ACC warm that is)? Especially at these type of rates of change? What is the model that says it should remain invariant?

  11. Ryan O said

    The problem isn’t so much that the circumpolar westerlies are warming as it is the strength has increased. I don’t know why this is, and at the moment, I am not particularly interested in finding out on my own. If any of you know a handy reference, though, it would be interesting to read about.

  12. Antonio San said

    Ryan O, I suggest you get familiar with the work of the late Professor Marcel Leroux. You can first read the seminal paper “The Mobile Polar High: a new concept explaining present mechanisms of meridional air-mass and energy exchanges and global propagation of palaeoclimatic changes ” available at science direct:

    http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6VF0-48DYTM7-W&_user=10&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=766756d3d7f4ca161b38dbeda14162d9

    I would also suggest you read his 2005 book “Global Warming: Myth or Reality, the erring ways of climatology” Springer Praxis available on amazon.

    I also suggest you get your hand on “Dynamic analysis of weather and climate” Wiley 1998 in a library as the book is now unavailable (a second edition, updated to 2008 is being prepared for 2010).

    You would find that there are very clear answers to the questions you asked in post #11. Moreover, once you’ve read Leroux, you get a general view of atmospheric circulation and how it really works (not the tri-cellular stuff). That knowledge in turn gives you a referential on which all climate news and papers can be assessed as his demonstration is as rigorous and cartesian as they can come, based on facts and observations. In my opinion, anyone who has not read this work is like a geologist who would not know about plate tectonics.

  13. Antonio San said

    ooops sorry for the link length.

  14. Ryan O said

    #12, 13 Thanks for the links. 😉

  15. Fluffy Clouds (Tim L) said

    as i see it there is NO real change in temps for 40-50 years

  16. Jim said

    I looked for information on the polar currents. I saw something that I can’t find again that said the waters feeding some of the polar currents were depleted in C14 due to the time taken for the water to make its way to the South Pole. It said live animals (seals and krill) that were carbon dated there would measure hundreds years older than they really were. I don’t have access to scientific journals or I would do more research on this. But if true, the heat content of the currents could be old also meaning any heating has nothing to do with current conditions or to some extent past CO2 build up. It is interesting that Ryan’s map matched the current path better than Steigs. I am wondering if that would be a good reality check on the Antarctic trends. It could get complicated because you would then have to know the history of the current temperature.

  17. Hal said

    I just learned something new. An explanation why the total sea ice area is meaningless, but that only the “rapid loss” in the northern hemisphere matters. Interesting reasoning.

    FROM
    Warnung aus Kopenhagen
    von Stefan Rahmstorf, 19. Juni 2009, 15:00

    A question along that line was asked on Dr. Stefan Rahmstorf (the teutonic voice on RC, of special smoothing fame exposed on CA)) own german blog (Klima lounge).

    http://www.wissenslogs.de/wblogs/blog/klimalounge/debatte/2009-06-19/warnung-aus-kopenhagen/page/4#comments

    Question:
    I would like to respectfully ask you, to interpret the following representation. It relates to the sea ice extent globally and Southern Hemisphere (SH) and the Northern Hemisphere(NH) views. We recognize no 30 year global trend, a decrease in the NH and an increase in ice surface around the Antarctic. In sum, nothing has changed except that we see a greater 2008 Global sea ice area than any time in the last 30 years. How does this fit in to the “accelerated” climate change, which is advancing even faster than thought?

    [Answer: To form the sum of the Arctic and Antarctica is not very useful, because (1) these are always dominated by the winter hemisphere, and (2) at both poles are affected by very different geographical conditions and other mechanisms. (At the North Pole ocean is surrounded by land, at the South Pole vice versa.) The strong disappearance of summer ice in the Arctic with the record minima of 2007 and 2008, to which the synthesis report relates to is caused by the strong warming in the Arctic region , while the ice expansion in Antarctica’s winter reacts heavily to the wind field which (not enclosed by land) blows the ice apart. Stefan Rahmstorf]

    I used Google translator and then corrected the gibberish part myself.

    Hal

    Harald (german version of my name)

  18. Jim said

    Hal said July 8, 2009 at 11:33 am:

    And yet NSIDC say winds do matter in the Arctic. Go figure …

    http://nsidc.org/arcticseaicenews/

  19. Jeff Id said

    #17, I love how they spin everything. Even the Cryosphere home page keeps global sea ice up prominently. The reason is simple. Computer models predict more warming at the poles so global sea ice is expected to shrink. So far – no melting and no dice. I calculated that global sea ice hit a near record high a couple of months ago. Pretending that the poles are on another planet from each other is simply convenient hand waiving.

  20. Jim said

    Found a database that gives the age of ocean water. Pretty cool. Looks like our ACC is 1200 years old.

    http://intcal.qub.ac.uk/marine/index.html?ulat=-47.989921667414165&rlon=-45.52734375&llat=-69.162557908105&llon=-77.16796875&clat=-60.239811169998916&clon=-61.34765625&zoom=4&Columns%5B%5D=Notes&Columns%5B%5D=ReservoirAge&Columns%5B%5D=ReservoirErr&Columns%5B%5D=C14age

  21. Jim said

    It would be nice to run this by someone who understands the deep ocean currents. If the ACC really took 1200 years to get to the Antarctic, then the heat it contains will also be from 1200 years ago.

  22. Nathan said

    Jeff
    “Even the Cryosphere home page keeps global sea ice up prominently. The reason is simple. Computer models predict more warming at the poles so global sea ice is expected to shrink. So far – no melting and no dice. I calculated that global sea ice hit a near record high a couple of months ago. Pretending that the poles are on another planet from each other is simply convenient hand waiving.”

    The North pole was always expected to be well advanced of the south in terms of warming. If you think otherwise, it is your misunderstanding.
    Remember the south pole is not only in the middle of a continent, but also at altitude (over 2000m), they are very different. You also need to consider the effect of the Ozone hole.

  23. curious said

    Nathan – did you see this post?:

    https://noconsensus.wordpress.com/2009/06/16/historic-variation-in-arctic-ice-tony-b/

  24. Nathan said

    Yes, I saw it, but didn’t read it all.

  25. Jim said

    @Nathan said July 9, 2009 at 2:43 am
    Hi Nathan,

    Won’t the sea in the Arctic have an influence on the temperature? It seems the heat energy from the sea would be superposed on any heat from the Sun.

    Also, the wind has a major effect on the ice there. If you are looking for air warming, it makes more sense just to look at the temperature instead of the sea ice which is influences by a number of factor that interact in a complex manner.

  26. Jeff Id said

    #22 Please stop assigning false beliefs to me and shoot them down. This is a straw man.

    Look hard Nathan, eventually you will find a real mistake. I honestly am just an engineer and am truly pissed off at the advocacy that’s so rife through climate science right now so there probably will be something. I tend to admit my mistakes and ask the same from guest posters.

    See this one for an example:
    https://noconsensus.wordpress.com/2009/03/29/mullagain/

  27. Jeff Id said

    #25, Jim,

    My own humble opinion is that water and air flow are the true reasons for any melting of ice. Aeronautical engineering is in large part a study of fluids. From memory, the equations for laminar flow heat transfer are quadratic with velocity because the boundary layer between water or air and ice becomes thinner as velocity increases.

    I will be shocked if in 30 years climatology is still attributing ice melt in our present time with 1 or 2 C of change or input from solar or cloud. That’s why climatologists still say current change. Where I don’t agree is when some are then attributing the current change to global warming with little evidence to back up the reasoning for it. – Ocean currents are poorly modeled. I’ve been wrong before though.

  28. Jim said

    #27Jeff,

    Looking at the Arctic UAH data, it appears the actual temperature has stabilized over the last several years. Maybe that’s why the warmingistas play the sea ice game. Even that could blow up in their faces.

  29. Bob H said

    Ryan,

    Thank you. This is the point I was trying to make. I’ve been on the road the last two days and out of internet range listening to the ICCC conference in March. One of the presenters was discussing the correlation of the PDO on various seemingly unrelated observations. Do you suppose the PDO could be related? I haven’t had a chance to follow Antonio’s link yet, so I don’t know whether the above question has already been answered.

  30. Ryan O said

    #29 I do not know. I wish I did, but I honestly don’t have any idea.

  31. Nathan said

    Jeff I am quoting your words. What false belief? Where is the strawman?

  32. Jeff Id said

    Nathan,

    Here.

    “well advanced of the south in terms of warming”

    Why are we discussing this? Did anything I said disagree with that. WTF?

    Actually, I wouldn’t mind pointing out the cherry picking and occasionally contradictory argument’s I’ve read about sea ice….. but why? There’s no point and I prefer to look at data. Your straw man is a waste of my limited mind space. Face it, the models are all over the place AND I never implied that your straw man wasn’t true.

    You need to focus if you want to take me down. Honestly, don’t worry you’ll find something. In the meantime I don’t care to argue this wildly distracting point.

    I just realized straw men are just homeless ad homs. – funny stuff.

  33. Nathan said

    You said this Jeff
    “Computer models predict more warming at the poles so global sea ice is expected to shrink. So far – no melting and no dice. I calculated that global sea ice hit a near record high a couple of months ago. Pretending that the poles are on another planet from each other is simply convenient hand waiving.”

    What does the last sentence mean?

    “I just realized straw men are just homeless ad homs. – funny stuff.”
    No they’re not. An ad hom is an attack on a person, rather than on what they discuss. A straw man is where you make up an argument that someone didn’t make.

    REPLY: Ad homs are attacks on a person with the intent of distracting an argument. A straw man is a made up argument with the intent of distracting an argument. Sorry you don’t like my analogy but I’m not surprised people with the intent to troll often miss the entertainment.

    The last sentence means that claiming you can’t add sea ice together at the poles is an argument of ‘convenience’ IMO rather than reality.

  34. Nathan said

    Jeff

    “Face it, the models are all over the place AND I never implied that your straw man wasn’t true.”
    what does this mean? It makes no sense.

    “I never implied that your strawn man wasn’t true”

    Does anyone know what this means?

    It means that you have implied that somehow I’ve argued both poles are equal or the south pole should melt greater than the north, you have created a straw man argument. In a troll like fashion you apparently cannot see it, this is what I meant about honesty at tAV. Do you admit that I have never implied the poles are equal or one or the other is of greater sensitivity?

  35. Nathan said

    “You need to focus if you want to take me down. Honestly, don’t worry you’ll find something.”

    That is paranoia.

    Ok, I’m out of here.

    Have a very pleasant day and good luck in your investagations.

  36. curious said

    Nathan – if you can give a succinct explanation and some supporting evidence of the “polar amplification” theory often talked about I’d like to hear it. This has been one of the primary motivators in the AGW debate as major ice cap loss would increase sea levels and that would be a major challenge for many settlements around the globe. However to date I am not at all convinced and I have tried to find it. Admittedly I’m just an armchair participant in all this so I could have missed some key element and I’d be pleased to have it flagged up.

    On this blog there are two (or three or more – I still haven’t had opportunity to follow up on the interesting works highlighted by Antonio San) items which really challenge your statement in 22:

    “The North pole was always expected to be well advanced of the south in terms of warming. If you think otherwise, it is your misunderstanding.”

    These items are the entry I highlighted to you by Tony B and Ryan and Jeff’s Antarctic recon. What is this “always expected” you refer to? Since when is “always”?

    In 10 I was asking where is the hypothesis which says that all should be (or is) invariant in the global climate? Seriously, can you summarise what this model might be and how it would hang together? Would it allow/imply that at any given point on the globe every July 4th or Dec 25th or Jan 1st will have identical weather? We are seeing some change in climate patterns but IMO we are still along way from knowing what this is or what it means. I tend to agree with the view it’s partly because we are measuring, comparing and communicating data in new more powerful ways that we are seeing problems where perhaps none exist. This is IMO similar to the arguments you are picking with our host – turning an off the cuff turn of phrase into a manifestation of “paranoia” is exactly what Jeff is highlighting: just another form of ad hom.

    Like I say I’m just sat in the backrow following along, but to be frank you are spoiling the show. There is a quality picture on and you are chucking popcorn everywhere and chewing loudly trying to kick off fights – please have a go at putting something worthwhile in. Thx

    Ok – advert break over – sorry all: back to the main event 🙂

  37. stumpy said

    dont forget the antarctic penninsula is also volcanic, you can see active volcanoes on google earth on the peninsula etc…

  38. Bob H said

    I was thinking about this a bit more and the Antarctic Circoumpolar Current (ACC) very well could be related to the PDO. Here’s my reasoning…

    Stieg’s paper runs from 1957 to 2006. However, the PDO was in it’s cool phase at the time the study begins (1957). In 1977 the PDO switched to a warm phase and remains there until about 2007. Stieg’s study happened to cover the cold to warm phase switch but not the warm to pahse shift. One should expect to see warming, even if minimal across the entire antarctic over this time span, and possibly greater warming in the peninsula.

    As Stumpy said in #37, the peninsula is actively volcanic as well. Depending on how much underwater volcanic activity there is, the ACC could be warmed considerably, which could make the peninsula a much warmer place, relatively speaking.

    So to review…the hypothesis is “The ACC is related to the PDO and Stieg’s paper covers less than a single cycle of the PDO. Volcaninc activity additionally raised the measured temperatures on the peninsula.”

    Proposed methodology to test: Rerun the analysis from 1957 to 1977, and then from 1977 until 2006. The best guess is that there would be little to no warming over the entire continent, except for the peninsula, from 1957 to 1976 or 1977. From 1977 until 2006, one could expect to see some warming as the continent would gradually warm from the PDO switch, again except for the peninsula. Actually, a shorter time span might be more revealing. Run analysis from 1957 to 1967, 1967 to 1977, 1977 to 1987, 1987 to 1997, 1997 to 2006. I suspect that the Antarctic, except for the peninsula, would show cooling for the first two time blocks, a rapid warming in the third block, and then reduced warming in the last two blocks.

    If the peninsula is analyzed separately, whatever the difference, it should be related to other sources, most likely volcanic activity.

    Your thoughts???

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

 
%d bloggers like this: