Click images for more details



Recent comments
Recent posts
Currently discussing

A few sites I've stumbled across recently....

Powered by Squarespace
« Parched earth policy | Main | In the news today - Josh 370 »

Gav loses it - Josh 371

When Steve MicIntyre writes "In the past few weeks, I’ve been re-examining the long-standing dispute over the discrepancy between models and observations in the tropical troposphere." you might think you were in for a bit of a technical post - which, of course, it is - but it is also also very funny and well worth reading. It also inspired the cartoon below.

Click image to enlarge

H/t commenter 'See owe to Rich' for the 'hide the gap' phrase.

Cartoons by Josh

PrintView Printer Friendly Version

Reader Comments (155)

The point of Christy's graph is to compare the rate of change of temperature over time.

No, the point was to cook up a graph to put in front of a Congressional committee to make the models look as bad as possible. There were no trend lines shown.

Hence the choice of height - mid troposphere - and region - tropics. Hence the lack of uncertainty ranges. Hence the smearing together of the observations. Hence the lack of labelling. Hence the irreproducibility (which models, which scenarios?).

All curiously endorsed by the Auditor. I wouldn't trust him to audit my shopping list.

Apr 24, 2016 at 2:44 PM | Unregistered CommenterPhil Clarke

Phil Clarke, yet you would trust the Hockey Team's science and models with your shopping list? What have they ever done but produce reports and graphs to present before Governments around the world with a proven track record of being wrong? It is the only area of their science with any consistency at all.

What are the bits you find so compelling, that 97% of the population find so repellant?

Apr 24, 2016 at 3:32 PM | Unregistered Commentergolf charlie

Schmidt proudly pronounces on his Twatter feed that he is an "occasional juggler".

Only "occasional"? Hmmmm

Apr 24, 2016 at 3:37 PM | Unregistered CommenterJeremy Poynton

If you use a more representative baseline, 1985-2005, include the 95% uncertainty range in the models, plot for the surface rather than the troposphere, include the whole globe and label the graph properly, it looks something like this.

Christy done right.

But doing it scientifically has rather less propaganda value.


(Last updated February since when observations are 0.2C or so warmer)

Apr 24, 2016 at 3:52 PM | Unregistered CommenterPhil Clarke

There were no trend lines shown.

ROFLMAO. You don't understand. What was shown is the divergence of observations and model against time. Which is still looking at temperature rate (same UNITS as trendlines), but a better way of doing it. McIntyre has direct comparisons of trends if you want (unsurprisingly, given they are just a different way of expressing the same data, they don't match either).

It is hilarious that you still don't understand how Christy works out his baseline. The baseline is based on the *trend* - not on a single year. His method is actually quite clever, because it uses the sufficient dataset to baseline without corruption by noise yet allows a comparison of divergence over thirty years. His method is far better than the method used by Mears to do the same thing.

Baselines which are closer than 30 years to the data being checked are not meaningful if you claim prediction accuracy improves at the 30 year timescale, which Christy's plot clearly shows is not the case. The other plots you choose are deliberately chosen to hide the gap by avoiding 30-year scale, the very scale that climate scientists have been demanding be used for as long as I can remember.

As for the choice of the tropics, it doesn't matter. The models need to be right everywhere, and the tropics are a key fingerprint of GHG warming according to models. The absence of that fingerprint falsifies the models. Finding a different plot that matches doesn't rescue the models; they're still wrong. It's kinda how science works. Shame you don't get that - but then you don't get much, you still think Christy's method is sensitive to variations in single years, despite having it explained to you why it isn't sensitive to such things. If you can't even get that simple thing right, how do you expect to get the more subtle issues right?

Apr 24, 2016 at 4:05 PM | Unregistered CommenterSpence_UK

If you use a more representative baseline, 1985-2005

Christy uses a wider range than this for his baseline - although I'm having trouble figuring exactly which years, McIntyre mentions 1975-2005 (but there are no satellite measurements in 1975, so this doesn't seem right) and Christy implies 1979-2015 in his PDF document; but either way, why would you use a *shorter* dataset than Christy did to define a baseline and declare it as better, if you are concerned about short term fluctuations influencing the graph?

include the 95% uncertainty range in the models

Neither plots show the 95% uncertainty in the models, which is unknown, they both show the spread of model results, which is something different to the uncertainty. I actually prefer Christy's method of showing the spread, as a spaghetti graph (commonly used in climate science, I believe) because it gives you a feel for the skew, kurtosis, mode and median, which you don't get from the 5%/95% values. In particular the skewness of the distribution looks non-zero, and this has a material impact on the interpretation that is made. But your sudden dislike of spaghetti graphs is noted.

plot for the surface rather than the troposphere, include the whole globe

Red herring. If the plots are wrong for the tropical troposphere, the models are broken. Aggregating data from elsewhere and hiding the gap until you get something with some tiny overlap doesn't fix the broken bit, especially when the broken bit is supposed to be a fundamental fingerprint of AGW.

Any other innumerate nonsense to pass our way, Phil?

Apr 24, 2016 at 4:30 PM | Unregistered CommenterSpence_UK

LOL. The only thing that needs rescuing is us, from the naked and nauseating dishonesty of Christy and Spencer. Always glad to amuse, personally I find your contortions in defending the indefensible a reliable source of mirth. I never did get that yes or no answer. You also seem to have failed to spot that baseline was not in the list of issues I gave above.

Red herring. If the plots are wrong for the tropical troposphere, the models are broken.

Nope. It could be that the observations are wrong, as they were for years under Christy's stewardship, or the comparison is inappropriate, for example are the models representing exactly the same quantity as is being measured by TMT? Fairly obvious point, really. Any more skewed nonsense to share? :-)

Apr 24, 2016 at 4:54 PM | Unregistered CommenterPhil Clarke

Gc. Forgot to add that the Climate Science award of which you wrote so lovingly - the Norfolk Turkey Prize - can be supplimented with a CRU bar, shaped like a sloping hockey stick ( commonly known as a decline) but this must never be exposed to the public view.

The Mann cluster is another greatly prized award.

Apr 24, 2016 at 4:57 PM | Unregistered CommenterAlan Kendall

Phil Clarke:

No, the point was to cook up a graph to put in front of a Congressional committee to make the models look as bad as possible. There were no trend lines shown.

Phil, could you mean as bad as possible with possible defined by the actual badness of the models? You didn't say to exaggerate, or misrepresent.

Apr 24, 2016 at 4:59 PM | Unregistered Commenterjferguson


Christy done right Pt2.

The choice of baseline has an impact on the relative positioning of the two series being compared, but apparently this is irrelevant.


Apr 24, 2016 at 5:53 PM | Unregistered CommenterPhil Clarke

personally I find your contortions in defending the indefensible a reliable source of mirth.

Phil, let's look back on this thread. You incorrectly claimed that Christy's method was sensitive to the observation value in 1979, showing you didn't understand what Christy has done. You still think a 20-year derived baseline is right but somehow Christy's 37-year derived baseline is wrong. I'm not the one who doesn't actually understand what Christy did - that would be you. It is an absolutely fundamental aspect of science that you have to understand what Christy did before you are able to criticise it - which is why your criticisms are pointless, as you do not understand what Christy did. Essentially every claim you made is based on ignorance.

I never did get that yes or no answer.

And this is just plain dishonest. As I noted, when you're comparing temperature *rates* (e.g. deg C per unit time) then the baseline doesn't matter, which answers your yes or no question quite adequately. However prediction interval DOES matter, and to properly test models you need to set the baseline at least 30 years from the period under test, or you're not actually testing anything. This is not from my claim - this is the claim from climate scientists. None of the plots you generated address this point at all.

You also seem to have failed to spot that baseline was not in the list of issues I gave above.

Are you completely delusional? As recently as your 3:52 post you said:
If you use a more representative baseline, 1985-2005

Phil, people here can read what you've posted. You've talked endlessly about the baseline, without once understanding what Christy actually did. And with respect to the uncertainties, I've already explained why Christy's spaghetti graph of model runs captures the model spread far better than 5%/95% points - although I don't suppose you understand higher order moments of distributions well enough to realise why. While it would be better to have uncertainties on the obs data as well, the uncertainties on those are trivial in comparison to the model spread and change nothing. The models still fail.

So on your list of claims, two are outright wrong, and one is a trivial point which would improve the graph but not change the result in any way. Well done! Another win for team innumerate.

Apr 24, 2016 at 6:18 PM | Unregistered CommenterSpence_UK

Phil, let's look back on this thread.

Yes, lets. Your argument was that Christy was demonstrating the difference in rates of change, trends in other words. If that were true, why not present, um, trendlines, as, to be fair he has done elsewhere, or better still a histogram such as prepared by Schmidt. He did not do this, indeed his choices such as the averaging of observational data sets, and the removal of any kind of confidence intervals, were all designed to put the models in the worst possible light. You don't need to be an Auditor to spot that, and it had the desired effect - Senator Cruz used a Christygraph to claim that the planet is not warming, and that humans do not impact climate.

You also stated that if there was a model/observations discrepancy, (and I do not for a moment dispute that there is a discrepancy in the Tropical TMT) the models must be wrong. Logical fail, as pointed out in the response from Santer and Mears

Senator Cruz used Exhibit A [The Christygraph] as the underpinning for the following chain of arguments: 1) Satellite TMT data do not show any significant warming over the last 18 years, and are more reliable than temperature measurements at Earth’s surface; 2) The apparent “pause” in tropospheric warming is independently corroborated by weather balloon temperatures; 3) Climate models show pronounced TMT increases over the “pause” period; and 4) The mismatch between modeled and observed tropospheric warming in the early 21st century has only one possible explanation – computer models are a factor of three too sensitive to human-caused changes in greenhouse gases (GHGs). [The Spence Logical Fail] Based on this chain of reasoning, Senator Cruz concluded that satellite data falsify all climate models, that the planet is not warming, and that humans do not impact climate.
This logic is wrong. First, satellites do not provide direct measurements of atmospheric temperature: they are not thermometers in space. The satellite TMT data plotted in Exhibit A were obtained from so-called Microwave Sounding Units (MSUs), which measure the microwave emissions of oxygenmolecules from broad atmospheric layers (2-4).[b] Converting this information to estimates of temperature trends has substantial uncertainties.[c] The major uncertainties arise because the satellite TMT record is based on measurements made by over 10 different satellites, most of which experience orbital decay (5) and orbital drift (6-8) over their lifetimes. These orbital changes affect the measurements of microwave emissions, primarily due to gradual shifts in the time of day at which measurements are made. As the scientific literature clearly documents, the adjustments for such shifts in measurement time are large,[d] and involve many subjective decisions (2-4, 6-8). Further adjustments to the raw data are necessary for drifts in the on-board calibration of the microwave measurements (9, 10), and for the transition between earlier and more sophisticated versions of the MSUs.[e]

[…]The hearing also failed to do justice to the complex issue of how to interpret differences between observed and model-simulated tropospheric warming over the last 18 years. Senator Cruz offered only one possible interpretation of these differences – the existence of large, fundamental errors in model physics (2, 21) [Now dubbed the Spence-Cruz argument] In addition to this possibility, there are at least three other plausible explanations for the warming rate differences shown in Exhibit A: errors in the human (22-25), volcanic (26-30), and solar influences (24, 31) used as input to the model simulations; errors in the observations (discussed above) (2-20); and different sequences of internal climate variability in the simulations and observations (23, 24, 30, 32-36). We refer to these four explanations as “model physics errors”, “model input errors”, “observational errors”, and “different variability sequences”. They are not mutually exclusive. There is hard scientific evidence that all four of these factors are in play (2-20, 22-36).


Apr 24, 2016 at 6:44 PM | Unregistered CommenterPhil Clarke

no warming since 1993:

Apr 24, 2016 at 7:18 PM | Unregistered CommenterIbrahim

Oh good grief, this would be so much easier if you actually understood numbers.

Your argument was that Christy was demonstrating the difference in rates of change, trends in other words. If that were true, why not present, um, trendlines

No. As I have made abundantly clear, the first derivative of temperature - rate of change of temperature - and trends are different things that share the same units. I actually don't like trends, there is no reason to expect climate to change on a linear slope and trends are not a particularly helpful way of representing change over time and scales. I would far rather see the magnitude of change over a given time span, which is exactly what Christy presents. Technically trends are another way, but it is absolutely false to see if you are analysing the change of temperature over time it has to be done with trends. Someone who has experience of crunching numbers would know this.

As for models being wrong, there are numerous papers that show models are wrong; including papers showing models are wrong against surface data sets. I agree that it is possible (perhaps even likely) that observations are also wrong in some ways; of course it is. But we can be clear on one thing. As the time horizon increases, models and observations increasingly diverge. That is, as we move from monthly, to yearly, to decadal, to multi-decadal, model and observations get increasingly bad as we move to the right of that scale.

This shouldn't surprise anyone. The hydrological cycle in models is an absolute joke, bearing no resemblance to reality at all, no pretty much any measure (rainfall, cloud cover, humidity etc, both in change and absolute value). Again, the longer the model runs, the more they diverge.

So your response doesn't address my point at all. Yes, there may be problems with the observations that we don't yet know about. But there are definitely problems with the physics in the models, which are hopelessly inadequate for making predictions or projections at the 30-year scale. So your wishful thinking that observation flaws will rescue models is uninteresting and flawed.

Furthermore, if the observations are wrong, for the models to be right we would have to model a complex system better than we can measure it. ROFLMAO. If we don't even know how to measure it properly, the idea that we could model it better is just laughably ignorant.

Every aspect of complex systems theory tells us as the time horizon increases, our ability to predict (or project) the future decreases. Every fair comparison of climate models agrees with that, as do the graphs shown by Christy (and indeed the other scientists, who use a shorter time horizon and therefore get "better" looking results). This is exactly what would be expected. Your resistance to accepting this simple concept is bizarre.

Apr 24, 2016 at 7:47 PM | Unregistered CommenterSpence_UK

That is, as we move from monthly, to yearly, to decadal, to multi-decadal, model and observations get increasingly bad as we move to the right of that scale.


Apr 24, 2016 at 7:58 PM | Unregistered CommenterPhil Clarke

eesh.... Phil C - why are you dancing on the head of a pin and bothering to wrestle like this to defend the broken models? - because they accord with your political / religious / philosophical perspective?

Let me remind you that climate models at best can be viewed as an aid to understanding complex processes that have *not* passed any formal / rigorous validation in pretty much any meaningful way - as understood by people who make critical decisions based on mathematical models in other disciplines which are almost invariably informed by that validation.

What part of - we are missing part(s) of the system, our knowledge is incomplete, the models are presently broken do you not get?

Gav's insistence that his models are skillfull and can make meaningful and verifiable predictions are simply insupportable. Just *look* at the bust between GISS sims and OCO-2 CO2 data ....

Buggering about like this wastes everybody's time and I would say getting it wrong and then trying brazen it out simply mires what should be honest science into tacky advocacy.

I doubt I am alone in my exasperation with people who do not have the basic honesty to say "we don't know" and seek to distort and subborn observation to a belief system.

You think the models work OK? - well more fool you pal.

Apr 24, 2016 at 7:59 PM | Registered Commentertomo

Phil Clarke is obviously a paid parody account, funded by big oil to make climate scientists and defenders of the consensus look stupid. No one can really be this daft or overly emotionally involved.

I think everyone should stop. It's not fair, leave the sad monomaniac alone.

Apr 24, 2016 at 8:31 PM | Unregistered CommenterAndy

ROFLMAO - hindcasts? An example where the result was known before the model was even written, much less executed?

Go learn why out-of-sample tests are important when assessing a model performance - it avoids self-delusion, something you appear to be good at.

Apr 24, 2016 at 8:51 PM | Unregistered CommenterSpence_UK

Watch out lads. Phil might post a link to his blog as evidence of how wrong you all are. Then you WILL be doomed!


Apr 24, 2016 at 8:55 PM | Unregistered CommenterMailman

Climate models made by real scientists are flawless. Phil Clarke is correct and all of you are wrong.

Apr 24, 2016 at 8:56 PM | Unregistered CommenterAyla

This Phil Clarke entity is almost clever (or has fast-track access to someone[s] who is/are) - he must be doing what he is doing on purpose? What a waste.

Apr 24, 2016 at 9:28 PM | Unregistered CommenterHenry Galt

Ayla, the perfection of climate models in portraying what they were programmed to portray, is unprecedented. It is wonderful for taxpayers to know the lavish lifestyles they have paid for, as such a modest contribution towards realising the fantasies of climate scientists.

If only these fantasy modellers could program their models to produce some real and useful information, everyone would be so much happier.

With the current rate of progress, it is now clear that mankind is going to have to find some means of adjusting the climate, to match the models, as the models can not be adjusted anymore, without the climate scientists looking particularly stupid and/or downright dishonest.

Apr 24, 2016 at 9:40 PM | Unregistered Commentergolf charlie

golf Charlie you presume to doubt Ayla? She proclaims climate models to be flawless yet within 12 minutes you broadcast blasphemous heresy. You falsely claim a need to modify the climate to fit the aforementioned flawless models.

Ayla has need of your turkish gullet, else your twisles would have undoubtedly been twisted beyond even the NHS's junior doctor's powers to heal. Repent and sin no more.

Apr 24, 2016 at 11:26 PM | Unregistered CommenterAlan Kendall

Still cannot see the "genius" of Phil Clarke at climate audit...could have been so easy Phil... :)
Just hand waving here...but still too busy to actually have a shot at the tile at CA.
cowardasur.. :)

Apr 25, 2016 at 12:00 AM | Unregistered CommenterDrapetomania

Alan Kendall, now you confusing me with technical terms. A turkish gulet is a large vessel for transporting loads, possibly including food. A turkey's gullet is a large vessel for transporting loads of food.

For some people, Turkish Delight is cruising in a gulet, but a stormy depression can lead to dumping the contents of the gullet on the gulet.Such Turkish Delight is rarely recycled.

Apr 25, 2016 at 12:21 AM | Unregistered Commentergolf charlie

To show that Phil Clarke never goes there. Climate Audit has this interesting quote from St Gavin
“As an aside, people are often confused by the ‘baseline period’ for the anomalies. In general, the baseline is irrelevant to the long-term trends in the temperatures since it just moves the zero line up and down, without changing the shape of the curve. Because of recent warming, baselines closer to the present will have smaller anomalies (i.e. an anomaly based on the 1981-2010 climatology period will have more negative values than the same data aligned to the 1951-1980 period which will have smaller values than those aligned to 1851-1880 etc.). While the baselines must be coherent if you are comparing values from different datasets, the trends are unchanged by the baseline.” – gavin
So the take home message is it is alright if we do it, but not if John Christy does it.

And note how the models always have anomalies - why not real world temperatures?. Could this be because they are just so wrong that the anomalies is just a fudge to hide it?

Apr 25, 2016 at 2:44 AM | Unregistered CommenterChrisM

Thanks ChrisM..nice
Sort of settles the case...
Phil Clarke is set and match.
I think phils anger is due to a sense of guilt that...
1/He is clearly out his depth..
2/He would be shredded on CA..and he knows it.......just another clueless idiot/coward for the $CAGW$ game.
3/He is connected to the grid and uses his self righteous hand morally bankrupt..and just..sad.
The shame..

Apr 25, 2016 at 8:01 AM | Unregistered CommenterDrapetomania

The caption should surely have read "Jockey tries to ride the gap".

Apr 25, 2016 at 8:04 AM | Unregistered CommenterPunksta

No gC it is my infernal spellchecker that creates these new technical terms behind my back. I do believe it could have defeated that Go Master long ago. Still gulet, gullet is it important?*

*once again the guardian of words initially wanted me to use gullet. But I fought it to a no points draw.

Apr 25, 2016 at 8:08 AM | Unregistered CommenterAlan Kendall

Phil's brief is inadequate. Were I he, I'd be angry.

Apr 25, 2016 at 8:38 AM | Unregistered Commenterkim

In general, the baseline is irrelevant to the long-term trends in the temperatures since it just moves the zero line up and down,

Thank you. That is exactly the point! When comparing two anomaly series that vary over time, where you choose to anchor them, ie the baseline or reference period - can move one higher or lower relative to the other, make differences appear greater or less. In the past Christy has supplied misleading charts for presentation to Congress where the series were crudely zeroed at 1979. Just another cherry-pick.


Now, about the uncertainties in observations ….

Apr 25, 2016 at 9:05 AM | Unregistered CommenterPhil Clarke

no warming since 1993:

Apr 24, 2016 at 7:18 PM | Ibrahim

It's been cooling for some 5000 years, so what the current fuss is about God only knows. Any temporary warming blips should be welcomed with open arms. Instead, thousands of public servants worldwide run around screaming the sky is falling, and telling us only more money - given to them - can fix this.


Apr 25, 2016 at 9:20 AM | Unregistered CommenterJeremy Poynton

To answer the question you posed up post
Spence - just to pin you down - you believe the comparison between observations and models is insensitive to the choice of baseline?

Yes or no will do."


Apr 25, 2016 at 9:24 AM | Unregistered CommenterChrisM

no warming since 1993:

Huh? RSS since 1993

Apr 25, 2016 at 9:29 AM | Unregistered CommenterPhil Clarke

Thanks, ChrisM.

Christy's graphs are for one purpose only - to provide a visual aid to the narrative that the models overestimate temperature, Judith Curry wrote that at the recent Congressional session they had one minute on each slide, so its all about the immediate visual impact on a non-technical audience. Every choice about the charts, from the choice of baseline to the erasing of uncertainty bands is designed to maximise that impact, and accuracy can go hang.

It is extraordinarily naïve to pretend otherwise, 'auditing' notwithstanding.

Apr 25, 2016 at 9:38 AM | Unregistered CommenterPhil Clarke

Yes Phil, Christy's graphs might over-estimate temperature differences , but the models they are being compared with can't even get the absolute temperature right, so maybe Christies Graphs should have shown the model spread from 12° to 16°
But then my final word on the subject is what your auditor posted
"There are several Schmidt diagrams – I should have been more clear. I haven’t been able to replicate the grey envelope of runs. I get a narrower envelope.

Also, RSS does not have a version 4.0 online for their lower troposphere series (TLT) which is the comparandum for the runs in the Christy diagram. There is a version 4.0 for TMT but that’s for a different level. I don’t know what the effect is.

Gavin’s use of envelopes and histograms also disguises the distributions and the improbabilities. It’s very cheeky of him to make accusations about Judy and others, when his own diagrams are so subject to the faults that he accuses others of"
As others have noted, You can redefine the English language as you will - no doubt you are Bill Clinton's lawyer moonlighting or are you just a wannabe.

Apr 25, 2016 at 10:13 AM | Unregistered CommenterChrisM

That's right, you crazy deniers!!!

Real scientists have never used models to provide a visual aid to any narrative in order to create an immediate visual impact on a non-technical audience.

And as Phil Clarke has ALSO PROVEN over and over, models do not overestimate temperature, they are perfect.

He is a hero and he's made fools of you all.

Apr 25, 2016 at 10:36 AM | Unregistered CommenterAyla


We have no need of visual aids, they have long been rejected.

Only the words of the all knowing, of whom Clarky boy is but a callow youth, now suffice.

We bathe in your perspicatitudinous glory.

The methane flow rate is now satisfactory?

Apr 25, 2016 at 11:00 AM | Unregistered CommenterAlan Kendall

Ayla 10:36 Yeah right.

My trusty bit of seaweed, dangling on a piece of string, is no wetter or drier, colder or warmer, than it has been for decades, at about this time of year. It is a bit smellier, and just like climate science, no longer fit for cooking. It needs replacing.

Apr 25, 2016 at 12:02 PM | Unregistered Commentergolf charlie

To answer the question you posed up post
Spence - just to pin you down - you believe the comparison between observations and models is insensitive to the choice of baseline?

Yes or no will do."


I disagree, although it is contingent on how I assess the results (which I believe is the only scientific way to do so). Let's take the two examples given here - Christy's plot and the plot at climate lab book (CLB).

Let's just note a few quick things about the plots. For all Clarke's whining, both plots use a spaghetti method to show the distribution of model spread, and neither explicitly show obs uncertainty beyond showing a few different lines. The other obvious difference is the baseline (for Christy, 1979 derived from 1979-2015 data, for CLB, 1995 ish from 1986-2005 data). Another difference is that Christy uses 5-year smooths (not a big deal, but I prefer annual data as per CLB). But there is another HUGE difference - Christy plots a model spread of 32 data sets, whereas the CLB plot shows 180 data sets for the model spread spaghetti. (FWIW the CLB plot spaghetti uses lines that are too thick - it is impossible to judge the shape of the distribution with so many thick lines. Thinner lines would have been better).

The model spread therefore is slightly different and this needs to be taken into account when assessing the graphs. The figures I give below are the standard deviation spread from normally distributed data. Now the results are not normally distributed - but assuming the kurtosis is not far from normal it will give a pretty good indication of relative spread. I've included "standard" 2-sigma spreads, and a 5% to 95% spread for comparison.

10 lines gives you a spread of approx. 5% to 95% or about +/- 1.65 sigma
20 lines gives you a spread of approx. 2.5% to 97.5% or about +/- 1.96 sigma
32 lines gives you a spread of approx. 1.6% to 98.4% or about +/- 2.15 sigma *Christy
180 lines gives you a spread of approx. 0.3% to 99.7% or about +/- 2.77 sigma *CLB

So the model spread on the CLB diagram is about 29% *wider* than the Christy diagram. Nothing wrong with this, other than the fact that we must take it into account when we interpret the results.

So, to make a like-for-like comparison of divergence of models and observations, we need to compare similar distances from the baseline centre (BAC). My recommendation is we consider 10 years, 20 years, 30 years from BAC. (You can see why I didn't abbreviate that to BC...)

For the Christy graph:
BAC + 0 years is 1979
BAC + 10 years is 1989
BAC + 20 years is 1999
BAC + 30 years is 2009

For the CLB graph

BAC + 0 years is 1995
BAC + 10 years is 2005
BAC + 20 years is 2015
BAC + 30 years is 2025

So, comparing the two graphs. After ten years, the observations on both charts are still within the model spread, although in both cases they look to be below the mean. (Nothing wrong with that, of course, if they are within the model spread). The Christy plot shows the obs near the edge of the distribution, but it is worth noting that that is a local low, and in fact the observations come back towards the distribution in the following 5 years. So at BAC+10 years, both charts agree that the observations are within the model spread.

At BAC+20 years, they appear to tell a slightly different story. At 2015, the CLB plot shows the data moving inside the distribution, but the satellite data is slightly outside it. But! The CLB is showing unsmoothed data, comparatively 2014 is right on the edge of the distribution. And this is where the difference in model spreads becomes important. A point at the edge of the distribution is 2.77 sigma for the CLB plot, but only 2.15 sigma on the Christy plot. When you realise this, the 20 year data are actually largely in agreement, other than the el nino spike in 2015. Both show around a divergence between models and observations of the order of 2.5 to 3.0 sigma around the 20 year mark.

This assumes that the 2015 el nino will not persist for much longer - of course, that may not be the case. We will have to wait and see.

For BAC+30 years, well only the Christy graph has this, and it shows continuing divergence. We will have to wait until 2025 to understand whether or not the CLB plot ends up showing the same thing - but when 2025 comes around, it will be important to note that the model spreads are around ~29% greater on the CLB plot.

So overall, the choice of different baseline makes no difference to my analysis whatsoever, beyond the fact that the later baseline requires us to wait longer for the data. Both show after ~10 years, the obs are within 2 sigma of the model spread; after ~20 years, they are moving outside the 2 sigma spread. After ~30 years, they should be comfortably outside the 2 sigma spread, but we'll have to wait for that to be confirmed for the CLB plot.

So what is different about the choice of baselines, other than how long we have to wait for data?

Apr 25, 2016 at 12:26 PM | Unregistered CommenterSpence_UK

Seaweed science has advanced enormously since you started using your Mark II model.
New models appear most months, now being 97% more accurate.
A new recalibrated algal model (from Mann & Co, purveyors of past climate heresy removal devices (PCHRDs, also known as "pilchards")) will reveal to you the greatly enhanced air temperatures your old model was incapable of detecting, and at the same time reveal to you just how acid the oceans have become.

Treat yourself gC, update and join the climate cognoscenti.

Apr 25, 2016 at 12:31 PM | Unregistered CommenterAlan Kendall

For convenience, the two graphs I was comparing:

Christy can be found at Judy Curry's place:

Climate lab book plot can be found at climate lab book here:

(I've posted this separately as it may get stopped by the spam trap since it contains multiple links)

Apr 25, 2016 at 1:58 PM | Unregistered CommenterSpence_UK

Alan Kendall, I have decided that the MkII, shall consist of Bladder Wrack (Fucus vasiculosus) In either the Latin, or anglo-saxon, it is a very handy set of pseudo expletives, and therefore very relevant for climate pseudo-science comparison purposes.

Apr 25, 2016 at 5:35 PM | Unregistered Commentergolf charlie



Apr 25, 2016 at 10:30 PM | Unregistered CommenterAlan Kendall

For convenience, the two graphs I was comparing:

Yeah, and in one the observations are more than 0.5C below the models, in the other they are splat in the middle. Because: baselines.

Phil 'wrong in just about every way imaginable' Clarke.

Apr 25, 2016 at 11:32 PM | Unregistered CommenterPhil Clarke

Yeah, and in one the observations are more than 0.5C below the models, in the other they are splat in the middle. Because: baselines.

No. Not even close to right.

One is "slap bang in the middle" (read: slightly below the middle) because it has annual data and is a strong el nino year (2015). Note that an equivalent el nino year (1998) near the baseline reference is actually off the top of the model spread. Note the three years prior to that 2015 (2012-2014 incl) are all at the edge of the model spread, which is at approx 2.8 sigma. But if you think 3 out of 4 data points at 2.8 sigma is just fine, you are welcome to your little delusion my friend. Back in the real world, that's a big problem.

The other reason is that it is 20 years from its reference point, and if you look at the data from Christy's graph, at 20 years from its reference point (circa 1999) the data is at around... 2.5 to 3.0 sigma. The exact same sigma relationship at the exact same distance from the reference point.

What a surprise! The two plots say the same thing, only the CLB plot doesn't show the failure of models at 30 years from reference point, which it most likely will by 2025, as I already explained.

Can you tell us Phil, what numbers did you that you feel the need to abuse them so?

Apr 26, 2016 at 12:26 AM | Unregistered CommenterSpence_UK

Alan Kendall, terribly sorry for my memory lapse, but Bladder Wrack is Fucus vEsisulosus, not vasiculosus. For people who appreciate a good Fucus, 'Up in Arms', may be their most likely response in letter to The Times.

The promotion of a good Fucus up into the upper echelons of climate science modelled predictions, is inevitable, if it has not occurred already.

It is that seaweed with the flotation bladders that can be popped like bubble wrap. It should not be used for wrapping delicate items, unless they really, really should smell like manky seaweed, or worse, like climate science.

Apr 26, 2016 at 12:49 AM | Unregistered Commentergolf charlie

Spence_UK, apologies for interrupting, but I thought I would post something as irrelevant as Phil Clarke.

Apr 26, 2016 at 12:57 AM | Unregistered Commentergolf charlie

Golf Charlie

Please feel free to interrupt! The asides are quite enjoyable. Although you do yourself a disservice - I believe your post makes infinitely more sense than anything yet posted by Mr Clarke.

Apr 26, 2016 at 2:21 AM | Unregistered CommenterSpence_UK

golf Charlie. Think nothing of it, what's a Fucus between friends? When you've seen one Fucus you've seen them all . I would not use the algal bubblewrap variety, however, because with all the anticipated heat waves that climate change will bring, those poppers could well explode, converting your Green-approved weather monitor into a claymore mine. Enough said!

Do try a pilchard* (see earlier post) they are so much healthier for you.

* my devil-spawn spellchecker refuses to recognize this word. I hate it when it does that, I never know if my spelling abilities are down to par, or if the word's just not in its vocabulary.

Apr 26, 2016 at 6:22 AM | Unregistered CommenterAlan Kendall

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>