Assessments of Liquefaction Triggering and Manifestation at Three Case History Sites in Napier Based on 1931 Hawke's Bay, New Zealand, Earthquake

The 1931 Hawke's Bay earthquake (∼ M S 7.8) stands as one of New Zealand's most devastating, causing widespread liquefaction and damage across the city of Napier. However, Napier remains underrepresented in liquefaction hazard studies as compared to other regions in the country. The most recent retrospective regional liquefaction study of the 1931 earthquake found that the cone penetration test (CPT)‐based simplified liquefaction triggering framework and manifestation severity index models, specifically liquefaction potential index (LPI) and liquefaction severity number (LSN), tend to overpredict severity relative to field observations from that event. This study investigates the tendency toward overprediction in the context of three case history sites, using CPT‐ and shear wave velocity ( V S )‐based liquefaction triggering and manifestation analyses, and incorporating sensitivity analyses to account for parameter selection uncertainties for the 1931 earthquake event. To address the likely factors contributing to overprediction, we incorporate: (1) the Ishihara‐inspired LPI ISH and LSN ISH to account for nonliquefiable top crust (H 1 ), (2) a modified soil behavior type index ( I B ) to define liquefaction susceptibility cutoff, (3) a New Zealand‐specific fines content (FC) ‐I C correlation, (4) corrections for partial saturation below the groundwater table as indicated by P‐wave velocity ( V P ), and (5) corrections for thin‐layer and transition zone effects. Key findings indicate that the use of an adjusted liquefaction susceptibility cutoff criteria and FC ‐I C correlations improved characterization of liquefaction behavior of transitional fine‐grained soil layers. Accounting for the influence of H 1 and partial saturation below the groundwater table had limited impact on severity predictions. Overall, the findings indicate that, even after extensive sensitivity analyses and adjustments to simplified liquefaction triggering and manifestation assessments, the models still tend to overestimate liquefaction severity at the three case history sites relative to historical observations.