Search for a command to run...
Abstract Life-loss vulnerability models, which estimate the probability of an individual being killed given landslide impact, are critical inputs for landslide risk assessments focused on life loss. Land-use decisions in western Canadian landslide-prone areas increasingly rely on these assessments, however, there are no published life-loss vulnerability models for individuals in typical Canadian buildings. To address this gap, we developed an empirical life-loss vulnerability curve for people inside wood-frame buildings as a function of impact pressures from flow-type landslides. It is based on a new method which combines estimates of building damage probability given landslide impact and life-loss probability for people in damaged buildings. It is produced from a dataset of 421 wood-frame buildings damaged by flow-type landslides and 242 people in those buildings. The results show that life-loss vulnerability increases with building damage state, and that occupants are much more likely to be killed in buildings that at least partially collapse, likely because occupants have fewer options to avoid harm as building damage increases. The vulnerability model indicates that life-loss vulnerability increases rapidly from impact pressures of 1–32 kPa because, over this range, at least partial collapse of the building increases to the most likely damage state. The resulting vulnerability curve likely overestimates vulnerability at low impact pressures and underestimates vulnerability at high impact pressures. We provide an annotated vulnerability curve to help users account for and communicate these limitations.