3D interpolation of grain size distributions in the upper 5Â m of the channel bed of three lower Rhine distributaries

This article compares two methods to interpolate grain size distributions to a 3D model (grid-size 25 × 25 × 0.2 m) of the upper 5 m of the riverbed of three lower Rhine distributaries in The Netherlands. These models have been used as input for morphological predictions of the effects of maintenance and restructuring plans of the riverbed. Both methods use field data from seismic surveys and vibrocore borings to construct a 2.5 D layer model. Full grain size distributions within grid cells of 25 × 25 × 0.2 m have been obtained by interpolation within layers, using 3D Kriging, thereby analyzing spatial variability over different scales. In the first method measured grain size distributions have been interpolated. The second method uses measured grain size distributions and synthetic grain size distributions from estimated soil sample parameters (silt content, gravel content, sand median). The combination of synthetic and measured GSD strongly enhances the spatial variation in the interpolation result and prevents a misrepresentation of lithologies in the model.