Abstract

In this work a random field model for shear-wave velocity profiles is presented. This model is compared with an existing model. Their relative merits in terms of their validity for use in site response analyzes are evaluated. A large database of site profiles from sites in Japan is used to calibrate the parameters of the random field models. Randomly generated profiles using both random field models are generated and compared with measured profiles. The uncertainties associated with the shear wave velocity measurements (i.e. epistemic uncertainty), are also included in the simulations and their effect on the resulting variability is analyzed. The randomly generated profiles are then used in site response analyses. Ground motion variability is also taken into account. The two models are compared in terms of the statistics of the shear-wave velocity random field and in terms of the predicted site response. © 2010 ASCE.