Estimating the economic cost of sea-level rise

Abstract

(cont.) In the case of a classical linear sea-level rise of one meter per century, the use of DIVA generally decreases the protection fraction of the coastline, and results in a smaller protection cost because of high spatial concentration of capital. As in past studies, wetland loss continues to be dominant for most regions, and the total cost does not decline appreciably where wetland loss remains about the same. The total cost for the United States is about $320 billion (in 1995 U.S. dollars), an estimate comparable with other studies. Nevertheless, capital loss and protection cost may not be negligible for developing countries, in light of their small gross domestic product. Using realistic sea-level rise scenarios based on the Integrated Global System Model (IGSM) simulations substantially reduce the cost of sea-level rise for two reasons: a smaller rise of sea level in 2100 and a nonlinear form of the path of sea-level rise. As in many of the past studies, the thesis employs conventional but rather unrealistic assumptions: perfect information about future sea-level rise and neglect of the stochastic nature of storm surges. The author suggests that future work should tackle uncertain and stochastic sea-level rise damages.

To improve the estimate of economic costs of future sea-level rise associated with global climate change, the thesis generalizes the sea-level rise cost function originally proposed by Fankhauser, and applies it to a new database on coastal vulnerability, Dynamic Interactive Vulnerability Assessment (DIVA). With the new cost function, a new estimate of the cost present values over the 21st century is produced. An analytic expression for the generalized sea-level rise cost function is obtained to explore the effect of various spatial distributions of capital and nonlinear sea-level rise scenarios. With its high spatial resolution, DIVA shows that capital is usually highly spatially concentrated along a nation's coastline, and that previous studies, which assumed linear marginal capital loss for lack of this information, probably overestimated the fraction of a nation's coastline to be protected and protection cost. In addition, the new function can treat a sea-level rise that is nonlinear in time. As a nonlinear sea-level rise causes more costs in the future than an equivalent linear sea-level rise scenario, using the new equation with a nonlinear scenario also reduces the estimated damage and protection fraction through discounting of the costs in later periods. Numerical calculations are performed, applying the cost function to DIVA and socio-economic scenarios from the MIT Emissions Prediction and Policy Analysis (EPPA) model.