A COMPREHENSIVE BENCHMARK OF LINEAR AND ENSEMBLE MACHINE LEARNING MODELS FOR GLOBAL CO₂ EMISSION FORECASTING

Year 2025, Volume: 11 Issue: 2, 247 - 262, 30.12.2025

Hasan Uzel , Feyyaz Alpsalaz , Emrah Aslan , Yıldırım Özüpak

https://doi.org/10.51477/mejs.1693320

Abstract

This study evaluates Linear Regression, Random Forest, XGBoost and CatBoost to forecast global CO₂ emissions from 2001 to 2021 using the Global Carbon Project dataset (accessed via Our World in Data). A leakage free pipeline standardizes preprocessing, prevents temporal spillover and applies a consistent train–test protocol. Performance is summarized with MSE, RMSE, MAE, MAPE and R² to enable fair, reproducible comparisons. Linear Regression delivers the strongest out of sample accuracy (R² = 0.94, RMSE = 3.81, MAPE = 12.9%), reflecting predominantly linear and autoregressive dynamics. Boosting models (XGBoost, CatBoost) follow closely (R² > 0.914), capturing nonlinear fluctuations, whereas Random Forest is comparatively weaker (R² = 0.879). Feature importance analysis highlights short-term lags (lag₁–lag₂) as dominant predictors, corroborated by autocorrelation, partial autocorrelation and Augmented Dickey–Fuller tests. Overall, the study provides a transparent global baseline and a standardized evaluation protocol that can be extended to country-granular analyses and policy experiments. By clarifying when simple statistical models suffice and when ensemble approaches add value, the results offer evidence-based, actionable guidance for researchers and policymakers seeking interpretable, scalable tools for emissions monitoring, planning and policy relevant scenario design.

Keywords

CO2 Emissions , Machine Learning , Statistical Model , Forecasting

Ethical Statement

Our study does not cause any harm to the environment and does not involve the use of animal or human subjects. Therefore, it was not necessary to obtain an Ethics Committee Report.

References

• Aras, S., Van, M. H., “An interpretable forecasting framework for energy consumption and CO₂ emissions”, Applied Energy, 328, 120163, 2022.
• Özüpak, Y., Alpsalaz, F., Aslan, E., Uzel, H., “Hybrid deep learning model for maize leaf disease classification with explainable AI”, New Zealand Journal of Crop and Horticultural Science, 53(5), 2942–2964, 2025.
• Yang, D., et al., “Time-series forecasting of a CO₂-EOR and CO₂ storage project using a data-driven approach”, Energies, 15(13), 4768, 2022.
• Aslan, E., “Araçlarda CO₂ emisyonlarının farklı yapay sinir ağı modelleri kullanılarak tahminlerinin karşılaştırılması”, Çukurova Üniversitesi Mühendislik Fakültesi Dergisi, 39(2), 309–324, 2024.
• Hassan, T., Mansour, K., Waseemullah, W., Memon, M. Q., “Carbon dioxide emission forecasting using machine learning and time series statistical models”, Proceedings of 25th International Arab Conference on Information Technology, pp. 1–6, 2024.
• Zhou, F., et al., “Fire prediction based on CatBoost algorithm”, Mathematical Problems in Engineering, 2021(1), 1929137, 2021.
• Kumari, S., Singh, S. K., “Machine learning-based time series models for effective CO₂ emission prediction in India”, Environmental Science and Pollution Research, 30(55), 116601–116616, 2023.
• Li, X., Zhang, X., “A comparative study of statistical and machine learning models on carbon dioxide emissions prediction of China”, Environmental Science and Pollution Research, 30(55), 117485–117502, 2023.
• Magazzino, C., Mele, M., “A new machine learning algorithm to explore the CO₂ emissions–energy use–economic growth trilemma”, Annals of Operations Research, 345(2), 665–683, 2022.
• Farahzadi, L., Kioumarsi, M., “Application of machine learning initiatives and intelligent perspectives for CO₂ emissions reduction in construction”, Journal of Cleaner Production, 384, 135504, 2023.
• Meng, Y., Noman, H., “Predicting CO₂ emission footprint using AI through machine learning”, Atmosphere, 13(11), 1871, 2022.
• Ayaz, İ., “Forecasting CO₂ emissions with machine learning methods: Türkiye example and future trends”, Nature Engineering Sciences, 5(2), 82–87, 2024.
• Bhatt, H., Davawala, M., Joshi, T., Shah, M., Unnarkat, A., “Forecasting and mitigation of global environmental carbon dioxide emission using machine learning techniques”, Cleaner Chemical Engineering, 5, 100095, 2023.
• Tawiah, K., Daniyal, M., Qureshi, M., “Pakistan CO₂ emission modelling and forecasting: A linear and nonlinear time series approach”, Journal of Environmental and Public Health, 2023(1), 5903362, 2023.
• Wen, T., Liu, Y., Bai, Y. H., Liu, H., “Modeling and forecasting CO₂ emissions in China and its regions using a novel ARIMA–LSTM model”, Heliyon, 9(11), e21241, 2023.
• Budennyy, S. A., et al., “eco2AI: Carbon emissions tracking of machine learning models as the first step towards sustainable AI”, Doklady Mathematics, 106(1), S118–S128, 2022.
• Tansini, A., Pavlovic, J., Fontaras, G., “Quantifying the real-world CO₂ emissions and energy consumption of modern plug-in hybrid vehicles”, Journal of Cleaner Production, 362, 132191, 2022.
• Ajala, A. A., Adeoye, O. L., Salami, O. M., Jimoh, A. Y., “An examination of daily CO₂ emissions prediction through comparative machine learning approaches”, Environmental Science and Pollution Research, 32(5), 2510–2535, 2025.
• Linardatos, P., Papastefanopoulos, V., Panagiotakopoulos, T., Kotsiantis, S., “CO₂ concentration forecasting in smart cities using a hybrid ARIMA–TFT model”, Scientific Reports, 13(1), 1–22, 2023.
• Li, X., Ren, A., Li, Q., “Exploring patterns of transportation-related CO₂ emissions using machine learning methods”, Sustainability, 14(8), 4588, 2022.
• Yuan, H., Ma, X., Ma, M., Ma, J., “Hybrid framework combining grey system model with Gaussian process and STL for CO₂ emissions forecasting”, Applied Energy, 360, 122824, 2024.
• Emami Javanmard, M., Ghaderi, S. F., “A hybrid model applying machine learning and optimization to forecast greenhouse gas emissions”, Sustainable Cities and Society, 82, 103886, 2022.
• Han, Y., et al., “Novel economy and carbon emissions prediction model using improved residual neural network”, Science of the Total Environment, 860, 160410, 2023.
• Yang, H., Yang, X., Li, G., “Forecasting carbon price in China using a novel hybrid model”, Journal of Cleaner Production, 401, 136701, 2023.
• Faruque, M. O., Rabby, M. A. J., Hossain, M. A., Islam, M. R., Rashid, M. M. U., Muyeen, S. M., “A comparative analysis to forecast carbon dioxide emissions”, Energy Reports, 8, 8046–8060, 2022.
• Iftikhar, H., Khan, M., Żywiołek, J., Khan, M., López-Gonzales, J. L., “Modeling and forecasting carbon dioxide emission in Pakistan”, Heliyon, 10(13), e33148, 2024.
• Ahmed, S., Khan, M., Ali, R., “CatBoost-based predictive modeling of global CO₂ emissions”, Environmental Modelling and Software, 170, 106998, 2023.
• Zhang, S., Li, M., Chen, J., “Hybrid deep learning and feature importance analysis for carbon emission forecasting”, Environmental Science and Pollution Research, 31, 55621–55635, 2024.
• Khan, K. A., Bhuiyan, R., Alam, F., “Benchmarking machine learning models for long-term CO₂ emission forecasting”, Energy Reports, 10, 1189–1203, 2023.
• Costantini, L., Marzocchi, M., Zollo, R. N., “Forecasting national CO₂ emissions worldwide”, Scientific Reports, 14(1), 12345, 2024.
• Begum, A. M., Rahman, K. S., Lee, J. K., “A machine learning approach to carbon emissions prediction of top emitters by 2030”, Scientific Reports, 15(1), 2023, 2025.
• Emissions by Country, “Global fossil CO₂ emissions by country (2002–2022)”, Kaggle Dataset [Online]. Available: https://www.kaggle.com/datasets/thedevastator/global-fossil-co2-emissions-by-country-2002-2022, Accessed: 2025.
• Aslan, E., Özüpak, Y., “Advanced skin cancer detection using convolutional neural networks and transfer learning”, Middle East Journal of Science, 10(2), 167–177, 2024.
• Alpsalaz, F., Özüpak, Y., Aslan, E., Uzel, H., “Classification of maize leaf diseases with deep learning and explainable artificial intelligence”, Chemometrics and Intelligent Laboratory Systems, 262, 1–14, 2025.