Abstract Many present-day alluvial floodplains display traces of abandoned meandering channel belts developed during the past millennia (i.e. mid to late Holocene). Deposits of these ancient rivers represent preferential pathways for groundwater flows and related environmental issues, such as contaminant propagation or saltwater intrusion in coastal areas. However, since formative bankfull flows in such old and abandoned routes are hard to estimate, fossil meanders have not been commonly addressed by morphodynamic numerical models, and most of them have been investigated following classical sedimentological approaches based mainly on punctual data derived from sedimentary cores. This study aims at investigating the sediment distribution within different fossil bends on the southern Venetian Plain (NE Italy), and relating such distribution to numerically modelled bed shear stresses used herein as a proxy of sediment sorting patterns. For this purpose, formative flows in the studied palaeomeanders are first inferred from measured sediment grain size and estimates of bend widths. Then, shear stress distributions are computed along the studied palaeobends using a 2D linearized model. Model results are finally compared with conductivity distributions gauged directly in the field through electromagnetic induction investigations in the frequency domain. Our results suggest significant correlations between shear stress distributions and sediment sorting estimated from conductivity data. We deem that the integration between sedimentological reconstructions and state-of-the-art numerical modelling can provide a solid contribution to predicting the spatial distribution of sediment properties within ancient meandering channel belts, with relevant implications for the understanding of shallow aquifer dynamics and soil management.

Abstract Characterization of the spatial variability of hydraulic properties of an aquifer is essential for reliable modeling of the fate of contaminants in the subsurface. Many geophysical methods offer the potential to derive such information, because of the high spatial density of sampling and the, albeit indirect, relationship between many geophysical and hydraulic parameters. In particular, borehole-to-borehole imaging may provide high-resolution sampling at scales which will permit detailed site characterization. Here, we examine the high-resolution spatial variability of electric and electromagnetic properties in the vadose zone at a specific field site in the Triassic Sherwood Sandstone aquifer in the UK, using cross-borehole radar. Assessment of spatial variability in the vadose zone is achieved through computation of experimental semivariograms of geophysical properties obtained from inter-well tomograms. The variability is compared with that deduced from conventional geophysical well-logging tools. The stratified nature of the site is clearly identified by gamma-log measurements. Single-hole and cross-hole geophysical measurements involving electrical properties are affected by both the stratification and the hydrological forcing conditions at the surface, and all show a similar spatial correlation structure. These data suggest that a representation of one-dimensional recharge processes is appropriate at the site. Data on moisture content from cross-hole radar are quantitatively compared against the results of stochastic unsaturated-flow simulations, accounting for the lithological spatial variability as described by gamma logs. The results demonstrate how improved conceptualization of the hydrological model of the site is achieved through incorporation of such geophysical data. Such a methodology permits improved assessment of mechanisms of recharge and transport at the site.