Abstract
This research utilizes the 2016 Carbon Disclosure Project (CDP) dataset and predictive analytics to explore water risk management in global supply chains within the landscape of ecological imperatives intersecting with economic realities. Employing a Random Forest (RF) model, the study investigates water risks from local to global scales, emphasizing the strategic importance of understanding and managing these risks. The findings reveal the RF model’s efficacy in predicting the financial impacts of water risks, highlighting the necessity for proactive risk management strategies in supply chains. This research not only demonstrates the application of machine learning in green supply chain management but also paves the way for future studies on comprehensive and adaptive environmental risk mitigation approaches.