Designing a multi-scale sampling system of stream–aquifer interfaces in a sedimentary basin

Highlights

• Methodological Framework to study stream–aquifer interaction.

• Imaging of structural heterogeneities at the regional and local scales.

• Geostatistical interpolation of piezometric head snapshot campaigns.

• Hyporheic zone temperature profiles used as a tracer of the flow.

• Quantification of water fluxes using thermo–hydro modeling.

Abstract

A methodological framework is proposed for designing a multi-scale sampling system to quantify the stream–aquifer interactions in a multi-layer aquifer system. First, geophysical and drilling investigations are performed to assess the regional structure of the aquifer system and the local connectivity between streams and aquifer units. At the catchment scale, the investigations permit to define the composition of the upper aquifer unit. At the local scale, the connectivity status between streams and aquifer units is evaluated using various settings for electrical resistivity tomographies. These geophysical investigations are then used to select local monitoring stations (LMSs) along the stream network. Moreover, piezometric head maps representative of low and high flow regimes are interpolated using geostatistics, which provides distributions of both piezometric heads and standard deviations of the estimation error (STEE). The map of STEE is used to define the location of new piezometers. Altogether, the sampling system allows for monitoring water exchanges on a 40 km² watershed along 6 km of the stream network, with a finer hydro-geophysical sampling at each LMS. Finally, temperature profiles in the HZ are interpreted with a coupled thermo–hydro finite element code at the upstream station of the domain. Multiple simulations indicate first proof of evidence for a gaining stream in the upstream part of the sampling domain.

Introduction

The stream–aquifer interface is nowadays considered a key transitional component characterized by a high spatio-temporal variability in terms of physical and biogeochemical processes (Brunke and Gonser, 1997, Krause et al., 2009). This interface needs further consideration for characterizing the hydrogeological behavior of basins (Hayashi and Rosenberry, 2002), and therefore continental hydrosystem functioning (Saleh et al., 2011).

From a conceptual point of view, stream–aquifer exchanges are driven by two main factors: the hydraulic gradient and the geological structure. The hydraulic gradient defines the water pathways (Winter, 1998), while the geological structure defines the conductive properties of the stream–aquifer interface (White, 1993, Dahm. et al., 2003). The timescale that is to be considered varies depending on the studied object (hyporheic zone itself or a sedimentary basin) (Harvey, 2002). The sampling frequency can also bias the quantification of processes (de Fouquet, 2012). Estimating the stream–aquifer exchanges at a sedimentary basin scale then requires the combination of various processes with different characteristic times or periods covering a wide range of temporal orders of magnitude (Blöschl and Sivapalan, 1995, Flipo et al., 2012, Massei et al., 2010): hour-day for river flow, year-decade for effective rainfall, decade-century for subsurface transit time. To address this, models are used as spatio-temporal interpolators.

Studying stream–aquifer interface thus requires to couple multi-scale sampling and monitoring strategies, spatio-temporal data analysis, interpretations and interpolations, as well as modeling techniques (Fig. 1). Many sampling methods are available that aim at understanding the stream–aquifer interactions (Kalbus et al., 2006). Most of them are indirect methods, which permit the localization and identification of the exchanges. Almost all of these methods are site specific and do not solely allow for quantifying water exchanges along the stream network, which requires a pluridisciplinary (Sophocleous, 2002, Winter, 1998, Woessner, 2000), multi-scale (Scanlon et al., 2002) approach to limit the errors and to validate the estimations (Fleckenstein et al., 2010). Among a selection of 39 papers dealing with multi-scale or/and multi-measurement stream–aquifer interaction studies (Table 1), only Kikuchi et al. (2012) integrate the three major spatial scales of interest (Fig. 1): hydrosystem, reach and HZ scales.

The goal of this paper is to provide a methodological framework to build up a multi-scale sampling network of stream–aquifer water exchanges that integrates the idea of spatially telescoping measurements (Kikuchi et al., 2012). The sampling network has to be coupled with data analysis and interpolations to provide temporally dense punctual datasets, the hydrogeological structure of the multi-layer sedimentary basin, the structure of the hyporheic zone (HZ), as well as spatial distributions of hydraulic head. These raw and interpreted data are used to run models of the whole hydrosystem and of a peculiar HZ (Fig. 1) in a second step.

In the first part, the hydrosystem of the Orgeval basin is presented together with the methods used to assess (i) the regional piezometric head distribution using geostatistics, (ii) the regional structure of the basin, and (iii) the connectivity status of stream–aquifer at five different locations of the stream network using various geophysical investigations and drilling core analysis. In the second part, the results of the data analysis and interpolations of those data are presented. In the final part, the design of the multi-scale sampling system is presented, with a special emphasis on the HZ monitoring stations. Preliminary results of temperature measurements coupled with a thermo–hydro finite element model show evidence of advective water fluxes from aquifer to stream, which proves the pertinence of the proposed framework.