Liquefaction Hazard Assessment

Site triggering analysis evaluates a single soil column from borehole or CPT data, computing a depth-by-depth factor of safety using the Youd and Idriss simplified procedure. Regional hazard assessment must cover areas too large to drill, so it replaces subsurface data with geospatial proxies — slope-derived VS30, topographic and hydrologic indices, surface geology — and predicts the spatial probability of liquefaction, as in the Zhu, Baise, and Thompson model. The two are complementary: the regional map shows where to worry and supports response and planning, while site analysis remains essential for designing individual structures.

Liquefaction requires loose, young, saturated granular soil plus strong shaking, and each ingredient has a mappable proxy. Soft, loose soils have low shear-wave velocity, which Zhu, Baise, and Thompson estimate from terrain slope (steeper terrain implies stiffer ground). Loose young sediments accumulate in flat valleys and floodplains, captured by a compound topographic index. Shallow groundwater is also likeliest in such low-lying settings, often represented by distance to water bodies. These proxies, combined with magnitude-adjusted peak ground acceleration from a ShakeMap, let the model predict liquefaction probability without site-specific borings.

Probability tells you whether liquefaction is likely; severity tells you how damaging it will be. Severity indices integrate the contribution of liquefiable layers over depth: the Liquefaction Potential Index and the Liquefaction Severity Number weight the predicted exceedance of cyclic resistance by depth, so shallow, thick liquefiable zones yield the highest values and the worst surface manifestations such as sand boils, settlement, and lateral spreading. These indices derive from the site-scale Youd and Idriss physics. Where regional subsurface models exist they are computed directly; otherwise geospatial models predict a severity or spatial-extent proxy alongside probability.