Empirically Constrained Climate Sensitivity and the Social Cost of Carbon Dioxide

The authors applied the 2015 Lewis and Curry equilibrium climate sensitivity (ECS) distribution to the widely-used DICE and FUND Integrated Assessment Models. Previously the developers of these models (and others) have relied on model-simulated distribution of ECS values. using the empirical ECS distribution, the estimated SCC drops substantially in both the DICE and FUND models, and in the latter there is a large probability it is no longer even positive. The FUND model calculates that emissions in 2010 using a 5% discount rate have a SCC of -$0.65/tCO2, that is, emissions are beneficial. The ECS used however, it too high because it fails to account for urban contamination of the surface temperature record nor natural long-term climate change.

This paper was authored by Kevin Dayaratna of the Heritage Foundation in Washington DC, Ross McKitrick of the University of Guelph and David Kreutzer of the EPA. See the paper here.

The 2015 Lewis and Curry ECS probability distribution has a best estimate of 1.64 °C, with a likely (17-83%) range of 1.25 to 2.45 °C. This estimate used out-dated aerosol estimates, did not account of the effects of urban warming and bad siting of surface temperature stations, nor the natural millennium cycle that caused a natural warming from the Little Ice Age. Correcting for these issues would reduce the ESC best estimate to about 1.0 °C. This would reduce the calculated SCC substantially lower than what is presented in the paper.

©2002-2022 Friends of Science Society