.png){kind=icon}
.png){kind=icon}
Return
A framework for carbon footprint computation and forecasting for Nigeria’s industrial decarbonization plan (NIDP)
.png){kind=icon}
OA
AI
B
Z
A
K
İ
DOI：10.3389/fenrg.2026.1717733.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This study presents a comprehensive comparative analysis of forecasting methodologies within a Machine Learning (Machine Learning)-driven framework for monitoring and predicting Nigeria’s industrial CO2 emissions in support of the nation’s 2060 net-zero target. Four primary model architectures; Multiple Linear Regression (MLR); Prophet; Support Vector Regression and Random Forest were evaluated using advanced time-series diagnostics and a recursive forecasting framework to assess their predictive fidelity. To ensure methodological rigor and mitigate look-ahead bias; feature scaling was fitted exclusively on training data; while non-stationary series were addressed through first-differencing and validated via Granger causality testing. The Multiple Linear Regression (MLR) model demonstrated superior predictive performance (R2=0.978; Mean Absolute Error = 0.66 Mt CO2); suggesting that Nigeria’s industrial emissions currently follow a strong deterministic linear trajectory driven by sectoral expansion rather than complex stochastic cycles. While the SVR (Recursive) model showed improved tracking over standard non-linear approaches (R2=0.212); stochastic models such as Prophet and Random Forest failed to generalize on the annual dataset; yielding negative R2 values due to the limited sample size and the lack of high-frequency seasonality. Sectoral diagnostics identified the Transport and Power Industry as the dominant drivers of industrial CO2 emissions. Utilizing an Auto-ARIMA-based recursive framework to project these predictors; the study forecasts that Nigeria’s industrial CO2 emissions will reach 135.65 Mt by 2025 under a business-as-usual scenario. This represents a significant upward departure from historical baselines and provides a critical “warning signal” for the National Industrial Decarbonization Plan (NIDP). The findings highlight an urgent need for targeted interventions in transport electrification and industrial grid greening to realign the sector with Paris Agreement commitments.
Keywords:
machine learning
Nigeria
time-series forecasting
decarbonization
comparative analysis
CO2 emissions
policy modelling
Journal
IF:
2.4
Papers:
923
Citations:
1.4W
Organization
Cited Papers