This study explores the dual role of artificial intelligence (AI) in addressing climate change by evaluating its potential to enhance climate prediction accuracy while assessing its environmental footprint. The research aims to inform sustainable policymaking aligned with Sustainable Development Goal 13 (Climate Action).
Using data from 72 developed and developing countries (2019–2022), this research employs three machine learning models—Multilayer Perceptron (MLP), Random Forest (RF), and M5P—to predict temperature changes. The methodology encompasses attribute selection via WEKA software, data preprocessing, and performance evaluation using Mean Absolute Error (MAE) and Root Mean Square Error (RMSE).
Results demonstrate that MLP, RF, and M5P models outperform alternative methods in predicting temperature changes, achieving lower MAE and RMSE values when key attributes such as AI investments and PM2.5 pollution levels are incorporated. While AI demonstrates significant potential for optimizing energy use and enhancing early warning systems for extreme weather events, its energy-intensive computational requirements present notable environmental challenges.
This study analyzes diverse attributes across 72 countries, offering a global perspective unlike regional or single-model studies. It links predictive modeling to “green AI” policy recommendations, minimizing environmental impact while enhancing climate mitigation, contributing to sustainable development and supporting evidence-based climate policy.
