Probabilistic Updating of Pore Pressure Fields

Abstract

A first‐order second‐moment method is proposed to quantify uncertainty about steady‐state pore pressures originating in random spatial variability of soil permeabilities. A Bayesian framework for systematic interpretation of piezometer observations is set out, which allows updating of the pore pressure field and, therefore, of geotechnical performance predictions. A probabilistic quantification of the observational method is attempted, so that probabilities assigned to different hypotheses (states of nature) regarding seepage can also be updated on the basis of piezometer readings. Application of the approach to the decision problem of sequential instrument location is suggested. An example illustrates the versatility of the proposed methodology.