Mitigation of China's carbon neutrality to global warming

The study addresses how much China’s pledged carbon neutrality (peaking CO2 before 2030 and reaching net-zero before 2060) could mitigate future global warming. Contextually, anthropogenic CO2 emissions are the dominant driver of observed warming since the preindustrial era, and without strong mitigation, global temperatures are likely to exceed 1.5 °C in the near future and reach 3–5 °C by 2100. Over 120 countries have pledged carbon neutrality following the Paris Agreement, highlighting the importance of quantifying individual national contributions to future warming mitigation to inform global stocktake and equitable climate responsibility. The specific research objective is to quantify the avoided global mean surface temperature increase attributable solely to China’s carbon neutrality pathway and to the combined pathway including associated reductions in CH4 and N2O, using a fully coupled Earth system model across multiple shared socioeconomic pathways (SSPs). This fills a gap left by earlier simplified-model estimates that suggested a 0.16–0.21 °C avoided warming by 2100 from China’s pledge but lacked comprehensive Earth system coupling.

The paper builds on IPCC assessments of historical and projected warming and on literature quantifying nations’ historical contributions to warming and global carbon budgets. It references analyses indicating the need for net-zero CO2 around mid-century to limit warming to 1.5 °C, and notes that many countries have set neutrality pledges. Prior to this work, a simplified climate model estimated China’s carbon neutrality could avoid 0.16–0.21 °C of warming by 2100. The paper also references work on SSP scenarios, critiques of high-end scenarios, internal variability in CESM, regional transient responses to cumulative CO2 emissions, and studies on aerosol forcing and near-term climate forcers, as well as analyses of policy, decarbonization pathways, ecological restoration, and the distinction between carbon and greenhouse-gas neutrality.

• Model: NCAR Community Earth System Model (CESM) version 2.1.3, fully coupled (atmosphere, land, land ice, ocean, sea ice, river, coupler). Prior validation runs (1850/1900–2014) show high consistency with historical GMST datasets (cross-correlation R=0.78–0.99).
• Scenarios: Four CMIP6 SSPs—SSP1-2.6 (low forcing), SSP2-4.5 (intermediate), SSP3-7.0 (medium-high), SSP5-8.5 (high). Time horizons: near term (2021–2040), mid-term (2041–2060), long term (2081–2100), relative to 1850–1900 baseline.
• Experimental design: For each SSP, two simulations are compared: (1) the default CMIP6 SSP with IAM-derived anthropogenic surface CO2 emissions; (2) a modified scenario where anthropogenic surface CO2 emissions within China are replaced by values consistent with China’s carbon neutrality policy target (CNCN) from Tsinghua University (2021). Outside China, emissions remain as in the default SSP. After 2050, CNCN CO2 emissions are held at 2050 levels. Differences between paired runs quantify CNCN’s mitigation effect on GMST.
• CNCN CO2 pathway: Peaks at 10.5 GtCO2/yr in 2030 and declines to 1.2 GtCO2/yr by 2050 (∼89% reduction). Spatial Downscaling: Future grid-level anthropogenic CO2 emissions within China are scaled proportionally to their default SSP spatial pattern to match national totals specified by the CNCN pathway (proportional scaling assumption).
• Extended scenario (CNCNext): To include concomitant changes in non-CO2 GHGs associated with the CNCN pathway, CH4 and N2O emissions are projected using GCAM v5.4 under the constraint of the CNCN CO2 pathway and assuming equal marginal abatement costs between CO2 and non-CO2 GHGs. CESM prescribes globally uniform surface concentrations for CH4 and N2O; therefore, concentrations under default SSPs are taken from CMIP6, while for CNCN the study derives concentrations empirically by fitting cubic relationships (R2≈1) between cumulative emissions (2015–2100) and concentrations across SSPs, then applying CNCN cumulative emissions to obtain concentrations. CESM runs with combined variations in CO2, CH4, and N2O constitute CNCNext; differences versus default quantify CNCNext’s total contribution; differences versus CNCN isolate CH4+N2O contributions.
• Statistical analysis: Significance of GMST and grid-level surface temperature differences assessed with paired t-tests (p<0.01 unless noted). Internal variability of CESM estimated at ~0.06–0.09 °C and considered when interpreting near- and mid-term signals. Spatial significance assessed via stippling in maps.
• Outputs: Global mean surface temperature (GMST) anomalies relative to 1850–1900; spatial maps of temperature differences; fraction of global area with significant differences.

• Baseline warming (default SSPs, 2081–2100 vs. 1850–1900): SSP1-2.6: +1.7 °C; SSP2-4.5: +2.7 °C; SSP3-7.0: +3.4 °C; SSP5-8.5: +4.7 °C.
• CNCN (CO2-only) mitigation of GMST:
  • Near term: No significant differences across SSPs.
  • Mid-term: Significant reduction only under SSP5-8.5: −0.17 °C (±0.05 °C).
  • Long term: Significant reductions for SSP2-4.5: −0.14 °C (±0.07 °C); SSP3-7.0: −0.48 °C (±0.09 °C), about 14% of the warming increase; SSP5-8.5: −0.40 °C (±0.09 °C), about 9% of the warming increase. SSP1-2.6 not significant.
• CNCNext (CO2+CH4+N2O) mitigation of GMST:
  • Near term: No significant impacts for all SSPs.
  • Mid-term: Significant reductions for SSP2-4.5: −0.18 °C (±0.09 °C) and SSP5-8.5: −0.13 °C (±0.07 °C); other SSPs not significant.
  • Long term: Significant net declines for all SSPs: SSP1-2.6: −0.21 °C (±0.17 °C); SSP2-4.5: −0.32 °C (±0.13 °C); SSP3-7.0: −0.50 °C (±0.21 °C); SSP5-8.5: −0.39 °C (±0.17 °C). These represent roughly 11–12% of total warming.
• Spatial patterns (CNCN): Temperature differences (CNCN − default) span −1.84 to +1.76 °C. Near/mid-term significant differences cover only ~0.2–4% of grids. Long-term: Significant areas up to ~10% (SSP1-2.6), 9% (SSP2-4.5), and much larger for high-forcing scenarios—53% (SSP3-7.0) and 34% (SSP5-8.5). Under SSP3-7.0, avoided warming shows polar amplification (stronger in Arctic/high latitudes). Under SSP5-8.5, strong avoided warming occurs in eastern Greenland, Greenland Sea, large parts of Russia, and various land/ocean regions.
• Spatial patterns (CNCNext): Near-term responses minimal except localized warming in Barents Sea region under SSP1-2.6; mid-term avoided warming evident but limited area (3–10% of global area depending on SSP); long-term avoided warming covers 8%, 22%, 25%, and 36% of global area for SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5, respectively.
• Robustness: Long-term CNCN/CNCNext mitigations under SSP3-7.0 and SSP5-8.5 exceed CESM internal variability (0.06–0.09 °C).

The study provides a first estimate using a fully coupled Earth system model of the mitigation of global warming attributable to an individual country’s carbon neutrality pledge. It shows that China’s CO2-only neutrality (CNCN) significantly reduces long-term warming under higher-forcing scenarios (SSP3-7.0 and SSP5-8.5) by 0.48 °C and 0.40 °C, representing 14% and 9% of respective warming increases, with limited near- and mid-term effects due to internal variability and cumulative historical emissions. Incorporating CH4 and N2O reductions associated with the CO2 pathway (CNCNext) increases mitigation to 0.21–0.50 °C across SSPs by the late century and yields statistically significant mid-term benefits under SSP2-4.5 and SSP5-8.5, underscoring the added value of multi-GHG strategies. Spatially, mitigation exhibits strong regional heterogeneity, with pronounced avoided warming in high latitudes, especially the Arctic (polar amplification), and varying regional responses across oceans and continents. These findings demonstrate that national net-zero pledges can make measurable contributions to limiting global warming, particularly under higher-emission pathways, and provide inputs to the global stocktake under the Paris Agreement. However, attributing future warming changes to a single country has inherent caveats given cumulative historical emissions and global interactions.

China’s carbon neutrality pledge, when implemented as a CO2-only pathway, is projected to reduce late-century global mean surface temperature by up to ~0.5 °C under higher-forcing scenarios, accounting for a substantial fraction of total projected warming. When associated reductions in CH4 and N2O are included, the avoided warming increases and mid-term benefits emerge under some scenarios. This study demonstrates, with a fully coupled Earth system model, the potential magnitude and spatial distribution of warming mitigation attributable to a single major emitter’s neutrality pathway, informing assessments of national contributions and the Paris Agreement’s global stocktake. Future work should consider coordinated global mitigation actions, a range of plausible national neutrality pathways, aerosol and short-lived climate forcer changes, larger ensembles to better quantify internal variability, and pathways targeting full GHG neutrality.

• Assumption that only China undertakes significant mitigation, whereas many countries have net-zero pledges; in reality, China’s relative contribution would be smaller under coordinated action.
• The CNCN pathway used is one of many possible emission trajectories; actual pathways depend on policy, sectoral decarbonization, deployment of renewables, and carbon capture/removal technologies.
• Aerosols and short-lived climate forcers co-emitted with fossil fuel use are not explicitly varied; their net climate effects are uncertain and could alter temperature responses.
• Attribution via paired simulations of CO2-only and combined CO2+CH4+N2O cannot isolate a single country’s effect from cumulative historical emissions; past emissions continue to influence future warming.
• Limited ensemble size; additional realizations would reduce uncertainty from internal variability and increase robustness.
• CH4 and N2O are prescribed as global mean concentrations derived from empirical fits to cumulative emissions; uncertainties in emissions-to-concentration relationships and spatially uniform concentration assumptions may affect results.