.

Sunday, April 21, 2013

Tisdale's Tricks

Sou | 1:13 PM Go to the first of 9 comments. Add a comment

Global Surface Temperature Changes - Model and Actual

See update below.

Bob Tisdale has an article on WUWT about CMIP5 model outputs for RCP 6.0 for the period 1880 to present.  He puts up this chart:


Looks not bad, doesn't it.  The models have picked up Pinatubo very well.  While they appear to be less closely aligned to earlier periods that is where there is greater uncertainty in the record.  They are pretty close from the mid-twentieth century onwards.   Pity the chart as he shows it has no probability ranges.  (Incidentally, I have no idea of whether the chart is an accurate representation of actual temperature or the CMIP5 model outputs. No reason to doubt it but I haven't checked.)

Choosing the trend

Anyway, Bob decides to see if he can find a way to make it look bad.  So what he does is draw some different coloured 'trend' lines on different sections, one blue line and one red line.  He doesn't seem to bother with proper trend analysis or anything fancy.  All he does is divide the chart into his own segments (eyeballing only based on his comments).  Note how his dividing lines are at peaks and troughs in the record. That pretty well ensures a greater slope for the record than the model output.  He picks start and end points for short sections such that the trend lines for actual and modelled have somewhat different slopes.  That means (he thinks) he can say "aha" and "gotcha".  He seems to be arguing that there should be a perfect alignment at every point on the chart in segments of his choosing.

While some WUWT readers will grab onto the 'gotcha', they'll be the ones who think any article on WUWT (no matter what it is about) means "all the science is wrong".

I think many casual readers of WUWT will be surprised at how close the model output is to the actual record of surface temperature change.

There's a recent article on RealClimate.org discussing models and regional trends - more difficult to model than global trends.  Worth a read.

Post Script:

This post was meant to be just a dig at Bob's fiddling with 'trends'.  But it's probably worth a caution. If you manage to get towards the end of Bob's post you'll read nonsense like this:
Atrocious, horrible and horrendous are words that could be used to describe the performance of the CMIP5-archived climate models during the early warming period of 1917 to 1944. See Figure 7. According to the models, if greenhouse gases were responsible for global warming, global surface temperatures should only have warmed at a rate of about +0.049 deg C/decade. BUT according to the new and improved GISS Land-Ocean Temperature Index (LOTI) data, global surface temperatures warmed at a rate that was approximately 3.4 times faster or about 0.166 deg C/decade.
This shows a couple of things.  Firstly, that Bob views surface temperature records as being very 'correct' in the early part of this century, ignoring the larger error bars on the actual data (he doesn't show any), as illustrated below from NASA (green bars):



Secondly, despite blogging about climate for quite some years, Bob hasn't learnt what climate studies show.  He seems to think that greenhouse gases are said to have been the only forcing since industrialisation.  That's not what scientists have found.  In fact, science suggests that earlier last century, although greenhouse gas increases contributed to some extent, other forces dominated.  Here is a chart indicating likely 'natural' and anthropogenic contributions - from SkepticalScience (click image for larger version).  Read more here.


Thirdly, Bob is more hung up on short term 'trends' than on the long term rise in temperature.  He fails to see the forest, being distracted by the leaves on individual trees as they flutter slightly in the breeze. (I even wonder if in Bob's world, he replaces the anomaly with the trend so to speak.  One could speculate that even if surface temperature went up 4 degrees it wouldn't worry Bob. He'd be happy as long as he could compare short term slopes along the way (probably putting the huge rise in temperature down to a magical ENSO "jump").  For example, later on and despite the CMIP5 model being closely aligned with the observed data, Bob still reckons something is wrong because he thinks the 'slope' isn't right. I don't know what he thinks the models are based on (it's physics).)

Finally, what does he mean by "according to the models...should only have warmed..."?  He's looking right at the model output and that's not what they indicate at all.  If anything CMIP5 models indicate there 'should' have been slightly more warming, not less, than was observed in the early part of last century!

So go look at Bob's charts and maps if you want to, but ignore what Bob writes.  He's talking through his hat.

Update

Bob is still complaining for no reason.  He's done another post of the difference between the annual surface temperature anomalies and the CMIP5 model outputs.  His chart shows that for most of the 130 year period, the model was within +/- 0.1 degree Celsisus of the observed annual anomaly.  It deviates as much as +/- 0.2 degrees in the earliest records (before 1930) and in the most recent year.  In my view, that's a remarkably close alignment.  To put this into context, from one year to the next, the observed anomaly can fluctuate by more than 0.2 degrees Celsius!  The models just keep getting better.

The weird thing is that while WUWT-ers will say that the rise of 0.8 degrees Celsius is "nothing to worry about", they will be quick to say that a deviation of model output of less than 0.1 degree Celsius is "shocking".  They are too easily "shocked" and yet not "shocked" enough (about climate change).  (And I guess DB Stealey/Smokey won't be singing his usual song this time.  He won't be asking Bob to show the difference in actual temperature (not anomalies) in degrees Kelvin with the vertical axis from zero.)

Have a read of this for another take on Bob's misrepresentation.  Scroll down to the bottom to get a better view of how close the models are to reality these days.

9 comments:

  1. HotWhopper, whose side are you on linking to WUWT, why send all your followers to a denier site?

    ReplyDelete
    Replies
    1. Hi Bender, can't critique without reference. When I think it's necessary, I usually warn people to put on a head vice.

      In this case it's not necessary except for some of the comments. A lot of Bob's article is interesting, the chart (without his additions) and the mapped data.

      What is odd that instead of being amazed that clever people can construct physical models that hindcast 130+ years of climate so well and giving them kudos, Bob complains it's not good enough (and he does mislead his readers with his wonky 'trend lines').

      (Like eventually happens to everyone who has a bit of an understanding of science, Anthony's banned me from WUWT or I might have commented there.)

      Delete
    2. BTW - not sure what you mean by "sides".

      WUWT is known among climate bloggers as the "world's most viewed anti-science website". I think Tony's even won a couple of anti-science awards (bloggies) for being a prolific "swift-boater" and ad hom-er of some of the world's top scientists.

      Going by my small band of twitter followers, most of my readers understand science very well. I doubt any of them would be fooled by Tony's bombastic and shameless displays of scientific illiteracy.

      Delete
    3. Thanks for your response, I understand where you are coming from.

      Delete
  2. I thought bender was busy cleaning up Moshpit's pool:

    > My pool boy is there, so I would normally be very close by. Bender sends his love.

    http://judithcurry.com/2013/04/13/open-thread-weekend-13/#comment-312520

    ReplyDelete
  3. BenderApril 21, 2013 at 7:42 PM

    HotWhopper, whose side are you on linking to WUWT, why send all your followers to a denier site?

    ----------------

    WHAT ALL 5 OF THEM? This site got more traffic from the link being on WUWT than it's had in a month.

    ReplyDelete
    Replies
    1. Lol :) - Actually no. Going by the site stats, not many WUWT readers follow links. HW gets a lot more visitors from elsewhere.

      Delete
  4. Ed Hawkins at NCAS has written some interesting posts about CIMP5 vs observations.

    ReplyDelete

Instead of commenting as "Anonymous", please comment using "Name/URL" and your name, initials or pseudonym or whatever. You can leave the "URL" box blank. This isn't mandatory. You can also sign in using your Google ID, Wordpress ID etc as indicated. NOTE: Some Wordpress users are having trouble signing in. If that's you, try signing in using Name/URL. Details here.

Click here to read the HotWhopper comment policy.