I have a climate change / climate modelling related question for you:
I have read that climate records that extend back further than 4 million years are irrelevant. Naturally this is coming from the AGW camp who want to ignore that 250 million years ago CO2 levels appear to have been very high.
The basis of this argument (and it does seem plausible) is that 4 million years ago the North and South American continents combined creating a distinct Atlantic and Pacific Ocean and thus dramatically altering global weather patters and therefore the biological carbon cycle.
Then the argument goes on to say that ice core records (only 2 million years worth, but extrapolated to 4 million) do not indicate that the earth has been outside of the 180 - 280 ppm envelope in that time period. Thus to be outside of that envelope is very dangerous.
There are a lot of extrapolations and estimates in this, but the basic premise of the separation of Atlantic and Pacific and limiting back tracking to 4 million years does not seem obsurd to me.
What are your thoughts?
I'm now moving away from my expertise, so take what I say with a grain of salt. I'm a modeler, not a climatologist (of course, a lot of the AGW folks aren't climatologists either).
This particular argument is again about parameters. The AGW folks have, with this argument, just added one or more additional parameters to their models. Call it the "Two Ocean" parameter. This parameter explains away historic data that is inconsistent with the models the same way that the IPCC's "masking factors" explain away 1998-2008.
This AGW's could be right on this--I don't have the expertise to know. What I do know is that, when theories start piling more and more explanatory parameters to make the data fit them, I get mighty suspicious that something other than science is going on.
Perhaps CO2 is historically high, if you ignore the one-ocean time periods. To figure out if this is a scary thing and if we can do anything about it, we have to answer two questions with real evidence.
The questions are: (1) what does high CO2 mean? and (2) are SUV's the cause?
Number 1 that takes us back to my original point. This whole edifice depends on how much you believe the models. Maybe they are wonderful models. But we don't have any empirical evidence worth spitting at that they are, especially when compared with other, simpler models.
Number 2 is even a bigger problem for the AGW modelers. Even the AGW folks admit that mankind is only a tiny contributor to the current level of CO2 (the problem is that if you add up human injected CO2, even with typical green exaggeration, you don't get all that much). So, the argument goes, "Yes, it's only a little but feedback makes man's CO2 the critical component in pushing us past a tipping point."
That means their models are so sensitive that they can detect not just gross CO2 effects, but discern the effect of the tiny percentage change caused by mankind's contribution from the effect of all that other CO2. Not only that, they are so sensitive they can tell when the "tipping point" would be reached and that we would not reach it unless we were driving SUV's.
But the dreaded "tipping point" is just a phase change of some kind or a move in a complex system to a different strange attractor. Such changes are notoriously hard to model because they are complex and depend on the ol' butterfly wings. For example, how can we measure when a hillside of snow turns into an avalanche (there's a lot of work on this issue)? That's a "tipping point" and predicting them is really, really difficult. Hillsides of snow are simple, transparent systems compared to climate. And, we have seen avalanches over and over. We actually know that's a tipping point that exists.
So now their models are so sensitive they can tell that the man-contributed CO2 puts us over a tipping point noone has ever seen or measured. And if their models are that sensitive, why don't we see forecasts from twenty years ago tracking temperature changes accurately?
Either their models are not that sensitive (likely) and/or climate is really noisy (also likely) and/or the bottom up modeling of climate is sufficiently complex that we do not have the physics or the computers to do so yet (typical complex system).
So now we're now back to the beginning. They are asking for control of the world economy to confront a looming catastrophe. But the system they are modeling is too noisy and complex and their models too coarse for us to know if their models are any good. And the evidence of the looming catastrophe exists only in their models. In fact, the evidence of the existence of a 'tipping point,' unlike avalanches, exists only in their models.
Most of them are aware of the paucity of model validation. And that's why you hear about the "precautionary principle." Basically, that means do what Greens want until someone proves they are wrong. But now we are out of the realm of science.
We really have NO idea what the effect of changing an input into a complex system like climate will be. It's perfectly possible that mankind's CO2 is holding off an ice age. Reducing our CO2 could be catastrophic because we would reach an "ice-age tipping point." I have no evidence validating that model. But the notion that there should be some presumption in favor of what Greens want doesn't strike me as having much validity.
BTW, this tipping point stuff is actually an admission by the AGW folks that their models are exponential and that we are still in the linear section of the exponential model. In other words, temperatures are going up in a more or less straight line (the last ten years excepted) today. But when the "tipping point" hits, they'll start going up at an ever increasing rate. So the tipping point argument admits my concern that their validation techniques are only validating the LINEAR portion of their models. Thus, the fact that an exponential model caused by positive feedback is the correct model is completely unvalidated.