The application of stochastic modelling to the evaluation of ground conditions in Late Quaternary Sediments in Göteborg, Sweden

Abstract

Statistical analysis based on the spatial analysis of variance has been applied to the characterisation of Late Quaternary sediments within the valley of Kvilledalen in Gfteborg, illustrated in this paper by the bed thickness of the marine clay. The analysis of the thickness demonstrates a low random disturbance which was attributable to small scale environmental variability rather than to errors of measurement, and a maximum range of influence of 300 metres from the location of each observation point. Beyond this distance there was no spatial dependency and reliable prediction of values was not possible using the existing data. Stochastic modelling solved using kriging has been used to estimate the distribution of each bed thereby giving a quantifiable degree of certainty and improving the basis for predicting ground conditions based upon the assessment of existing information.