Seismic risk for spatially distributed infrastructure is driven mainly by ground failure, defined as permanent ground displacements from mechanisms such as landslides, liquefaction, and seismic compression. Most forms of ground failure are a consequence of soil responses to ground shaking, which should be evaluated on a hazard-consistent scenario basis to represent spatial correlations of intensity measures. A companion paper describes a methodology for identifying hazard-consistent event scenarios. Seismic ground failure responses are evaluated based on regionally-accessible information on geology, groundwater hydrology, and terrain. Given these inputs, liquefaction and landslide displacements are predicted point-by-point on a 10 m grid using customized analysis procedures and logic trees for each ground failure type. For each point, these analyses provide probabilities that the hazard exists, probabilistic distributions (accounting for epistemic uncertainties) of related displacements, and displacement directions (azimuths). Series of points expected to move together (e.g., in a single lateral spread) are grouped into polygons. Ground failure features (landslides, lateral spreads) of varying sizes may occur within these polygons. The output of these analyses are feature locations, sizes, displacement amounts, and displacement azimuths, which can be applied in subsequent fragility and risk analysis of distributed infrastructure systems.
