Scoring, selecting, and developing physical impact models for multi-hazard risk assessment

Abstract

This study focuses on scoring, selecting, and developing physical fragility (i.e., the probability of reaching or exceeding a certain damage state given a specific hazard intensity) and/or vulnerability (i.e., the probability of impact - or consequence - given a specific hazard intensity) models for assets of interest, with particular emphasis on buildings. Given a set of multiple relevant hazards for a selected case-study region, the proposed procedure involves 1) mapping the relevant asset classes (i.e., construction types for a given occupancy) in the region to a set of existing candidate fragility, vulnerability and/or damage-to-impact models, also accounting for specific modelling requirements (e.g., time dependency due to ageing/deterioration, multi-hazard interactions); 2) scoring the candidate models according to relevant criteria to select the most suitable models for a given application; or 3) using state-of-the-art numerical or empirical methods to develop fragility/vulnerability models not already available. The approach is demonstrated for the buildings of the virtual urban testbed “Tomorrowville”, considering earthquakes, floods, and debris flows as case-study hazards.