Forecasting the Economic and Environmental Impact of Low-Carbon Technology Trade in the United States Using Machine Learning

Abstract

The global shift towards sustainability has increased the strategic significance of low-carbon technologies in international trade. This research introduces a machine learning-driven framework for forecasting the economic and environmental impacts of low-carbon technology trade in the United States, utilizing real-world data related to the economy, trade, and emissions.The study begins by integrating the import and export records of green technologies, such as solar panels and wind turbines, with national economic indicators (like GDP contribution and clean energy jobs) and environmental data (including sectoral CO₂ emissions and the share of renewable energy) to create a comprehensive analytical dataset. Through extensive feature engineering, we derive metrics such as temporal trade lags, emissions intensities, GDP-to-trade ratios, and sectoral growth indicators. We apply time-series decomposition and smoothing techniques to uncover seasonal and trend-based dynamics. Next, we train and evaluate a series of regression and hybrid models, including Random Forest, XGBoost, and LSTM networks, to forecast future economic gains and carbon reduction outcomes associated with clean technology trade patterns. We use evaluation metrics such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and R² to compare model accuracy in relation to both economic and environmental targets. The top-performing hybrid model, which combines LSTM and Random Forest, achieves an RMSE of 0.34 and an R² of 0.95 for GDP impact prediction, as well as an MAE of 0.48 and an R² of 0.92 for CO₂ reduction forecasting. Feature importance analysis using SHAP values indicates that carbon tariffs, trade volume, and policy indices are significant predictors of environmental impact. Finally, we conduct scenario modeling to simulate trade policy shifts and global price shocks to evaluate their effects on sustainability outcomes. Our framework offers a predictive foundation for policymakers and investors to assess and optimize the trade-offs between economic growth and climate objectives within the clean technology sector.