Facies characterisation of sediments from the East Frisian Wadden Sea (Germany): new insights from down-core scanning techniques

Sediment facies provide fundamental information to interpret palaeoenvironments, climatic variation, archaeological aspects and natural resource potentials since they are summary products of depositional processes, environmental conditions and biological activities for a given time and location. The conventional method of facies discrimination relies on macroscopic and/or microscopic determination of sediment structures combined with basic physical, chemical and biological information. It is a qualitative measure, depending on observer-dependent sedimentological descriptions, which cannot be reanalysed readily by further studies. Quantitative laboratory measurements can overcome this disadvantage, but are in need of large sample numbers and/or high temporal resolution, and are time-, labour- and cost-intensive. In order to facilitate an observer-independent and efficient method of facies classification, our study evaluates the potential of combining four non-destructive down-core scanning techniques: magnetic susceptibility (MS), X-ray computed tomography (CT), X-ray fluorescence (XRF) and digital photography. These techniques were applied on selected sections of sediment cores recovered around the island of Norderney (East Frisian Wadden Sea, Germany). We process and integrate the acquired scanning measurements of XRF elemental intensities, represented by principal components, MS, CT density and lightness of eight sediment facies previously recognised by conventional facies analysis: moraine, eolian/fluvial, soil, peat, lagoonal, sand flat, channel fill and beach-foreshore. A novel type of density plot is introduced to visualise the digitised sediment information that allows an observer-independent differentiation of these facies types. Thus, the presented methodology provides the first step towards automated supervised facies classification with the potential to reproduce human assessments in a fully reproducible and quantitative manner.

Automated summary

Traditional methods of sediment facies discrimination rely on observer-dependent descriptions and cannot be easily reanalyzed. Quantitative laboratory measurements can overcome this limitation but are time-, labor-, and cost-intensive. This study evaluates the potential of combining non-destructive down-core scanning techniques for automated supervised facies classification, providing a novel approach towards reproducible and quantitative analyses.