9 thoughts on “SciAm: Mathematicians Unite to Tackle Climate Change”

  1. Some statisticians have been coopted into the AGW fold already; I’m sure some other mathmeticians will go along for the ride. But so far the math says that the AGW theory is wrong.
    Using math to develop alternative scenarios for climate change (the real thing, not the AGW stuff), crop development, population changes and other real issues would be valuable if people would let facts be facts and would then vote for policymakrs who would act on the facts. What a concept.

  2. You can’t be a good scientist without recognizing the limits of statistical analysis. I suspect the reverse is equally true.

  3. I dropped my subscription to SciAm after their whole-hearted embrace of the anthropocentric global warming myth. One of the true values of the scientific method is the initial skepticism regarding any hypothesis until there are considerable independent verifications of the particular hypothesis. Simply saying “those guys know what they’re doing so they must be right” is not scientific analysis. It’s sad that a publication that pretends to report on science is so quick to accept leftist/ecologist based ideas.

  4. As long as two competing very competent weather bureaus, one of the worlds largest computers at hand (in a small European country), are off by almost 10 centigrade in their forecast for next week’s temperature there are serious doubts as to the feasibility of shoveling more money on “predicting” global climate, including temperature. The problem is not one of more maths, it is one of defining and formulating the model correctly, INCLUDING THE INITIAL CONDITIONS. Try this: take consequtive squares n(1)=1, n(2)=2*2=4,.., up to n(10)=10*10 = 100. that is, 1,4,9,16,25,36,…,100. Calculate the numbers n(i+1)-2*n(i)+n(i-1) for i=2,3,..9. You should get the result 2 each time. Assume then that you have a uniformly distributed error in your numbers corresponding to at most+/- half a percent of the largest one, that, is +/-0.5. Then you might get numbers like -0.489 0.835 3.679 8.849 15.588 24.665 35.569 48.579 63.512 80.99 99.744
    Try and calculate the same differences. Your results will now flucutate between 1.28 and 2.54!! One half a percent error in your starting values will give you an error margin of something like – 36 % – +27 % in your final values. That’s predicting climate, and even weather. Why this n(i+1)-2*n(i)+n(i-1) stuff? Because it gives the relation between cause and effect for any dynamical system (plus a bit of other things, too). Your 480 hp 1987 Buick Regal GN moves according to it, as does the water in the San Francisco Bay and the snow gliding down the Alps, the Buick a bit more predictably than the waves and the snow. Then, what if you measure those half percents? You can’t: there are too many of them. You also must measure how fast they are changing when you start your measurements. For the Buick, you can predict the trap speed because you know that it 1) starts at the starting line 2) with zero speed. Not so with all the snow in ana avalanche or all the water in the Bay. Well, there are lots of other phenomena involved that make things even more complicated: clouds, aerosols, soot on polar ice etc. etc.

Comments are closed.