Field data-based modeling of lateral ground surface deformations due to earthquake-induced liquefaction

Abstract

Liquefaction is one of the most damaging consequences of an earthquake. Lateral ground surface deformations occur due to soil liquefaction. The magnitude of lateral deformations can be small or large depending on the seismic characteristics, soil deposit properties, and ground slope. A safe design in the liquefaction-prone regions requires an accurate estimation of lateral ground deformations. In this study, a wide-ranging data of real lateral ground deformations induced by past earthquakes were gathered and analyzed. Using the adaptive neuro-fuzzy inference system, a robust model was developed to predict lateral ground deformations. The sensitivity analysis was performed in order to investigate the behavior of the developed model under various conditions. Finally, the proposed model was compared with the available relationships for the estimation of lateral ground surface deformations. The results demonstrate reasonable accuracy of the proposed model for estimating lateral ground deformations induced by liquefaction. Indeed, this can be considered as an effective step to reduce uncertainty in geotechnical earthquake analyses in liquefaction-prone areas.