Uncertainty Analysis of the Effect of In-Stream Water Level Fluctuations on the Distributed Quantification of Stream-Aquifer Exchanges at the Regional Scale

Abstract:

The classical approach to the modeling of stream-aquifer exchanges at the regional scale (> 10 000 km²) is a conductance model in which river stages are assumed to be constant. However, time fluctuations of river stages have a significant impact. The implementation of variable river stages in hydrological models requires parameters which are difficult to estimate at regional scale. This work aims at analyzing how the quantification of stream-aquifer exchanges at the regional scale is affected by the uncertainties on the parameterization of the process.

A real case study (Loire basin, 90 000 km², France) is considered. The length of the simulated river network is 16141 km, 32% of which is in contact with an underlying aquifer. The surface and groundwater flow in the basin are simulated with EauDyssée, an integrated, distributed, physically-based hydrological model. In-stream water level fluctuations are simulated using a simplified Manning-Strickler approach. Stream-aquifer exchanges are evaluated on a 17 year period (1996-2013) at the daily time step over the river network at a resolution of 1 km.

The spatial distributions and the time fluctuations of stream-aquifer exchanges obtained with different values of the quantities in Manning’s equation are compared with the results of the simulation with fixed river stages. The water fluxes at the stream-aquifer interface are shown to be sensitive to the approach taken for their quantification (constant or variable river stages): in-stream water level fluctuations determine temporary reversals of the gaining or losing regime for some river reaches. Moreover, the impact of the uncertainties on the input quantities of Manning’s equation is assessed.

Finally, it is shown that the modeled stream-aquifer exchanges along the Loire are consistent with the longitudinal temperature profile estimated with the satellite based thermal infrared images (LANDSAT): the groundwater discharge into the Loire warms the river in winter and cools it in summer.