Early Prediction System for Glacier Lake Outburst Floods

Abstract

“Glacial Lake Outburst Floods (GLOFs)” are sudden, high-magnitude floods resulting from the breach of natural dams containing glacial lakes. With climate change accelerating glacial lake formation, GLOFs present growing threats to downstream communities and infrastructure. This study introduces an AI-powered chatbot system designed for real-time GLOF risk prediction and assessment. By leveraging predictive models such as ARIMA, GRU, and LSTM, the system analyses user-provided environmental data, including temperature and water levels, to generate actionable insights. A simple and intuitive chatbot interface allows users to input key parameters, which are processed using predefined thresholds to categorize risk levels as high, moderate, or low. The system enhances disaster preparedness and management through accessible and timely decision support. Future work aims to integrate IoT-enabled real-time data streams to improve prediction accuracy further, making this solution a scalable and effective tool for mitigating GLOF risks in vulnerable regions.