Modeling spatial dependencies of natural hazards in coastal regions: a nonstationary approach with barriers

• DOI: 10.1007/s00477-023-02519-9
• Publication Date: 2023-08-01T15:01:38Z
• Authors: Somnath Chaudhuri; Pablo Juan; Laura Serra Saurina; Diego Varga; Marc Saez
• Keywords: Natural hazard; Smoothing

Abstract Natural hazards like floods, cyclones, earthquakes, or, tsunamis have deep impacts on the environment and society causing damage to both life property. These events can cause widespread destruction lead long-term socio-economic disruption often affecting most vulnerable populations in society. Computational modeling provides an essential tool estimate by incorporating spatial uncertainties examining global risk assessments. Classical stationary models statistics assume isotropy stationarity. It causes inappropriate smoothing over features having boundaries, holes, or physical barriers. Despite this, nonstationary barrier model been little explored context of natural disasters complex land structures. The principal objective current study is evaluate influence compared classical analysing incidence regions islands coastal areas. In study, we used tsunami records from island nation Maldives. For seven atoll groups considered our implemented three distinct categories stochastic partial differential equation meshes, two for one that corresponds concept. results show when assessing variance at scale, outperforms other while maintaining same computational cost as models. broader picture, this research work contributes relatively new field intends establish a robust framework explore phenomena, particularly hazards,