The Oz and Andrew Bolt have a tag team attack on me today (Google it if you want). Most of it consists of quotations, with lots of ellipses, that are meant to show me as a dangerous radical. I can’t say I’m too upset by that – from their perspective, it’s a fair assessment. But Bolt also repeats his claim that I made a factor-of-5 error in my estimate of the impact of Australia’s current 2020 target on global temperatures.
This is a striking piece of chutzpah, given that this estimate was made in the process of correcting a calculation by Bolt, which was out by two orders of magnitude. But it has finally provoked me to clear up some of the confusion on this. The starting point was this post by Bolt who used a calculation by Damon Matthews that each tonne of CO2 emitted into the atmosphere changes the equilibrium temperature by 0.000 000 000 0015 degrees, that is 1.5*10^-12 in scientific notation. Noting that the carbon price is expected to reduce emissions by 160 million tonnes per year by 2020, Bolt made the calculation that the emissions avoided in the year 2020 will reduce equilibrium temperature by 2.4*10^-4 or 0.00024 degrees, and treats this as an estimate of the impact of the policy.
This is an amazing howler on Bolt’s part. He’s only counted one year of emissions reductions for a policy that is supposed to permanently reduce emissions. I made the very quick calculation that, if the policy stays in place until 2100 and that the 2020 reduction in emissions was maintained over this period, the number used by Bolt would imply a reduction of 0.02 degrees. I did another rough calculation that came out the same way.
Bolt came back with a lower estimate by Roger Jones, who suggest that the policy would reduce temperature by only 0.004 degrees, lower by a factor of 5 than my estimate, but higher by a factor of 20 than Bolt’s silly calculation.
At this point I slipped up. As a result of a misunderstood conversation with Roger, I gave an incorrect explanation for the discrepancy. Roger subsequently advised that he had made his calculation using a standard modelling tool called MAGICC. I finally got around to downloading MAGICC, and trying it out, so I can now give an explanation for why our estimates differ. There are three main points
(1) The most important factor is that we are estimating two different things. MAGICC produces estimates of the temperature change by 2100, but the atmosphere takes a long time to reach equilibrium. For reductions in CO2 emissions spread out over the rest of this century, the change by 2100 is only about half the long run equilibrium change.
(2) Estimates of the sensitivity of the global climate to changes in CO2 concentrations vary. The most common measure is the equilibrium temperature change for a doubling in atmospheric CO2. Until recently MAGICC used 2.6 degrees as the default, on the low side of most estimates. I used 3.5, which gives a value around 30 per cent higher
(3) Finally, while it’s obviously silly to assume, like Bolt, that the policy is in effect for only one year, it’s not entirely clear how we should project its impact into the future. That depends on baseline projections of emissions from which to calculate percentage reductions. My simple estimate takes a constant reduction over 80 years, which is probably a bit on the high side. If you assumed that emissions were going to decline anyway over the second half of this century, the effect of the policy would be reduced, perhaps by half.
Those three factors, taken together, would account for the discrepancy in the two estimates. I don’t claim that I’ve got them exactly right and there may be points I’ve missed. But for someone like Bolt to pontificate on a subject like this, when he is incapable of avoiding or correcting even the most absurd errors, brings to mind Matthew 7:3-5.
Read more…
Recent Comments