SKLGP大讲堂第81期 | 马国涛— Uncertainty Quantification of Postfailure Behaviours of Landslides Considering Complex Spatial Variabilities

报告题目：Uncertainty Quantification of Postfailure Behaviours of Landslides Considering Complex Spatial Variabilities

报告人：马国涛

单位：英国华威大学

时间：2024年5月10日16:00—17:00（周五）

地点：新实验楼211教室（珙桐对面）

报告人简介：

Dr. Derek Guotao Ma is an Assistant Professor at the University of Warwick, specializes in Geo-Engineering and Data Science. His acclaimed interdisciplinary research, meriting a prestigious early career fellowship, focuses on the application of data science to geohazards. Honoured with the Global Talent awarded by the Royal Academy of Engineering, Dr. Ma has significantly contributed to policy advisory roles with the Welsh Government's Coal Tips Safety Taskforces. His expertise further encompasses a key position as a corresponding member of the ISSMGE's TC309 technical committee, dedicated to Machine Learning and Big Data. Dr. Ma is recognized for his innovative research in geohazards, particularly through pioneering work in stochastic computational catastrophe modeling and risk uncertainty quantification. His editorial prowess is evident in his contributions to leading publications, including the new Journal of Machine Learning and Data Science in Geotechnics.

报告内容简介：

The challenge of accurately predicting the postfailure behaviour of landslides, crucial for risk management, is exacerbated by the inherent and complex heterogeneity of geomaterials along with their spatially variable mechanical properties. This seminar presents an integrative probabilistic risk assessment framework combining the Generalized Interpolation Material Point (GIMP) method and a novel multivariate stochastic method to evaluate landslide risks. The framework employs non-Gaussian copula-based cross-correlated multivariate random fields, derived from sparse field data, to account for soil spatial heterogeneity and geotechnical uncertainties in the prediction of landslide influence zones. By implementing the GIMP with Monte Carlo simulation, this study demonstrates the significant impact of considering interdependency and spatial variability on the post-failure analysis. A risk factor is also introduced to quantify the degree of threat to nearby infrastructures, offering a comprehensive view of the potential danger zones and enhancing the predictability of mass flow’s impact.