Guest Post by Willis Eschenbach
A learned man was arguing with a rube named Nasruddin. The learned man asked “What holds up the Earth?” Nasruddin said “It sits on the back of a giant turtle.” The learned man knew he had Nasruddin then. The learned man asked “But what holds up the turtle”, expecting Nasruddin to be flustered by the question. Nasruddin simply smiled. “Sure, and as your worship must know being a learned man, it’s turtles all the way down …”
I’ve written before of the dangers of mistaking the results of the ERA-40 and other “re-analysis” computer models for observations or data. If we just compare models to models and not to data, then it’s “models all the way down,” not resting on real world data anywhere.
I was wondering where on the planet I could demonstrate the problems with ERA-40. I happened to download the list of stations used in the CRUTEM3 analysis, and the first one was Jan Mayen Island. “Perfect”, I thought. Middle of nowhere, tiny dot, no other stations for many gridcells in any direction.
Figure 1. Location of Jan Mayen Island, 70.9°N, 8.7°W. White area in the upper left is Greenland. Gridpoints for the ERA-40 analysis shown as red diamonds. Center gridpoint data used for comparisons.
How does the ERA-40 reanalysis data stack up against the Jan Mayen ground data?
Figure 2. Actual temperature data for Jan Mayen Island and ERA-40 nearest gridpoint reanalysis “data”. NCAR data from KNMI. Jan Mayen data from GISS.
It’s not pretty. The ERA-40 simulated data runs consistently warmer than the observations in both the summer and the winter. The 95% confidence intervals of the two means (averages) don’t overlap, meaning that they come from distinct populations. Often the ERA-40 data is two or more degrees warmer in the winter. But occasionally and unpredictably, ERA-40 is 3 to 5 degrees cooler in winter. Jan Mayen’s year-round average is below freezing. The average of the ERA-40 is above freezing. The annual cycle of the two, as shown in Figure 3 below, is also revealing.
Figure 3. Two annual cycles (Jan-Dec) of the ERA-40 synthetic data and Jan Mayen temperature. Photo Source
The ERA-40 synthetic data runs warmer than the observations in every single month of the year. On average, it is 1.3°C warmer . In addition, the distinctive winter signature of Jan Mayen (February averages warmer than either January or March) is not captured at all in the ERA-40 synthetic data.
So that’s why I say, don’t be fooled by people talking about “reanalysis data”. It is a reanalysis model, and from first indications not all that good a reanalysis model. If you want to understand the actual winter weather in Jan Mayen, you’d be well-advised to avoid the ERA-40, or February will bite you in the ice.
The use of “reanalysis data” has some advantages. Because the reanalysis data is gridded, it can be compared directly to model outputs. It is mathematically more challenging to compare the model outputs to point data.
But that should be a stimulus to develop better mathematical comparison methods. It shouldn’t be a reason to interpose a second model in between the first model and the data. All that can do is increase the uncertainty.
In addition, due to the fact that both models involved (various GCMs and the ERA-40) are related conceptually (being current generation climate models), we would expect the correlations to be artificially high. In other words, a model’s output is likely to have a better fit to another related model’s output than it does to observational data. Data is ugly and has sudden jumps and changes. Computer model output is smooth and continuous. Which will fit better?
My conclusion? The ERA-40 is unsuited for the purpose of validating model results. Compare model results to real data, not to the ERA-40. Comparing models to models is a non-starter.
Regards to everyone,
w.
[UPDATE] Several people have asked about the sea surface temperatures in the area. Here they are:
Figure 4. As in Figure 2, but including HadSST sea surface temperature (SST) data for the gridcell containing Jan Mayen. SST data from KNMI
Figure 5. As in Figure 3, but including HadSST sea surface temperature (SST) data for the gridcell containing Jan Mayen. SST data from KNMI
Note that SST is always higher than the Jan Mayen temperature. This is not true for the ERA-40 reconstruction model output.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Vidar says:
March 9, 2011 at 3:21 am
As I fist said, validation of models is very important, and robust validation techniques are needed. But unfortunately, as I see it, Willis here demonstrates how NOT to validate a model. It could have been an interesting excersice, but he fails to recognize the limitations of his study, and he also reveals a severe lack of understanding when designing his analysis. And I am a bit worried when I see all the praise he gets for his (failure of an) attempt. This is NOT the kind of analysis I would like to use to replace what is being done at institutes around the world. But obviously, the “crowd” of climate sceptics would more than welcome such “science”. Am I right?
No, Vidar , you are wrong! The “crowd” of people who are sceptical of climatism wants facts, real observations, conclusions drawn from carefully relating observations and assessing the outcome, in short: THE SCEPTICAL CROWD WANTS REAL SCIENCE! And not untestable models which use highly questionable input figures instead of data, as Willis rightly pionted out.
I want to thank Colonel Sun (March 9, 2011 @ur momisugly 6:45 am) for his effort in sourcing the information on the Argo resolution! Has it ever occurred to you, Vidar, to question the ERA-40 resolution, did you ever attempt what Willis did – to question the model and try to validate it? If so, it would be nice to share your findings. If you didn´t, then it might be time to begin. Go for it, question the models as well as you can, help all of us – not just the “crowd” of sceptics – to better understand these issues. AND TRY TO BEHAVE LIKE A TRUE SCIENTIST: Look for the facts, try as hard as you can to find facts which invalidate your models and theories.
As I see it, ERA-40 can not be validated at all! If this is true, then why is it used? Is there no scientist around who has the courage to stand up for science, for sound and solid labor? Eisenhower was just all too right!
Izen says:
“I would like to see a link to a quote of Micheal Mann using the ‘runaway global warming’ meme in the sense of unlimited and accelerating warming from CO2 or temperatures increasing exponentially.”
Here you go.
And obviously you didn’t read [or maybe didn’t comprehend] the link I provided with the definitions of a Conjecture, a Hypothesis and a Theory. There is no possible way that the CAGW conjecture could fit the definition of a theory.
Your argument is simply a consensus argument, with no factual evidence. Arrhenius recanted his 1896 paper with a newer 1906 paper – which reduced climate sensitivity to less than the lower end of the IPCC’s current guesstimate. The claim that Arrhenius formulated a “theory” [which by definition must be able to make reliable and reasonably accurate predictions] has been debunked.
If you have a problem with the definition of a theory, Dr Glassman has a link at the end of his article. You can ask him directly to explain the differences to you. Words matter, and in this case you are using incorrect words.
Izen says:
“Only the scientifically ignorant would think that RUNAWAY global warming is possible.”
Is it possible on VENUS?
I think the scientifically ignorant (IPCC, consensus…) are being exposed.
With all the discussion/counter-discussion of Willis’ blurb, the one point that seems to be missing from the discussion is the resolution is too poor from a Nyquist sampling point of view. All the fun things they’re doing to upscale the data, just means they’re trying to get around the lack of resolution. Peanut buttering data across wide swaths of the planet doesn’t get anything but useless, garbage, numerical collections – I won’t dignify the results as datasets, they’re not. As a starting point for GIGO, they are, however, perfect.
There is no substitute for measured data with error bands and a knowledge of the time sequence of the samples. When we get measurements with decent temporal resolution, and reasonable error bands, then we can start seriously discussing how the data can help provide insight into what’s going on with the climate.
wsbriggs wrote:
March 9, 2011 at 9:13 am
“With all the discussion/counter-discussion of Willis’ blurb, the one point that seems to be missing from the discussion is the resolution is too poor from a Nyquist sampling point of view. All the fun things they’re doing to upscale the data, just means they’re trying to get around the lack of resolution. Peanut buttering data across wide swaths of the planet doesn’t get anything but useless, garbage, numerical collections – I won’t dignify the results as datasets, they’re not. As a starting point for GIGO, they are, however, perfect.
There is no substitute for measured data with error bands and a knowledge of the time sequence of the samples. When we get measurements with decent temporal resolution, and reasonable error bands, then we can start seriously discussing how the data can help provide insight into what’s going on with the climate.”
Well said.
Thank you for succintly making the point I was belabouring.
I see much angst in some of the comments regarding the temps recorded on Jan Mayen vs the temps over the ocean. Being a simplistic guy, I tend to think, well, simplistically.
We are measuring AIR temps, here, right? Well, the isle is 34 miles long (please forgive the units. Even after decades in research, I still prefer the mile to km) and VERY narrow. A quick search shows that the AVERAGE wind speed on the isle is over 14 mph. So, on average, the island air is replaced with air from over the ocean in less than 4 seconds, max (more likely less than one second unless the wind is in the exact direction of the isle’s length), unless you believe the wind is simply rotating only over the isle (that would be a neat trick for nature!).
So why wouldn’t the AIR temps of the isle be exactly the same as the ocean AIR temps? I would welcome a reasoned response as to why the air temps should differ considering the size of the isle and the wind speed.
Alas, in rereading my post I see that my fingers got ahead of my brain. Those seconds should read hours, and to be more precise, at average windspeeds, the maximum time any given air stays over the isle ranges from about twenty minutes to 2.5 hours, with the probability of 20 minutes being closer to reality than hours (look at the shape of the island).
The jest of the argument remains. How much of a temp change will the air pick up while over the isle? Also consider where the temperature is being measured. If the temperatures are measured at a location in the center of the isle, the air would have been over the land from about 10 minutes to 1 and one-quarter hours. I would posit that the island/ocean air is too well-mixed to have any significant temperature deviation between them.
Hi Willis,
I’m just trying to get my head around all of this and learn a bit of basic statistics along the way, and I wonder if you could clarify something?
I don’t understand how you computed your 95% confidence intervals on the data. I was under the impression that the 95% CI for normally distributed data (which I am guessing you assume) is 1.96*SD (where SD is the standard deviation). For both the data sets you present the SD is fairly large, order 4 degrees I think. So how did you possibly get such tiny numbers as your “error” on the mean of the datasets? I expected it to be much larger from looking at the scatter in the data. Please could you clarify?
many thanks!!
So why wouldn’t the AIR temps of the isle be exactly the same as the ocean AIR temps? I would welcome a reasoned response as to why the air temps should differ considering the size of the isle and the wind speed.
Indeed why ?
And why should the rest of the air which is only in contact with the ocean (hundreds of thousands km² of it) differ significantly from the temperature of the ocean considering its size and the wind speeds ?
There is only one answer :
we have no clue what it is elsewhere than on exactly 1 point which happens to be the station on the Mayen island .
For me THIS is the point Willis made : a “reconstruction” of an average air temperature over 10 000 km² of ocean where only 1 point is measured and this point happens NOT to be above the ocean is just a numerical artefact which might be about anything you want .
If you feel like that you can say that it follows closely the SST . If you don’t then you can say that it follows closely the Mayen data . Or anything in between .
Whatever you choose can’t be falsified by real measures anyway pretty much per definition .
That’s what wbriggs called GIGO .
Vidar says:
March 9, 2011 at 3:21 am
I picked a gridcell with only a single data point in it for ease of comparison. I picked it because, if there’s only one temperature datapoint there, it makes a good test for how they are comparing the model and the planet. I’m sorry that you didn’t understand that, other folks seemed to get it.
Vidar, you do understand that:
a) You are calling me a liar?
b) Historically I (and other honest men) don’t respond well to being called a liar by some random fool on the internet?
c) When you call someone a liar, you automatically lose the argument and look like an idiot to boot?
d) Even if you might be right, if you call someone a liar everyone laughs at you and pays no attention?
Because if you don’t, I recommend some real fast study if you don’t want to be lisping in later life. You may not know this, as it appears from your actions that you don’t number them among your friends, but calling an honest man a liar is not a good business plan for your front teeth, and while I’m a nice guy, not everyone is.
w.
PS – Regarding your claims, if you believe them, go out and do the kind of analysis that you claim I’m doing so terribly wrong and show us the results … because so far all we have is your word for it that you have the inside track on the great plan. You have a big and unpleasant mouth, but so far, you’re just talk. Ugly talk to be sure, but still just talk.
And as your actions have shown to date, your talk isn’t worth anything. So far all I know about you is that you’re the kind of jerk that accuses people of lying when you’ve never met them and have no clue about them. That doesn’t increase your credibility, Vidar, quite the opposite. After that, the idea that a reasonable person should then blindly believe your claims is ludicrous.
Mark W says:
March 9, 2011 at 6:51 pm
What I quoted was the confidence interval (CI) of the mean (average) of the data, not the confidence interval on the data itself, which you discuss. The CI of the mean is equal to
SD/sqrt(N-1)
where SD is the standard deviation of the data, and N is the number of datapoints.
Keep asking questions, it’s the only way to learn.
w.