Field-testing liquefaction models based on geospatial vs. geotechnical data

Abstract

This study assesses the relative efficacy of liquefaction models based on geospatial vs. geotechnical data. In particular, state-of-practice geotechnical models based on the Cone Penetration Test (CPT) are compared to geospatial models that use readily available no-cost data. This assessment is performed using a database of 9,623 liquefaction case studies compiled from the 2010-2016 Canterbury, New Zealand, Earthquakes. While the top-performing model is CPT-based, the geospatial models perform surprisingly well given their simplicity. In particular, a region-specific geospatial model out-performs some CPT-based methods. While further research is needed, the presented findings are provocative considering the relative cost and complexity of the geotechnical models. Accordingly, performance assessments of geospatial vs. geotechnical models are ongoing for more than 20 additional earthquakes.