Model-based classification of CPT data and automated lithostratigraphic mapping for high-resolution characterization of a heterogeneous sedimentary aquifer.

摘要: Cone penetration testing (CPT) is one of the most efficient and versatile methods currently available for geotechnical, lithostratigraphic hydrogeological site characterization. Currently soil behaviour type classification (SBT) CPT data however have severe limitations, often restricting their application to a local scale. For parameterization regional groundwater flow or geotechnical models, delineation hydro- lithostratigraphy, SBT would be very useful. This paper investigates use model-based clustering classification, influence different approaches on properties spatial distribution obtained classes. We additionally propose methodology automated mapping regionally occurring sedimentary units using classification. The applied large dataset, covering basin ~60 km2 with predominantly unconsolidated sandy sediments in northern Belgium. Results show that approach superior detecting true lithological classes when compared more frequently unsupervised literature diagrams. demonstrate advanced techniques can provide gain efficiency, time-consuming manual yields at least equally accurate results.