Analitik Prediktif untuk Mitigasi Risiko Iklim pada Rantai Pasok Pangan: Model Ketahanan Agroindustri Berbasis Data

Abstract

Climate change has induced systemic disruptions in global food supply chains, manifested through increasing extreme weather events, harvest uncertainty, and commodity price volatility. This study aims to develop a data-driven agroindustry resilience model by integrating predictive analytics for climate risk mitigation in food supply chains. Employing a systematic literature review and conceptual analysis of twenty Scopus-indexed studies (2019–2025), this research identifies climate risk transmission mechanisms, evaluates machine learning algorithm capabilities for early disruption detection, and constructs a resilience framework grounded in three capability dimensions: absorptive, adaptive, and restorative. Key findings demonstrate that integrating Double Machine Learning, Long Short-Term Memory networks, and Bayesian Networks yields superior climate risk prediction accuracy compared to single-model approaches. This study contributes theoretically by consolidating Dynamic Capabilities Theory, Supply Chain Resilience Theory, and the Climate Risk Framework into a coherent analytical architecture, while offering practical implications for agroindustry policymakers in climate-vulnerable developing countries.