Spatial-temporal prediction of secondary compression using random field theory

A methodology is presented for observation-based settlement predictions by considering the spatial correlation structure of soil. The spatial correlation is introduced among the settlement model parameters, and the settlements at various points are spatially correlated through these geotechnical parameters, which naturally describe the phenomenon. The method is based on Bayesian estimations, considering both prior information, including spatial correlation, and observed settlements, to search for the best estimates of the parameters. Within the Bayesian framework, the optimized selection of the auto-correlation distance, by Akaike's Bayesian Information Criterion (ABIC), and the spatial interpolation of the model parameters, by the kriging method, are also proposed. The application of the proposed approach in secondary compression settlement predictions, based on the linear relationship between settlement and the logarithm of time, is presented in this paper. Several case studies are carried out using both simulated settlement data and actual field observation data. It is concluded that the accuracy of settlement predictions can be improved by taking into account the spatial correlation structure, especially when the spacing of the observation points is shorter than half of the auto-correlation distance, and that the proposed approach produces rational predictions of settlements at any location and at any time with quantified errors. (C) 2012. The Japanese Geotechnical Society. Production and hosting by Elsevier B.V. All rights reserved.