Abstract: Addressing climate change is ultimately a challenge of risk management, which requires an understanding of the likelihood of potential outcomes. We provide integrated, probabilistic socio-economic and climate projections obtained using updated estimates of probability distributions for key parameters in both the human and Earth system components of the MIT Integrated Global System Model (IGSM).

The Reference scenario results in median end-of-century warming of 3.5^oC and a 90% range of 2.8-4.3^oC, which is lower than the median of 5.7^oC from a prior study using a previous version of the IGSM. About 0.5^oC of the difference is due to updated estimates in the human system and the rest of the difference is explained by changes in Earth system estimates.

Our results show that climate policy lowers the upper tail of temperature change distributions more than the median, and that even relatively modest policies can significantly reduce the likelihood of high global temperature outcomes. Human system uncertainties contribute more to uncertainty in projected CO₂ concentrations and total radiative forcing, while Earth system uncertainties have the greatest influence on temperature and precipitation. Including additional uncertain inputs does not automatically increase the outcome range because uncertainties can offset one another.

Results also show how policy costs can vary greatly among regions.  As we improve understanding of underlying technology and economic factors as well as Earth system response to human forcing, further updating of these estimates of uncertainty can make an important contribution to decision-making about mitigation and adaptation.