Climate Models

My article this week at Forbes.com digs into some fundamental flaws of climate models

When I looked at historic temperature and CO2 levels, it was impossible for me to see how they could be in any way consistent with the high climate sensitivities that were coming out of the IPCC models.  Even if all past warming were attributed to CO2  (a heroic acertion in and of itself) the temperature increases we have seen in the past imply a climate sensitivity closer to 1 rather than 3 or 5 or even 10  (I show this analysis in more depth in this video).

My skepticism was increased when several skeptics pointed out a problem that should have been obvious.  The ten or twelve IPCC climate models all had very different climate sensitivities — how, if they have different climate sensitivities, do they all nearly exactly model past temperatures?  If each embodies a correct model of the climate, and each has a different climate sensitivity, only one (at most) should replicate observed data.  But they all do.  It is like someone saying she has ten clocks all showing a different time but asserting that all are correct (or worse, as the IPCC does, claiming that the average must be the right time).

The answer to this paradox came in a 2007 study by climate modeler Jeffrey Kiehl.  To understand his findings, we need to understand a bit of background on aerosols.  Aerosols are man-made pollutants, mainly combustion products, that are thought to have the effect of cooling the Earth’s climate.

What Kiehl demonstrated was that these aerosols are likely the answer to my old question about how models with high sensitivities are able to accurately model historic temperatures.  When simulating history, scientists add aerosols to their high-sensitivity models in sufficient quantities to cool them to match historic temperatures.  Then, since such aerosols are much easier to eliminate as combustion products than is CO2, they assume these aerosols go away in the future, allowing their models to produce enormous amounts of future warming.

Specifically, when he looked at the climate models used by the IPCC, Kiehl found they all used very different assumptions for aerosol cooling and, most significantly, he found that each of these varying assumptions were exactly what was required to combine with that model’s unique sensitivity assumptions to reproduce historical temperatures.  In my terminology, aerosol cooling was the plug variable.

  • netdr

    Renewable Guy:

    As PaulD would say lets get back to atmospheric aerosols in models. In section 4.2 of the paper NetDr gave they talk about how the atmopheric aerosols were used in the model. There is no talk of adjusting the aerosols to make the model work.
    *************
    The point is they never say explicitly that they adjusted the aerosols, that would be way too honest.

    Since both the amount and effect of the aerosols are unknown it is the perfect “plug variable” it can be whatever it needs to be to match reality.

    In the real world only one value is correct and all others are simply wrong.

    RE:
    ********Interesting article on paleoclimatology. 1 deg C warmer difference from our climate today with oceans 15 feet higher than what we have at present. They had only 300ppm co2.**********

    This makes the incorrect assumptions that we live in a one variable world and that CO2 is a major determinant of the temperature of the earth. Since we have 380 PPM right now that fact must be apparent.

    I still don’t understand:
    1.25/6 decades = .21 C/decade [Where does this come from ?]
    1.00/6 decades = .17 C/decade [Where does this come from ?]

    http://cstpr.colorado.edu/prometheus/archives/hansenscenarios.png

    Taking 1960 to 2020 scenario “B” predicts .17 per decade.
    [Even that is unfair because it is 1988 when he publishes his model and he knows the correct answer for 28 years. No sale on that one.]

    The other number is a complete mystery to me!

    Where in the world does the .21 come from ?

  • Renewable Guy

    The point is they never say explicitly that they adjusted the aerosols, that would be way too honest.

    Since both the amount and effect of the aerosols are unknown it is the perfect “plug variable” it can be whatever it needs to be to match reality.

    In the real world only one value is correct and all others are simply wrong.

    #################################################

    If you can show Hansen is being dishonest, it would be interesting. If Hansen is doing something blatantly wrong, other scientists are there to review the work to keep the quality high. This paper has been out for over 25 years. That is plenty of time to catch cheating plagarism and lieing, let along mistaken assumptions or laziness.

    Richard Lindzen’s

    ####################################################

  • Renewable Guy

    Richard LIndzen’s last paper was reviewed by Trenberth and given critical remarks for its poor quality of science work.

    ######################################################

    http://www.skepticalscience.com/climate-models-intermediate.htm

    figure 2 in the intermediate section

    Lets pick a different set of numbers then. Just easy numbers eyeballing the graph.

    The average isn’t the endpoints, its an imaginary line going through the middle of the wiggling upward slope of three different lines this time.

    lets go between 1990 and 2005 off the graph

    scenario b 1990 average point .35
    2005 average point .70
    .70 – .35 = .35
    15 years = 1.5 decades
    .35/1.5 decades = .23 C /decade

    station data 1990 average point .25
    2005 average point .70
    .70 – .25 = .50
    .50/1.5 decades = .33 C/decade

    land ocean 1990 average point .20
    2005 average point .60
    .60 – .20 = .40
    .40/1.5 decdes = .27 C/decade

    Crude estimates, does that give you any idea where I am coming from? Given the data sets, the average slopes of each set could be obtained which would give us trends of the climate.

  • Renewable Guy

    http://www.metoffice.gov.uk/hadobs/hadat/images/update_images/tropical_upper_air.png

    One of the data sets that supports AGW theory is the statosphere will cool while the troposphere warms. The models predicted this before it was measured.

  • Ted Rado

    netdr:

    If I were to start studying chemical engineering all over again, I would take advanced courses entitled FUDGE FACTORS. I could then skip all the other classes, as this would enable me to fit any data to any theory. The magic bullet of science and engineering: the fudge factor!!!

  • netdr

    Renewable.

    My calendar says it is 2011 not 2005.

    Dr Hansen’s model looked pretty good as of 2005 [no one disputes that point] but it looks terrible as of today.

    http://cstpr.colorado.edu/prometheus/archives/hansenscenarios.png

    The model predicted quite a bit of warming but the world didn’t warm between 2005 and today.

    2005 = .63
    2020 = .63
    2011 so far .48

    [so it actually cooled from 2005 to 2011 so far.]

    The model will look even worse after 2011 is on the charts.

    Your “average point” method is unusual and error prone.

    The mathematicians use a least mean squares method to make trending less end point sensitive.

    They draw a line through the graph and manipulate the line until the square of the distance to each point is a minimum. Woodfortrees.org uses this method.

    From 1988 to present:

    #Least squares trend line; slope = 0.0153971 per year or .153 per decade

    http://www.woodfortrees.org/plot/hadcrut3vgl/from:1988/to:2012/plot/hadcrut3vgl/from:1988/to:2012/trend

    Hansen scenario “B” shows 23 years .55 ° C change in temp = .239 per decade

    You could take the yearly predictions of the model and put them in an Excel spreadsheet and do LMS on them.

    [It is pretty easy to do, but I couldn’t show you the results and you would think I cheated when the results showed the model off by about 215 %.]

  • netdr

    Renewable
    scenario b 1990 average point .35
    2005 average point .70
    .70 – .35 = .35
    15 years = 1.5 decades
    .35/1.5 decades = .23 C /decade

    station data 1990 average point .25
    2005 average point .70
    .70 – .25 = .50
    .50/1.5 decades = .33 C/decade

    land ocean 1990 average point .20
    2005 average point .60
    .60 – .20 = .40
    .40/1.5 decdes = .27 C/decade
    **************
    Why are all of your data points 2005 ?

    Is that a magic year or something ?

    I like 2011 much better. So far the anomaly is only .48 which is actually cooler than 1988.

    [Did the model predict cooling and I just missed it?]

    http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts+dSST.txt

    Like most models the longer they run the greater the error.

    Claiming to be able to predict 100 years in the future is ridiculous when they obviously can’t do it for 23 years.

    The alarmists like to pretend that errors tend to even out but instead they compound.

    A climate model is more like a professor that gives a 100 question final and the answer to question 1 is the input to question 2 etc etc.

    Since the processes are poorly understood there are substantial errors in each computation and far from cancelling out by the time you get to problem 100 no two students would get even close to the same answer.

    Almost no models include ocean currents which drastically change the atmospheric temperature which is itself an input to the next iteration so the errors compound. Those that do include them show 20 to 30 years of cooling stating about now.

    The alarmists need to strike now before the cooling becomes apparent to even the dullest.

  • netdr

    Renewable

    I collect climate model predictions and compare them to reality.

    Either there are no models that are correct or they don’t allow them to be published.
    [If you or anyone else has a correct one I would love to see it.]

    Here are several more EPIC FAILURES.

    http://www.c3headlines.com/2011/06/a-spectacular-failure-latest-hadcrut-nasa-temperatures-significantly-below-ipcc-climate-model-predic.html

    When will they get it right ?

    The answer is to predict no warming or some cooling for 30 years and then if they want to scare people they remove the aerosol [plug] and let the temperature climb like a homesick angel.

    By the time 30 years have elapsed the modeler should be safely retired.

    Simple isn’t it ?? They can be correct for 30 years and still be scary.