Spread in climate policy scenarios unravelled.

Analysis of climate policy scenarios has become an important tool for identifying mitigation strategies, as shown in the latest Intergovernmental Panel on Climate Change Working Group III report1. The key outcomes of these scenarios differ substantially not only because of model and climate target differences but also because of different assumptions on behavioural, technological and socio-economic developments2-4. A comprehensive attribution of the spread in climate policy scenarios helps policymakers, stakeholders and scientists to cope with large uncertainties in this field. Here we attribute this spread to the underlying drivers using Sobol decomposition5, yielding the importance of each driver for scenario outcomes. As expected, the climate target explains most of the spread in greenhouse gas emissions, total and sectoral fossil fuel use, total renewable energy and total carbon capture and storage in electricity generation. Unexpectedly, model differences drive variation of most other scenario outcomes, for example, in individual renewable and carbon capture and storage technologies, and energy in demand sectors, reflecting intrinsic uncertainties about long-term developments and the range of possible mitigation strategies. Only a few scenario outcomes, such as hydrogen use, are driven by other scenario assumptions, reflecting the need for more scenario differentiation. This attribution analysis distinguishes areas of consensus as well as strong model dependency, providing a crucial step in correctly interpreting scenario results for robust decision-making.

On the optimality of 2°C targets and a decomposition of uncertainty.

Determining international climate mitigation response strategies is a complex task. Integrated Assessment Models support this process by analysing the interplay of the most relevant factors, including socio-economic developments, climate system uncertainty, damage estimates, mitigation costs and discount rates. Here, we develop a meta-model that disentangles the uncertainties of these factors using full literature ranges. This model allows comparing insights of the cost-minimising and cost-benefit modelling communities. Typically, mitigation scenarios focus on minimum-cost pathways achieving the Paris Agreement without accounting for damages; our analysis shows doing so could double the initial carbon price. In a full cost-benefit setting, we show that the optimal temperature target does not exceed 2.5 °C when considering medium damages and low discount rates, even with high mitigation costs. With low mitigation costs, optimal temperature change drops to 1.5 °C or less. The most important factor determining the optimal temperature is the damage function, accounting for 50% of the uncertainty.