the Air Vent

Because the world needs another opinion

Two Civil Conversations on Other Blogs

Posted by Jeff Id on October 1, 2010

I’ve been enjoying a couple of conversations away from tAV.  One where I’m discussing with James Annon the McKitrick McIntyre and Herman paper which shows models far exceeding trends.  His blog is moderated so the pace is very slow.  The link is here. James has made (and published) the claim that model means produce a non-physical result and is therefore not worth examining closely.  I replied that while model mean may be non-physical the magnitude of the difference is the issue.   It has been a fun conversation to date.

The second conversation is more entertaining and has been over at Bart Verheggens open thread.  We’ve been discussing the usual hockey sticks with unusual civility.  Scott Mandia asked for the simplest explanation of why some critique the hockey sticks – so I took a crack at it below.

Temperature proxies are millenia long series of suspected but unknown temperature sensitivity among other things (noise). Other things include moisture, soil condition, CO2, weather pattern changes, disease and unexpected local unpredictable events. Proxies are things like tree growth rates, sediment rates/types, boreholes, isotope measures etc.

In an attempt to detect a temperature signal in noisy proxy data, today’s climatologists use math to choose data which most closely match the recently thermometer measured temperature (calibration range) 1850-present.

The series are scaled and/or eliminated according to their best match to measured temperature which has an upslope. The result of this sorting is a preferential selection of noise in the calibration range that matches the upslope, whereas the pre-calibration time has both temperature and unsorted noise. Unsorted noise naturally cancels randomly (the flat handle), sorted noise (the blade) is additive and will average to the signal sorted for.

The blade of the hockey stick is the sorted signal and noise in post 1850 times, the handle has a flattened shape due to the canceling of the noise. The difference between the blade and handle is referred to as variance loss.

Anyway check them out if you’re interested.

13 Responses to “Two Civil Conversations on Other Blogs”

  1. Kenneth Fritsch said

    Jeff ID, I am interested in the subject matter, but those threads are a big waste of time for me to read in attempting to sort out any useful tidbits of information.

    In these threads there is the prevailing tone of the black versus white and that what I see proves/disproves something very convincing about the past temperatures and temperature changes and the edgy personal references. I see over arching statements about the similarities in the reconstructions when, in fact, given that the reconstructions lose variance amplitudes, the striking feature is how much they do not agree. Even the evolving Mann et al reconstructions over the years have changed significantly and/or the CIs are actually ceiling to floor.

  2. kim said

    I’ll do my usual and comment without educating myself by reading your conversations, other than lightly skimming Bart’s yesterday. It amazes me that trained climatologists have these questions. The magnitude of the discrepancy between the models and reality is clearly the problem, and why does Scott not understand the mess that the hockey stick is in. Even Gavin’s beaten the retreat on that one.

    And last month I found an erudite commenter at lucia’s who’d never heard of the Livingston and Penn findings.

    Just as locked in as the producers of 10:10’s ‘No Pressure’.

  3. Stilgar said

    Which means that any proxy which was affected by anything other than temperature during 1850-present that could have changed the rate of formation of the proxy into a trend similar to that of temperature change will used as a temperature proxy.

    Sediment layers changed by dams/diches/bridges, temperature proxy.
    Trees healing after they have been damaged, temperature proxy.

    My favorite is that you can have 2 series of the same type of proxy (2 ice cores series or 2 pine tree series) and one is eliminated and the other is not. It doesn’t matter if they are a relative “stone’s throw” away from each other. You don’t have to know why one is considered a thermometer and the other isn’t even though they were both in places with very similar characteristics. Nope the statistical meat grinder says one is a temp proxy and that is enough for climatologists.

    Hell, some climatologists don’t even care if other scientists have proven for a fact that a certain proxy is in no way a temperature proxy. If the trends match then they will be used in some way.

  4. Eric Anderson said

    Exactly, Stilgar.

    Unless there is independent, controlled verification that a particular proxy functions as a temperature proxy, regardless of other perturbing events, then there is *no* reason we should believe that the particular proxy is representing temperature when we plug it into our calculations. It might represent temperature, but it might just as well not.

    How this simple concept can be lost on some of the “experts” is astonishing.

  5. John Norris said

    The average jet analogy is irrelevant and misdirecting from the point. If one were to average jets one would use measurements, not photographs. You could then produce an average jet size and shape that would be clear. The result may not represent an aircraft that would fly well, but it would be clear. Of course climate science uses measurements from thermometers, proxies, and models. A combination of model results may also not fly well, but the results can be clear. It can also be meaningful.

    If you say you have a big jet, unless the average jet size is somewhat established, your claim is just a claim. If you say the models are simulating reality, unless you have an objective comparison to some reality, your claim is just a claim.

  6. Steve Fitzpatrick said

    The thing about most proxy reconstructions that is simply bizarre is that there is usually no reasoned basis for the “calibration”. I mean, if trees rings near the limit of their temperature range are good indications of average temperature, then you damned well ought to have controlled data (like growth rates under controlled conditions) to back up this supposition. Finding a smallish number of proxies out of a much larger pool which happen to fit the instrument temperature data, and discarding all those that do not fit the instrument data, is just unbelievably bad methodology; if you look hard enough for correlation… you will find it, even if it is spurious. I can’t understand why such ridiculous methods would ever be funded.

  7. Graeme said

    The models are a theoretical construct – not a test.

    Hence Theory + Models does not prove anything.

    The gap between the models and observations is critical for the predictive value of the theory.

    Three possibilities. (1) The theory and the models are wrong, or (2) the models are wrong in terms if incorrectly repesenting the theory, oir (3) the observations are wrong.

    My money is on (1) or (2), with (1) being most likely

  8. steveta_uk said

    Can anyone tell me whether a reconstruction using tree-ring data has ever been performed/published on unselected data?

    It’s rather obvious (I would have thought) that selecting samples by their match to recent temperature trends, which as Jeff says could be simply selecting matching noise, then back-projecting will produce a hockey-stick, due to the historical noise cancelling itself out.

    So are there any plots available of the unselected tree-ring data back projected? Is this simply a flat line, and hence has always been ignored, or are there any long-term trends visible?

  9. KevinUK said

    I’m sure most readers here are already aware of this but I’ll post it anyway.

    Just in case anyone is wondering about the significance of ‘variance loss’ as referred to in Jeff ID’s quote above ‘The difference between the blade and handle is referred to as variance loss.’, remember that we are talking about the claim by Mann et al that the late 20th century warming period has been ‘unprecedented in the last 1000 years’. Well if you suppress the variance in the pre-20th century data then of course it’s going to look like the late 20th century warming is ‘unprecedented’ because you’ve damped down if not removed any variance that would have been present otherwuse that would most likely have included somewhat inconveniently (for the CAGW ‘warmista’)the Medieval Warm Period (MWP) and the Little Ice Age (LIA). If you start to put back in the ‘rejected’ proxies (and preferably don’t use tree ring proxies at all), rejected because they weren’t selected by your algorithm that is looking for ‘unprecented late 20th century warming’ and if found selects and weights that proxy above all others e.g. de-centred PCA) then surprise, surprise the inconvenient MWP and LIA appear.

  10. Rob R said

    I wonder when some enterprising blogger will notice what Roger Pielke Snr has been saying about oceanic heat and conclude that for climate the temperatures which really matter are those of the sea-surface.

    What is needed is a comprehensive study of the last 2000 years of sea surface temperature. There are some reasonably direct geochemical proxies for water temperature including the Mg/Ca ratio in calcite from marine organisms. The organisms include but are not limited to coral and foraminifera.

    A collection high-resolution of SST records would likely be a better diagnostic of the state of global climate than the likes of the Mann et al tree-ring fairy stories. After all, water temperature is a much better metric of heat content than air temperature.

    One of the problems with sea surface temperature records is that this work typically focusses on changes over many thousands to 10’s or 100’s of thousands of years. But there are an increasing number of peer-reviewed papers that cover the more recent past with resolution down to just a few years. In fact some are now examining water temperature to a seasonal level. I suspect this would be a productive topic for a bunch of “citizen scientists” to begin investigating.

    I also suspect that the SST records are being compiled by scientists who are less heavily invested in the whole alarmist IPCC bandwagon than are the tree-ring clique. So there is a good prospect of finding a substantial number of quality temperature series that collectively diagnose changes in global climate.

  11. ianl8888 said


    I’ve read through the Bart thread, ostensibly on Loehle’s view that Lindvquist’s paper supported him

    Essential players were you, MapleLeaf, Mandias and Mosher towards the end. I don’t really regard MapleLeaf’s attitude as “civil”, rather as barely restrained contempt, but at least the discussion did continue for a bit

    Although it finished in a muted stalemate, MapleLeaf did make some minor concessions to variance across the palaeo record, while you and Mosher came to a lukewarm status

    The point of this post – I’m at the lukewarm status for the reason that the basic laboratory physics attributes a range between 0.8-1.2C to 2x increase in atmospheric CO2 (mostly I use Lindzen’s various papers for this, since he is unlikely to promote exaggeration – but he regards this increase as trivial)

    So where do you stand on the vexed question of feedbacks, +/- ? This seems to me the absolute crux, yet the Bart thread did not mention this

  12. Jeff Id said

    I’m not done with the Loehle thing yet, just haven’t had time. Regarding feedbacks, I really don’t know. Sorry if that is boring. I’m really not a lukewarmer either b/c negative feedbacks may flatten the CO2 to an undetectable level. I don’t think Mapleleaf is an honest broker and he never did admit his failure to understand until claiming he had made concessions. He just attacks with generalities and moves on.

  13. ianl8888 said


    Not boring – not surprising either

    Mapleleaf did not “fail” to understand. he understood very well, but insisted with his point until the “high MWP = high climate sensitivity” connection was placed in the open

    Although I didn’t like his undercurrent of contempt, the thread did expose some interesting discussion

Leave a Reply

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

You are commenting using your 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: