Most of the integrated assessment modelling literature focuses on cost-effective pathways towards given temperature goals. Conversely, using seven diverse integrated assessment models, we project global energy CO₂ emissions trajectories on the basis of near-term mitigation efforts and two assumptions on how these efforts continue post-2030. Despite finding a wide range of emissions by 2050, nearly all the scenarios have median warming of less than 3 °C in 2100. However, the most optimistic scenario is still insufficient to limit global warming to 2 °C. We furthermore highlight key modelling choices inherent to projecting where emissions are headed. First, emissions are more sensitive to the choice of integrated assessment model than to the assumed mitigation effort, highlighting the importance of heterogeneous model intercomparisons. Differences across models reflect diversity in baseline assumptions and impacts of near-term mitigation efforts. Second, the common practice of using economy-wide carbon prices to represent policy exaggerates carbon capture and storage use compared with explicitly modelling policies.

大多数综合评估建模文献都侧重于实现给定温度目标的具有成本效益的途径。相反，我们使用七种不同的综合评估模型，预测全球能源 CO ₂基于近期缓解努力的排放轨迹以及关于这些努力如何在 2030 年后继续进行的两个假设。尽管发现到 2050 年的排放范围很广，但几乎所有情景在 2100 年的平均升温幅度都低于 3°C。然而，最乐观的情景仍然不足以将全球变暖限制在 2°C。我们进一步强调了预测排放走向所固有的关键建模选择。首先，排放对综合评估模型的选择比对假设的缓解努力更敏感，突出了异构模型比对的重要性。模型之间的差异反映了基线假设的多样性和近期缓解工作的影响。第二，