Climate module disparities explain inconsistent estimates of the social cost of carbon in integrated assessment models

Abstract

•Uncertainties in SCC estimates are attributed to modeling choices in IAMs •The differences in climate modules explain 60%–95% of SCC variations •Non-linear climate system responses result in higher SCC estimates •Unified climate models substantially improve the consistency of SCC estimates The social cost of carbon (SCC) determines the quantifiable cost (in monetary value) of emitting CO2 and has been employed by policy makers to evaluate the benefits of climate mitigation against the costs of cutting emissions. SCC is generally estimated using integrated assessment models (IAMs), primarily the DICE, FUND, and PAGE models. However, these models often report different SCC estimates. Inconsistent SCCs can lead to over-/underestimation of the costs and benefits of emission reduction, further undermining the effectiveness of climate policies. The reasons behind such inconsistent SCCs, however, remain unclear. Here, we quantitatively decompose two core components of IAMs, climate modules and damage functions, and find that the core reason for SCC disparities lies in differences in climate modules. These differences explain 60%–95% of SCC variations. SCC estimates are more consistent when DICE/FUND/PAGE IAMs are coupled with a reduced-complexity climate module. Our study highlights the importance of considering a common modeling framework by synthesizing state-of-the-art information on climate science to provide more consistent science-based advice to policy makers. The social cost of carbon (SCC) determines the quantifiable cost (in monetary value) of emitting CO2 and has been employed by policy makers to evaluate the benefits of climate mitigation against the costs of cutting emissions. SCC is generally estimated using integrated assessment models (IAMs), primarily the DICE, FUND, and PAGE models. However, these models often report different SCC estimates. Inconsistent SCCs can lead to over-/underestimation of the costs and benefits of emission reduction, further undermining the effectiveness of climate policies. The reasons behind such inconsistent SCCs, however, remain unclear. Here, we quantitatively decompose two core components of IAMs, climate modules and damage functions, and find that the core reason for SCC disparities lies in differences in climate modules. These differences explain 60%–95% of SCC variations. SCC estimates are more consistent when DICE/FUND/PAGE IAMs are coupled with a reduced-complexity climate module. Our study highlights the importance of considering a common modeling framework by synthesizing state-of-the-art information on climate science to provide more consistent science-based advice to policy makers.