PIM-TSUNAMI
A global TSUNAMI hazard amplification model considering site-specific factors and applying Physics-Informed Machine learning
A global TSUNAMI hazard amplification model considering site-specific factors and applying Physics-Informed Machine learning
Project PIM-TSUNAMI aims to develop a model that incorporates the effects of site-specific factors on tsunami heights, using Physics-Informed Machine Learning simulations and offering a user-friendly interface. Megathrust earthquakes in subduction zones, along with their cascading tsunamis, have led to some of the most devastating hazards on Earth, such as the 2004 Indian Ocean tsunami, which claimed more than 220,000 lives, and the 2011 Tohoku tsunami in Japan, resulting in over 20,000 fatalities. Recent understanding has revealed that the tsunami potential of megathrust earthquakes depends on site-specific factors such as slab curvature, bathymetry steepness, and sediment thickness. Despite significant advancements in tsunami hazard analysis over the past 30 years, the critical issue of site-specific factors, which can amplify tsunami hazards two- to threefold, remains a major gap in research and practice. By addressing this critical gap, PIM-TSUNAMI offers significant improvements to global tsunami hazard assessments and contributes to the safety of millions of people worldwide who live at risk of tsunamis.
PIM-TSUNAMI is funded by the European Commission, under the Horizon scheme.
PIM-TSUNAMI is managed by the CERLAB team.