Trade flows and carbon emissions: how far are we from the climate targets?

Abstract

Policymakers are increasingly calling for tangible actions to meet the United Nations targets for reducing emissions to 1990s levels. This study assesses whether climate targets are attainable by 2030 for a selected set of countries, with a particular focus on the transportation sector. PyCaret’s Python library is used to train and compare a range of time series models and derive transport emission forecasts of countries up to 2030. We employ a range of machine learning models, including ARIMA, AdaBoost, and Theta forecaster, using historical transport emissions data to forecast CO2 emissions and assess the likelihood of countries achieving the EU Green Deal’s decarbonization objectives. To further support policy design, we analyze freight modal emissions using trade flow data and group countries using the K-Means clustering based on their emission profiles. Our findings show that only two countries—Greece and Italy—are currently on track to meet the 2030 targets. The remaining nations require urgent and targeted policy reform, particularly in shifting freight transport toward lower-emission modes. We conclude by offering realistic, data-driven mitigation strategies for those unlikely to meet their commitments.