Evaluation of uncertainties in the existing empirical models and probabilistic prediction of liquefaction-induced lateral spreading

The empirical models of liquefaction-induced lateral ground deformation have the most common applications for the routine projects. Beside the inherent uncertainties of such models, estimation of geotechnical, geometrical, and design earthquake parameters also involves measurement errors. These sets of error may considerably affect the final estimation of ground deformation and lead to questionable designs. Several empirical models are in-use in the current practice; however, a comprehensive study seems to be required in order to evaluate and compare their performances. This papers aims to present: (1) parametric study of the available empirical models of lateral spreading, consideration of the most recent database, and developing a new empirical model, and (2) Reliability analysis of the developed model through FORM, SORM, and Monte Carlo methods. In the first part, unreliable data points with only one observation borehole located farther than 5m from the point of ground deformation are removed from the database. The new empirical model has acceptable accuracy compared with the previous models. In the second part, effect of the geotechnical parameters variability in the estimated values of ground deformations are estimated and the results are presented in terms of probability of exceedence from a certain level of permissible displacement versus the ground slope.