Friday, December 23, 2022

Inflating the Cost of Carbon 2

In my previous post I sketched my criticisms of Rennert et. al. 2017, a recent Nature article that attempts to calculate the total cost imposed by an additional ton of CO2. I have now written an article on the subject that I plan to submit to the EPA in response to their request for comments on Rennert and have webbed the current, I think almost final, draft. 

Comments welcome.


SB said...

"Since their calculations were based on national figures they ignored differing
incomes within nations, hence probably underestimated the effect."

It's not obvious to me why this would be an underestimate. Is it because the effect of income on temperature-related mortality is sublinear -- adding a dollar to a poor person's income makes more difference than adding a dollar to a rich person's income -- and therefore an income gain spread evenly across an income-diverse nation makes more difference than the same income gain applied to an income-homogeneous nation? If so, are you sure that effect isn't already accounted for in Bressler et al?

"Increases due to anthropogenic climate change, as projected in the IPCC reports, are greater in winter than summer. The data in the articles used by Cromar et. al. to deduce temperature-related mortality are from temperature variation in recent years, over either time or space, so have no reason to reproduce that pattern. If we assume that the increases in Cromar et. al. are uniform..."

Why do data from "recent years" "have no reason to reproduce that pattern"? Is warming in "recent years" qualitatively different from that in IPCC projections? Is it possible to look at the temperature variations used in Cromar et al and check whether they're uniform over the year, rather than speculating?

"Long et. al., however, found increases of 12%, 13%, and 14% for rice, wheat, and soybeans respectively for an increase to 550 ppm from the ambient concentration, which implies an increase of about 17.5% for a doubling.16 Kimball 2016,17 a survey of FACE studies, found that “Yields of 3 grain crops were increased on average about 19%” by increasing CO2 from 353 ppm to 550, which implies a 23% increase for a doubling."

I'm guessing that your "which implies an increase of X for a doubling" calculations are linear extrapolations. Linearity is probably not a bad assumption on this scale. Still, it might be worth adding an "assuming linearity".

On the other hand, Taylor and Schlenker, starting from natural variation rather than artificial CO2 augmentation, quotes figures "per added ppm", and assuming linearity from there to "doubling" is much more dubious, and I presume that's why you didn't extrapolate any specific numbers from it.

The comments on the impossibility of projecting technological change over 300 years seem spot-on :-)

David Friedman said...

"Why do data from "recent years" "have no reason to reproduce that pattern"?"
Cromar isn't generating his own numbers, he is taking a bunch of different articles that contain information linking temperature variation to mortality and trying to combine their results.
So there isn't a webbed database showing temperatures used by Cromar that I could look at.

None of them, so far as I know, were comparing pairs of years far enough apart so that the main cause of the difference was climate change. Some of them compared days of different temperature in the same year and looked at deaths in the next few days. Some compared different localities in the same year.

Carleton et. al., in contrast, used a massive database of detailed temperature and mortality data to get their values for the effect on mortality of lots of things including temperature.

David Friedman said...

"I'm guessing that your "which implies an increase of X for a doubling" calculations are linear extrapolations."
I thought I had a footnote citing a source for that, but it seems to have gotten lost in the process of editing, so I will have to find it and put it back in. Thanks.

David Friedman said...

Taylor and Schlenker were looking at a small range of CO2 variation, since their data was what happened naturally, not an artificial experiment. I don't think extrapolating from the effect of increasing CO2 by 10 ppm to extrapolating it by 500ppm is very plausible. For one thing, they are combining the effect on yield directly and the effect via water requirements — that's why they get a substantial increase for maize, which is a C4 crop. The biggest increase you can get that way is to bring yield back up to the level with no water constraints.

What their results show, assuming they are right, is that the effect of small increases is substantially greater than the FACE numbers imply.

David Friedman said...

I found the footnote, which got lost between drafts:

“Such an adjustment is justified because to a first approximation growth responses by plants to elevated CO2 are generally linear between 300 and 900 ppm” (Kimball 2016).

SB said...

Peripherally related, some of the harder-to-predict effects of climate change are through an expected increase in zoonotic diseases such as AIDS, SARS, and COVID. See this post and its linked sequels, which summarize and cite a lot of research papers. Some of the effects described in the research are specific to global warming, while others would occur in any climate change that caused widespread animal migration.