Climate Model Validation

I am sorry that posting has been light, but I am currently working to migrate this site to WordPress from hosted Typepad.  This is a real hassle, as described at my other blog where I just completed a succesful migration.  I hope to have this blog moved over this weekend.

In the mean time, I thought my readers might need some help understanding James Hansen’s recent comments that flat world temperatures over the last 10 years and substantially cooler temperatures in 2008 were entirely consistent with the climate models that forecast  0.2-0.3C (or more) warming for this decade.  Most other natural sciences are stuck in the old and outdated practice of questioning forecasts when actual observational data diverges from the forecast by several standard deviations.  Not so modern, enlightened, consensus-based climate science.  Below is my graphical representation of how climate scientists evaluate their models in light of new data.


21 thoughts on “Climate Model Validation”

  1. I’d say you got that about right.

    Is there any actual data that won’t be consistent with the models? Do we have to drop below 1978 temps before alarmists stop claiming the world is still warming because the 30-average is still up?

  2. Jeff, you’re real mean.

    I find climate models perform miracly well given that :
    “None of the models used by IPCC are initialized to the observed state and none of the climate states in the models correspond even remotely to the current observed climate”.

    No no, that’s not stated by some big-oil shill but by Kevin Trenberth, IPCC author with impeccable alarmist credentials

  3. Does anyone else have a problem with the images on this site? they simply won’t load for me. I am behind a firewall that blocks active content, but the firewall always labels what it blocks. The images here just come up as a red X and the alt text.

  4. “flat world temperatures over the last 10 years”: please tell me which of these is flat:

    By the way, it seemed impossible to make comments on your earlier post ‘Steve Chu: “Climate More Sensitive Than We Thought”‘. I’ll have to say here that I was appalled by your dishonesty in ignoring the Hansen 1988 scenarios B and C. Really, you should be ashamed of yourself. And in case you didn’t realise, your graph entitled “High feedbacks greatly over-predict past warming” is quite meaningless. No climate study of any sort has ever predicted that temperatures would respond to transient forcing by following the curves you plotted.

  5. Jenny:

    It all depends on your choice of starting point. Looking at the trend starting in 2000, unless next year is a real scorcher, the 10 year trend will be almost flat by the end of 2009. Also, if you go back and start looking at the trend from 1999 or 1998 the results are quite different.

    Which is not to substantiate the claim of recent flat trends – the results appear to be highly sensitive to your choice of starting point – which only indicates the data are quite noisy.

  6. Nice try, Jennifer.
    Your links were from 1999 for which temperature estimate had significant drops from 1998. Here is the flat — or decreasing trend — since 1998:
    Of course, 1998 had a large El Nino bump, so we are all familiar with that impact. Therefore, I am not impressed with claims that temperatures have been flat or cooling since 1998.
    Also, the OLS trends are highly dependent upon whether one uses summer 2008 or November 2008 as the end point. And the temperature estimates for the next few months will greatly affect OLS estimates in the coming months. So let’s be alert to cherry picking!

  7. An Inquirer: nice try, you say. Nice try at what?

    This graph shows you why any trend of less than about 15 years is not very useful. Trends over the last 29, 30 and 31 years differ almost indistinguishably. The upward trend over 30 years does not depend on your choice of starting year. Trends over the last 19, 20 and 21 years differ a little more, but still, the 20 year trend is not dependent on the choice of starting year. Trends over the last 9, 10 and 11 years are radically different, and it’s clear that there really is no such thing as a ‘ten year trend’. It is entirely dependent on your choice of starting year, and so it’s subjective and not reflective of reality.

  8. Question: I take it that the proponents of the “consensus” view regarding anthropocentric global warming explain the recent global temperatures as a transitory weather phenomenon of some sort, that will only last 10 years or something like that. My question is, did the proponents of the “consensus” view predict that there would be such a temporary temperature drop in advance of it actually occurring, or do they use this explanation only after the fact? If the latter, what is their explanation for their failure to anticipate the phenomenon in advance?



  9. “The upward trend over 30 years does not depend on your choice of starting year.”
    That’s nonsense!
    This graph shows trends over 30 years DO depend on the starting year. So 15 year trends are as “usefull” as 30 year trends, ie useless. Because trends (linear regressions) are simply irrelevant to represent cyclic changes in climate.

  10. Jennifer,
    I believe that we are in violent agreement that one should be suspecious of GMT trends over ten years. However, it appears that we have significant disagreement of GMT trends for 30 years. First of all, I am not sure that GMT is a meaningful concept. Second, more signficantly, the earth’s climate features oscillations, and perhaps the most significant one is 60 or maybe 70 years long. To take a linear trend over the positive 30-year-phase produces most misleading results. Third, I am not persuaded that we have a reliable record of temperature estimates for 60 or 100 years that we know within a couple of tenths Celsius what the real GMT was. Nothing that I have read from Hansen / GISS, Jones / HadCrut, (or UAH or RSS) convinces me that we do. Temperature readings that are flat in the U.S. go through computer gymnastics to show an upward trend. Ancedotal evidence throughout the U.S. shows the 1930s summers to be much more extreme than now. My limited contacts outside the U.S. do not lead me to believe that the U.S. was an isolated case.

  11. Demesure: you misunderstand. Plotting 30-year trends for two different periods of different climate behaviour does not tell us that 30 year trends are not robust, it tells us that the climate behaviour was different between the two periods. Change your start and end dates by one year in each phase, and you’ll get the same result.

    An Inquirer: I am sure we agree about ten year ‘trends’, but I wouldn’t say just be suspicious – I’d say disregard them entirely. I have a number of questions for you, because quite honestly I don’t understand the statements you are making or what evidence you are basing them on. How could global temperature not be a meaningful concept? Why doesn’t a 60-70 year oscillation show up in the instrumental climate record? How could you trust anecdotal evidence over instrumental data? Can you quantify ‘much more extreme’? Why would you characterise necessary calibrations as ‘computer gymnastics’? Do you think there is an ‘upward trend’ in the US temperature record? What makes you believe the instrumental record is not reliable?

  12. I don’t like the label skeptic because it conveys that we are not totally sure that the AGW theory is flawed. However, the AGW advocates have transcended beyond data to embracing beliefs. This defines mysticism as opposed to science so it would be appropriate for us to label them as mystics. In the world of science, I prefer being called a skeptic over a mystic any day.

  13. So are you saying when all you have left is lies, you lie?

    Works for me. Remember, the hardest thing for government to say is the three little words — “we don’t know”.

    There was a day when science was open, methods and conclusions were openly debated and discussed. Now we hide data, obfuscate results and try and convey a conclusion not supported by fact. The new age is upon us.

    I suggest that modern science take up where we once were, and take to the blog to become open once more.

  14. Why would you characterise necessary calibrations as ‘computer gymnastics’?

    Because the surface stations often turn up in close proximity to barbeque pits, and the calibrations add artificial heating to a system which should be corrected for a century of unremitted encroachment by UHI, while also rewriting the pre 1970 history to make the past colder.

  15. Jennifer according to nasan faq page for sat data they feel that we need minum 50 year data to determin climate. I disagree. Due to the cyclic nature of weather we would need to determin the length of the longest cycle and have accurate records for a minum of three cycles. I understand that the AMO Or the NAO could be as long as 80 years.
    I agree that there is nothing that can falsify the models as they have already falsified each other. Based on the fact that if none of the models are good enough to stand alone for the IPCC theory then combining them is just piling junk upon junk. GIGO
    Ray: concider your self a Realist. I no longer think of the other group that belives in agw as anything but Alarm Bell Ringers and will use the phrase ring that bell when conversing with them.
    So all of you that belive that we are getting warmer: Ring That Bell

  16. Jennifer, you say “Why doesn’t a 60-70 year oscillation show up in the instrumental climate record? “. If you look at the curves from hadcru and giss, there is a very clear trend of a 60 year oscillation superimposed on an upward trend. I cannot see how you can fail to identify this? The 60 year oscillation also means that – if it continues – the upward trend during the last 30 years was aided by the cyclical trend and will now be dampened by that oscillation.

  17. How could GMT not be a meaningful concept? For several reasons: First, if my head is in the oven, and my feet are in the freezer, the average temperature might be acceptable. Turn up the oven’s heat, and turn down the freezer’s cold, and the average stays the same, but I get worse. Second, if we have an average, should there be equal weight given to each geographical area? We know that GISS would like to do this, but temperature records do not exist to make this possible. GISS makes some heroic assumptions to accomplish this, but those assumptions are most suspicious. Also, the number of reporting stations has fallen drastically over periods of time, which makes equal weighting especially difficult. Third, the temperature for each geographical spot is not measured in a consistent method over the years. This, in addition to human influence, prompts the computer gymnastics. I would highly recommend you read up on the process of what happens to a raw temperature reading on its way to a GISS contribution. (If at one point in time, temperature readings are rounded to the nearest degree, we are supposed to get excited about a tenth-of-a-degree movement?) Fourth, the results of the computer gymnastics are often counter-intuitive. Consider Central Park in New York City where actual temperatures have been flat to declining over the years, but the adjustments show an upward trend. I have other thoughts, but for sake of time, I will move on.
    It is not that I would simply rely on anecdotal evidence and disregard instrumental reading. In fact, the anecdotal evidence that the 30s had hotter summers than the 90s does agree with the actual instrumental readings with which I am familiar. It is the temperature estimates produced by computer gymnastics that do not.
    Anecdotal evidence and temperature readings do suggest that the last 30 years have been warming, so I — and climate skeptics – have no problem in saying that the U.S. has been in a warming trend for decades. (In fact, mainline skeptics have no problem in saying that CO2 does have some impact on temperatures but disagree on the issues such as magnitude, positive feedbacks, and run-away scenarios.) What I do not know is whether the 90s were hotter than the 30s. Certainly the climate/weather impact on agriculture was far more disastrous in the 30s.
    If I – or others — have not addressed any of your questions, I apologize, but I run off. By the way, I do believe that your questions were good and politely asked.

  18. What you have neatly summed up in one graph is a perfect illustration of the non-falisfiable nature of AGW.

  19. @Jennifer,
    @An Inquirer
    Personally, I do take a different approach: I do take GISS, HadCrut, RSS and UAH monthlies and
    plot them from the current month, back as far No Trend is there. The sum – in months – is
    increasing since about 2.5 years.
    Another way is: getting the rate-of-change – I compare the monthly value to the same month in previous years and take the difference. Then you get the change on a currently existing trend – if it’s getting
    stronger, weaker and no change. By God, are the trends getting weak.
    Just a suggestion: Try it with CO² global monthly mean data
    Just my IMHO: the ROC may turn neutral within five to seven years

Comments are closed.