⚠ Could not load forecast data — displaying cached model output
Key Metrics
2025 Observations
GtCO₂
2026 Prediction
GtCO₂
2050 Most Likely
GtCO₂
Peak Emissions Year
Individual Method Projections — 2050
Growth Rate Extrap.
Gaussian Process
RW-Quad 25yr
RW-Quad 20yr
Structural Deceleration in Emissions Growth
Annual & 5-Year Rolling CO₂ Growth Rate (%/yr)

Annual CO₂ Emissions Graph — Four Carbon Emissions Forecast Methods

Fossil CO₂ Projections to 2050 (GtCO₂/yr)

Global CO₂ Data — Method Envelope & Median Trajectory

Combined Projection Envelope (GtCO₂/yr)
Method Comparison
Method Approach Peak Year 2050 Central (GtCO₂) 95% CI
Loading model data…
Sources: Global Carbon Budget 2025 (Friedlingstein et al., 2025) for 2000–2025 observed fossil CO₂ · RCMIP Phase 3 Emissions v1.0.0 (Emissions|CO2|Energy and Industrial Processes, World region) · IEA World Energy Outlook 2025 STEPS for reference benchmark · globalcarbonbudget.org
Methods: COVID years 2020–2021 excluded from all regression training · Recency weighting: 5× for 2022–2025, 2× for 2010–2019, 1× earlier · 95% CI via bootstrap (2,000 resamples) for Growth Rate Extrapolation; Bayesian posterior std for RW-Quad methods · Gaussian Process: Matérn ν=2.5 kernel, sklearn GaussianProcessRegressor · All cumulative totals via trapezoidal integration · RCMIP historical (2000–2024) spliced with GCB observed before scenario branch at 2025.
Note: These are statistical extrapolations, not policy forecasts. Real outcomes depend on policy, technology, geopolitics and economic cycles. RCMIP SSP scenarios use different IAMs (IMAGE, MESSAGE-GLOBIOM, GCAM4, AIM/CGE, REMIND-MAGPIE).
Get climate signals when new forecasts drop
Quantitative updates from Dr Ben McNeil — new data, model updates, and findings on CO₂, warming, and carbon markets.
Subscribe →