Rapid tsunami force prediction by mode-decomposition-based surrogate modeling

Abstract. This study presents a framework for rapid tsunami force predictions by the application of mode-decomposition-based surrogate modeling with 2D–3D coupled numerical simulations. A limited number large-scale analyses are performed selection scenarios variations in fault parameters to capture distribution tendencies target risk indicators. Then, proper orthogonal decomposition (POD) is applied analysis results extract principal modes that represent temporal and spatial characteristics forces. model then constructed linear combination these modes, whose coefficients defined as functions selected input parameters. example presented demonstrate applicability proposed one tsunami-affected areas during Great East Japan Earthquake 2011. Combining 2D 3D versions stabilized finite element method, we carry out series high-precision different obtain set time history data forces acting on buildings inundation depths. POD construct capable providing equivalent simulation almost instantaneously. Based acceptable accuracy obtained results, it was confirmed useful tool evaluating time-series hydrodynamic buildings.