Data Analytics and Machine Learning Approach for Tsunami Prediction from Satellite and Hydrographic Data

Tsunamis pose significant threats to coastal regions, necessitating effective disaster preparedness and response efforts. This research project aims to understand and predict tsunami characteristics and arrival time. The first stage involves comprehensive data collection spanning historical earthquake, tsunami, and landslide data from 1800 to 2023. Through reputable databases, scientific publications, historical archives, and government agencies, crucial information on location, magnitude, and timing is gathered. The collected data form the foundation for subsequent analyses and predictions. Leveraging machine learning techniques, including the Random Forest Regressor model and K-means clustering, precise forecasts of tsunami wave speed, travel time, and vulnerability are achieved. Real-time data integration and climate change considerations further enhance prediction capabilities. The research culminates in comprehensive visualizations, including maps highlighting tsunami-prone regions and charts depicting the most affected months. These insights aid disaster management authorities and coastal communities in formulating targeted disaster preparedness and mitigation strategies. The study's global collaboration and advanced predictive models help protect vulnerable areas from tsunamis, offering crucial insights for disaster risk reduction and coastal resilience decision-making.

Keywords— K-means clustering, Tsunamis, Seismic Events, Geographical Location, Sea Surface Depth, Tsunami Wave Speed