Achieving China's carbon neutrality: Predicting driving factors of CO2 emission by artificial neural network

Abstract

After China announced its commitment to peak carbon emissions by 2030 and carbon neutrality around 2060, concerns arose over its CO 2 emission paths. The feasibility of net-zero emission in China has been assessed, yet how emission-driving factors may behave throughout different paths remains explored. Based on the Logarithmic Mean Divisia Index decomposition model, the present study examined the driving factors from 2005 to 2016 and applied the artificial neural network for factor prediction from 2016 to 2060. Energy efficiency plays a vital role in reducing CO 2 emissions by 4.90 Gigatons (Gt), while economic growth, as the decisive promoting factor, encourages emissions by 8.58 Gt. In the pre-peak phase 2016–2030, energy intensity is the leading emission counterforce decreasing CO 2 emissions by up to a maximum of 11.3 Gt before sliding to the second position after 2030. During the period of 2030–2060, industrial structure exerts a significant negative effect eliminating up to 6.78–6.87 Gt of CO 2 emissions, meanwhile showing an accelerated increase (0.167–0.172 Gt/yr in 2030–2050, and 0.333–0.352 Gt/yr in 2050–2060). From an economic perspective, negative emission technology shows little advantage before 2030, but thereafter offers a lower-cost emission reduction until 2060. Sustainable scenarios' cumulative emissions are totally 420.1–506.3 Gt CO 2 between 2005 and 2060, with emission peaks at 9.46–11.58 Gt around 2030. Carbon sinks & carbon capture and storage (CCS) and BECCS (biomass energy and CCS) are preferable for China to accomplish carbon neutrality, contributing 1.33–5.09 Gt CO 2 in 2060. Projection of CO 2 emission drivers could highlight the sensitive variables during emission mitigation and neutralization, and benefit global green development.