Climate related uncertainties in integrated assessment models (IAMs)

Uncertainties resulting from the estimates of climate sensitivities are usually used to represent climate related uncertainties in integrated assessment models (IAMs). However, climate related uncertainties are also significantly affected by present understanding of carbon cycles, including biosphere carbon cycle and ocean carbon cycle, and climate change impacts induced by land use change, pollutants and aerosols, and various natural sources like solar irradiance and volcanic aerosols. In this study, we try to quantify the differences between climate related uncertainties and uncertainties derived from only climate sensitivities, using an integrated assessment model, the Simple Climate Model for OPTimization (SCM4OPT). We update the simpleclimate module to be consistent with the CMIP5 simulations, and use the inter-model differences to represent the climate-related uncertainties. We define two scenarios, 1) a scenario with climate related uncertainties reflecting CMIP5 simulations, 2) a scenario with climate sensitivity likely range of (1.5,4.5), which is suggested by the IPCC Fifth Assessment Report (AR5), and other parameters use default estimates. The results show that, the climate related uncertainties of the estimates of temperature increase have a notable wider range than those uncertainties using only climate sensitivities, due to uncertainties existing in land use change, pollutants and aerosols, and various natural sources. The differences of the mean values between the two scenarios are large in early period, particularly, the mean values derived from climate related uncertainties are lower than those climate sensitivity uncertainties, though become narrower in recent period. It is because the fossil fuel CO₂emissions in early period are relatively small, and uncertainties induced by other sources dominate the uncertainty estimates, especially for pollutants and aerosols, some of these emissions cause cooling effects to climate change and lower the mean values of temperature estimates. Assessing uncertainties regarding climate change in IAMs needs to distinguish the contributing factor and observe how it will change the final results.

Figure 1: A comparison of temperature increases above pre-industrial level between climate related uncertainties and climate sensitivity uncertainties.The ranges indicate the 17th and 83th percentiles; and the solid-colored line indicates the 50th percentile.