Carbon Neutrality

As Paris Agreement entered into force, countries take further steps to limit Greenhouse Gases emissions. China pledges to peak its CO₂ emissions by 2030 and achieves net-zero no later than 2060, implying rapid and dramatic decarbonization actions across all sectors. Transportation accounts for about 25% of global carbon emissions and 10% of China’s emission.

Modeling scenarios of deep decarbonization in the transportation sector, as stipulated by large-scale EV adoption, requires a combined assessment of technical and behavioral factors. Here, we develop a machine-learning-based method (RL-GCAM) to model the optimal technical and behavioral pathways to directly achieve a specific share of EV in China. In contrast with traditional stylized perturbation scenarios, we automate and improve the parameterization of technical and behavioral factors within a well-established global integrated assessment model through a reinforcement learning framework.