An alternative approach to capture uncertainties embedded in the estimation of social cost of carbon

The continuous increase in global surface temperature, which has been triggered by the increase in atmospheric CO2 concentration from anthropogenic activity, results in damage not only in the short term but also in the long term. This damage, called the social cost of carbon (SCC), has been widely estimated using integrated assessment models (IAMs). A large range of estimated SCC values have been observed because of uncertainties in IAMs' parameters. This study provides a comprehensive review of these uncertainties after dividing IAM modules into four categories: climate sensitivity, damage function, discount rate, and regional–sectoral validation. The review was conducted by comparing key ideas considered by various IAMs: socioeconomic conditions in relation to projected CO2 emissions, estimation of the atmospheric concentration of CO2, estimation of total radiative forcing, parameters of the temperature function, parameters of the damage function, and discount rate value. In addition, this study presents an alternative approach to capture the uncertainties embedded in the SCC estimation, using a machine learning approach. This enables a probabilistic evaluation of a specific level of SCC and improves our comprehension of the implication of the calculated SCC using IAMs. This alternative approach provides a basis for further study of SCC.