Extended carbon emission efficiency of coal resource-based cities in China: From the perspective of water-land-energy-food pressure and vegetation carbon sinks

This dataset and its accompanying R scripts assess the green low-carbon transition in China's coal resource-based cities (CRBCs). We test the hypothesis that integrating water-land-energy-food (WLEF) pressure and vegetation carbon sinks (VCS) provides a more accurate and systemic evaluation of carbon emission efficiency. Spanning 2011–2021, the dataset includes Extended Carbon Emission Efficiency (ECEE) scores calculated via a Super-SBM model, alongside relevant socio-economic variables.

To address the limitations of conventional econometric models in driver analysis, the repository also evaluates five machine learning algorithms: Random Forest, LightGBM, XGBoost, Decision Tree, and SVM. Researchers can use these resources to fully reproduce the model selection process and the optimal predictive models. This facilitates a deeper understanding of external driving mechanisms and supports evidence-based policy for synergistic carbon-resource governance.