A Sedimentological Model to Characterize Braided River Deposits for Hydrogeological Applications

Abstract

Braided river deposits form important aquifers in many parts of the world, and their heterogeneity strongly influences groundwater flow and mass transport processes. To accurately characterize these coarse gravelly aquifers, it is important to understand the erosional and depositional processes that form these sediments. Moreover, it is important to evaluate the relative importance of various parameters that determine the preservation potential of different depositional elements over geological time scales. These objectives may be achieved by developing techniques that allow for the integration of different quality data into quantitative models. Information concerning sedimentary textures and the spatial continuity of sedimentary structures in braided river deposits, inherent in depositional facies descriptions, allows the spatial variability of hydrogeological properties (e.g. hydraulic conductivity and porosity) to be predicted. Depositional elements in gravel deposits can contain a restricted range of textures, which form a limited number of sedimentary structures. These depositional elements are bounded by erosional and/or lithological surfaces. The frequency, size and shape of different elements in a sedimentary sequence depend on several factors, including aggradation rate, channel belt mobility on the kilometre scale, gravel-sheet/scour activity at the scale of hundreds of metres and topographic position of the different elements within an evolving system. Preserved shape and size of the elements affect the correlation lengths and the standard deviations of the aquifer properties, such as hydraulic conductivity and porosity. Different quality data sets that may be used in characterizing braided river deposits can be recognized in outcrop, boreholes and on ground-penetrating radar (GPR) sections. This paper proposes a means of integrating outcrop, borehole and GPR data into a stochastic framework of sedimentary structures and the distribution of hydraulic aquifer properties. Data integration results in variable degrees of uncertainty when assigning values to hydraulic properties and characterizing the geometry of sedimentary structures. An application of this approach is illustrated using a data set (400 m × 550 m) from the northeastern part of Switzerland at the confluence of the Rhine and Wiese rivers. The data set includes drill-core data from five boreholes and 14 GPR sections with a total length of 3040 m. The results of the variogram analysis provide the orientation of sedimentary structure types representing the main flow direction of the River Rhine in the lower part of the aquifer, and of the River Wiese in the upper part. The analysis also results in large ranges of spatial correlation, ranging from a few metres up to tens of metres for the different sedimentary structure types.