Towards Net-Zero Emissions in OECD Countries: Forecasting AI by Machine Learning Methods

Abstract

Achieving net-zero emissions is a paramount objective for Organisation for Economic Co-operation and Development (OECD) countries in combating climate change and fostering sustainable development. This study provides an overview of the strategies, opportunities, and challenges facing OECD countries in their transition towards achieving net-zero emissions. This study delineates the OECD's commitment to ambitious climate targets, including the overarching goal of achieving carbon neutrality by 2040. While some existing models provide reasonably accurate predictions of CO2 emissions, the model presented in this study offer improved prediction capabilities for all OECD countries. This study successfully predicts historical emissions, current emissions, and future emissions from 1990 to 2022 using Machine Learning (ML) methodology. The study forecasts global CO2 emissions for all OECD countries from 2022 to 2042 (near future) using prediction models using SARIMA (Seasonal Autoregressive Integrated Moving Average) based on ARIMA (Autoregressive Integrated Moving Average). The primary aim is to compare these models and identify the most effective one for predicting the transition to net-zero emissions for all OECD countries. These predictions highlight that policymakers should thoroughly evaluate the measures and strategies to promote a transition to net-zero emissions and reduce the levels of CO2 emissions. Furthermore, it highlights the potential co-benefits of transitioning to a low-carbon economy, including improved air quality, enhanced energy security, and job creation.

Keywords

• Environmental sustainability
• Carbon emissions
• Net-zero emissions
• Machine learning (ML)

References

1. Celik, S. (2024). Towards net-zero emissions in OECD countries: Forecasting AI by machine learning methods. The Eurasia Proceedings of Science, Technology, Engineering & Mathematics (EPSTEM), 27, 99-107.