ISTerre seminar

---

Multi-hazard building exposure modeling using AI methods

Tuesday 14 January 2025 - 14h00

Laurens Oostwegel - GFZ / ISTERRE
--- 
Risk assessment is a highly spatial practice: hazards occur on geographically bound area, where the intensity can vary highly from site to site. At the same time, hazards do not manifest in a vacuum, but arise within the dynamics of urban environments. Cities are complex and to get a better understanding of the interaction of exposure and hazard, we need to look at the inherent differences within the built environment: we tend to find the same housing styles in the same areas; zoning can largely determine the shape of a city; similar people tend to live together. This thesis is an exploration of leveraging the abundance of open data sources for new perspectives into exposure modeling: how can we use the urban fabric to better calculate risk? How do we include time as a dimension of exposure, e.g. seasonal variations based on touristic information? Can we identify physical and socioeconomic vulnerabilities to hazards using mere open data and AI algorithms? By exploring these questions, this thesis aims to address new ways to deal with risk assessment, in a world that is urbanizing at a fast pace and where hazards are intensifying due to the increase of extreme weather events.

---

Organizing team : Ondes et structures

Salle Dolomieu, Maison des Géosciences, 38400 Saint Martin d'Hères