Einsatz geostatischer Verfahren zur optimalen Erkundung und modellhaften Beschreibung des Untergrundes von Deponien Endbericht

Geostatistical methods have been developed to optimize investigation and underground modelling especially of waste disposal sites. Analysis of the inaccuracy of sampling data takes account of the soil, rock and groundwater parameters as regionalized variables. Starting with an empirical semivariogram or variogram, the site being investigated is divided into polygons and the variance in the area of each polygon is determined. This is done to identify the spatial distribution of the errors in the available empirical data. Subsequently, areas of statistically low confidence can then be further sampled. This can be repeated until the necessary accuracy has been obtained. The method has also been extended to include multivariate analysis. The geostatistical tools can also be used to compare several different exploration strategies in order to determine the possible increase in accuracy. But in areas of complicated geology it is necessary to have a relatively dense borehole grid (spacing about 100 m) before a reasonable fit of the variogram to the theoretical model can be obtained. Within this dense grid it is possible to optimize the number and spacing of the sampling sites. Parameter values that express heterogeneity in the data are determined for stochastic simulation models, assessment of risk, and the design of a multi-barrier safety concept. (orig.)SIGLEAvailable from TIB Hannover: F96B849+a / FIZ - Fachinformationszzentrum Karlsruhe / TIB - Technische InformationsbibliothekBundesministerium fuer Forschung und Technologie (BMFT), Bonn (Germany)DEGerman