Wednesday, November 08, 2006

Jan Veizen’s cosmic ray climatology

Last week we had Jan Veizen, a long-ago ANU alumnus, come back to give a talk about how hew thinks that Phanerozoic climate is primarily governed by cosmic rays. I am not a climatologist, so I can’t really get into the nitty-gritty details. Additionally, it was a fairly qualitative presentation, so this reflection is similarly mushy. There was a fast one that he tried to pull, and there was an abundance of sketchy science that is worth repeating.

First of all, Professor Veizen did repeatedly state that he was not in favor of polluting, and that he thought it was a generally bad idea. But he also concluded that CO2 climate sensitivity was overestimated because it did not create a water vapor feedback. The trouble is, even if we accept his overall model, the application of it to anthropogenic climate change is not necessarily appropriate.

Professor Veizen’s thesis basically goes like this:
Cosmic rays encourage cloud formation, by increasing nucleation.
Cloud formation effects the hydrological cycle, by changing albedo and precipitation.
The hydrological cycle drives CO2 levels, by regulating CO2 sequestration by land plants.

So CO2, humidity, and albedo are all increasingly direct feedbacks from cosmic ray abundance.

Even if we accept this model for the Phanerozoic, it still doesn’t tell us anything about 21st century climate change, for the following reason:
It assumes that CO2 is a passive positive feedback- an amplifier of humidity and/or precipitation.
The problem is that we know the current CO2 increase is NOT due to a feedback; it is due to fossil fuel burning. Therefore, we are already operating outside of the Veizen model. Since his model assumes that CO2 reacts passively, this model cannot be used for predicting the results of a forced change on CO2 content, which is our current situation.

As for the presentation, the only really deceptive bit was when he said that the 1 W/m2 dry CO2 forcing was equivalent to the 1 W/m2 variation in solar irradiance. As realclimate people have occasionally pointed out, the CO2 acts on surface area, which for a sphere is 4 times the cross section (4pi*r2 vs pi*r2). So using W/m2 on effects that operate over different numbers of square meters is dishonest. Interestingly, none of the 60 or so professional geoscientists in the audience called him on this.

The main criticism of the talk from our paleoclimate people was that he showed a lot of correlations of normalized, detrended data where his residual wiggles were way smaller than the magnitude of the detrending. For example, he showed a 1 per mil d18O curve without mentioning that the detrending removed 10 per mil of change.

My main problem was the lack of quantification of processes. For example, he was very coy about how much longer (or shorter) it took a cloud to nucleate in a high radiation vs. low radiation environment. Is it seconds, or hours, or years?

Furthermore, he did not discuss the magnitudes of changes that he claimed to observe on different timescales. For example, his observations on the 100 Ma timescale had 10 degree changes, while his observations on the century and decadal timescale were a degree or less. He correlated these with cosmic ray wiggles, but the magnitude of those cosmic wiggles on the various timescales was not linked to the magnitude of the climactic changes.

Finally, he failed to address any of the non-CO2 alternative hypotheses generally invoked to explain long term changes in paleoclimate. In order to get a new theory accepted, it is generally necessary to find a flaw in the current state of knowledge. He never mentioned the effect of traditional long timescale climate drivers like tectonics.

It was an interesting theory, and one worth trying to quantify, but running to the New York Times might have been a little bit premature.

No comments: