Replies

Interesting data, and it addresses one but not both of the points.

Taking the scatter in the various models as a measure of their collective accuracy (an assumption, not a fact, but a place to start), you can see that the consistency among the models leads to an estimate of about half a degree in uncertainty. The actual data show a man-made impact slightly outside that range, which is at least some support for the significance of man-made effects. However, consistency among the models is no guarantee that any of them are accurate.

In addition, that doesn't really address the second point, which I couldn't extract from the data. Were any of those simulations set up at a single point in time, based on conditions observed at that point and extrapolations from that data using assumptions similar to those being used today, then allowed to run for a century?

I get the impression that the methodology was the opposite. On a much shorter timeline - like year by year - it looks like they took a look at all observed data and fed it into the model to get an impact assessment for the various observatios for that year. Thus impacts like Mt. Pinatubo are explained. Of course, a prediction started in 1900 would not (likely) have predicted a Mt. Pinatubo event.

This is better than nothing at attempting to validate the models. It shows (or doesn't, as applicable) that the temperature variations follow observed variations in key predictive variables such as solar and volcanic activity. If we knew now what those variations would be from now until 2100, we'd have some confidence that the models could provide a correlation between that future volcanic activity and future temperature, for example.

But we don't know how those factors will vary in the future.

I'm all for research, if it's honest, scientific inquiry without agenda. But the burden of proof should be squarely on those who advocate massive economic harm to the US particularly and without corresponding impact to our competitors (i.e. Kyoto Protocol), or who advocate massive centralization of power into the hands of unelected bureaucrats (i.e. Kyoto Protocol). These data do not provide that proof.

One interesting note: A single volcanic eruption (Mt. Pinatubo) is reported to have had twice the effect of all man-made warming combined for that corresponding year. Perhaps the solution to global warming is to detonate a few nuclear devices in not-quite-active volcanos. One or two of those a year and the man-made effects are more than cancelled out. That 'solution' is offered at least partially in jest (the science might work, but I know the politics would not) but the key issue is that we haven't really considered how to address the issue in a positive way. The 'limits to growth' philosophy just rolls over and dies, accepting that for all future generations, the standard of living will be less then ours. We should be looking at data like this for solutions that increase the overall wealth of society, not eliminate it.

I'd like to see, for example, an analysis of what hydrogen as a fuel, or nuclear power does to these data. Both can reduce the amount of carbon entering the atmosphere, but they increase water vapor (which is a worse greenhouse gas) though not on a pound-for-pound basis. Are those good or bad ideas for power, as a net effect? That's a matter for data, not the opinion of mankind-is-bad-and-manmade-carbon-will-kill-us-all zealots.

Interesting article, and thanks for pointing it out. When I get home this evening, I'll post you another note with a link to a paper identifying some compensatory atmospheric effects that tend to reduce global warming (and may have been the reason the predicted ice age didn't show up). Compensatory factors may be buried in the simulations reference in your article, but I couldn't see any explicit recognition of them.