by Makari Krause
Sea level rise is one of the most concerning facets of climate change, Pycroft et al. (2014), examine its effects on the social cost of carbon. Rapid ice sheet melting or collapse which would cause rapid, significant sea level rise is hard to incorporate into climate models because it is difficult to gauge the likelihood of such events. This difficulty stems from the fact that the underlying processes are, themselves, hard to model. Pycoroft et al. use an integrated assessment model to examine the impacts and costs of large-scale damage associated with sea level rise. In their model they adjust the physical aspects that contribute to sea level rise and the economic consequences of those aspects. Their model shows that incorporating extreme sea level rise significantly increases the social cost of carbon.
Pycroft et al. begin with a review of economic and scientific literature surrounding sea level rise. There is continuing uncertainty about the extent of temperature rise and even more uncertainty about how changing temperatures will affect sea levels. Through looking at the literature Pycroft et al. determine that sea level rise of more than 1m would require a significant increase in ice flow dynamic and that sea level rise over 2 m is highly unlikely. For this reason their 90% confidence interval is between 1 and 2 m and sea level rise parameters have been adjusted accordingly for global temperatures. (The standard model uses values of 0.4 and 1 m)
The integrated assessment model used in this study, PAGE09, separates impacts induced by climate change into four categories: economic, non-economic, sea-level rise, and discontinuity. Economic and non-economic damages depend on temperature, sea-level rise damages are explicitly modeled through sea level rise, and discontinuity damages capture everything that doesn’t fall into the other categories, such as catastrophic events like ice sheet collapse. Pycroft et al. calculate the social cost of carbon by first running the model as normal and then running it with a slight decrease in the amount of CO2 in the first year. They then can calculate the marginal impact per ton of carbon dioxide and from that the cost.
Pycroft et al. replace discontinuity damages in the original PAGE09 model with tails, allowing for the possibility that economic damage estimates are much higher than the central estimates. Under the original PAGE09 model sea level rises by an average of 0.64 m by 2100 and 1.61 m by 2200. When tails are added to the model, sea level rise in 2200 increases by 8% (thin tail), 9% (intermediate tail), and 12% (fat tail). Under the revised sea level rise parameters mean sea level rise increases from 0.64 m to 1.43 m in 2100 and from 1.61 to 3.98 m in 2200. The revised parameters clearly increase the response of sea level to temperature. Translated into mean dollar values for the cost of carbon, the thin tailed parameters yield $135/ton, the normal tailed yield $147/ton, and the fat tailed yield $218/ton. When revised sea-level parameters are introduced these values change to $149/ton, 161$/ton, and $218/ton respectively.
The revised sea level parameters add between 10 and $14/ton of CO2 to the mean values for the social cost of carbon dioxide. This estimate increases considerably when the revised sea level parameters are combined with the fat tailed distribution. In this case the 95th percentile estimate increases from $54/ton to $893/ton. This shows that the combination of revised parameters and fat tails gives a lot more weight to damages caused by rare, extreme climate events. More research needs to be done to determine the accuracy of the different tail sizes and determine which variant of the model will most reflect reality.
Pycroft, J., Vergano, L., & Hope, C., 2014. The economic impact of extreme sea-level rise: Ice sheet vulnerability and the social cost of carbon dioxide. Global environmental change 24, 99-107. http://bit.ly/1MejAyK