Slashdot

>>Nonsense. This is only true if it's a linear relationship. Given that the greenhouse effect involves a complex feedback cycle, that is not a valid assumption.

Yes, as we all know from Al Gore's memorable definition of what a "non-linear system" is: "It's a fancy way they have of saying that the changes are not all just gradual. Some of them come suddenly in big jumps."

I used to work doing modeling of both ocean seawater and other things (like heart cells or full cardiac cycles) which attempted to accurately simulate whatever ODE or whatever it was we were simulating. These models were incredibly sensitive to the various constants used, and what the starting assumptions were. They'd fly off into incoherent-land if these values were not very precise, or if the constants didn't match each other. The only way we could calibrate or test our simulation was by, say, pulling out a rabbit's heart, wiring it up, flooding it with some solution, and having the severed heart beat for us when driven by impulses at different frequencies and amplitude. Testing and experimentation is the only way to truly know something, as Feynman said. If we just relied on the models without doing followup experimentation with them, we'd have gotten wildly inaccurate results.

Climatology, on the other hand, is "science-y", but not really science. It wants to be science, it really does - and goes through the window dressings of having peer reviewed journals and conferences and all of that - but ultimately it is not science. There is no experimentation involved (or if you will, there is one large experiment running all the time), and there is no control for the experiment. Forgive me if I do not allow your models to substitute for actual experimentation, for the reasons listed above.

As one of my professors once said, never listen to anyone who claims to be really accurate over the sample data set. It's real easy to be accurate on a sample data set. Hell, you can always just spit back out the original numbers if you want - for my neural net spam filter, we could have just returned the classifications of each email and claimed 100% accuracy, for example. If you don't think that climate researchers actually make bullshit claims like this, check out the wikipedia page on global climate modeling, and look at, say, this graph: http://en.wikipedia.org/wiki/File:GCM_temp_anomalies_3_2000.jpg [wikipedia.org] There's charts like that everywhere on wikipedia, showing how accurate the climate models are, even back in 1930, decades before the models were created.

What is important is the accuracy going forward into new data, and as they do, they've found numerous glaring problems with the predictive ability of climate models (such as rainfall changes being 25% of what is expected). (For some fun laughs, read predictions of what life would be like in 2010 written 10, 20 or 30 years ago.)

The simple fact of the matter is, I don't believe any (self-described) scientist who claims he knows how much temperature will move in the next 100 years, unless he says it will range somewhere between absolute zero and the temperature of the sun.

And if it sounds like I'm picking on climate "scientists", well, I am, but I had a number of friends who worked in the field at SIO, and they're generally smart and nice guys, and think there's a serious problem. Their problem lies in claiming more knowledge than they actually know. (Again, this is not how actual science works.) And it's not like other fields have looked enviously at the tremendous success of real scientific fields, like physics, over the last hundred years. Psychology, sociology, hell even scientology and philosophy have tried to co-opt the patina of science for themselves. (Nearly every modern philosopher since Wittgenstein calls themselves an analytic philosopher, which was a movement to directly make philosophy more "scientific" and less heads-in-the-cloudsy.)

Actually, you needn't look to such minute differences. Different isotopes do react at slightly different rates, so biological processes often enrich molecules in one isotope over another. I don't know of any way to use this to trace CO2's source, but it has been used to chemically trace the earliest appearances of photosynthesis on Earth, for example.

That said, your post is right: you can reasonably accurately measure and sum the man-made carbon sources.

Isotopic composition is a good test. For fossil carbon, all of the C-14 will have decayed, so if the fraction of C-14 has gone down over time then that's a good indicator that the increase is from a fossil fource.

What makes him think that since the measurement method is unreliable for the last 20 years, it is reliable for the rest of the time period hundreds of years back? It throws his entire theory out of the window. He is doing specious reasoning by cherry picking the results that fit his theory better.

If anything, that's the real black eye that the recent data swipe reveals. The emails between AGW scientists specifically mention bullying publications into not accepting/publishing papers that don't support AGW, and subsequently use the lack of published, peer-reviewed articles against those scientists whose conclusions differ from their own.

The one I find convincing is the melting of the ice. It's a crude kind of measure, lacking in detail, but any explanation that doesn't account for that is obviously unacceptable.

The North-West passage is currently open to ice-breakers, and it is projected to be open to normal passenger liners soon. This doesn't give me many data points, but the ones it give are irrefutable.