Water flow and solute transport in the stream corridor: hyporheic flow directions, parameters identifiability and transient storage processes

Abstract

The continuum of a stream channel, its permeable bed sediments and the adjacent groundwater is termed the stream corridor. The study of how water moves within the stream corridor is crucial to understanding the spatio-temporal evolution of solute transport and nutrient cycling affecting river water quality, and is thus essential for more efficient water resources management.However, the current understanding of stream corridor processes is limited in that past studies have focused only on certain hydrologic conditions and often did not include observations of the near-stream groundwater table. The key objective of this thesis is to understand and describe the spatio-temporal role of the drivers controlling the water transport in the stream corridor in a headwater stream reach. To address this aim, this thesis employed high/frequency groundwater measurements in the near-stream domain via a novel spatial-dense well network, proposes a new iterative modelling approach, and explores water transport via in-stream tracer experiments across a wide spectrum of hydrologic conditions.Chapter 1 presents groundwater data collected near the Weierbach stream in Luxembourg to analyse the drivers of stream-groundwater interactions over a wide spectrum of hydrologic conditions. The data suggest that groundwater inflow from upstream locations, together with the storage capacity in the regolith and the bedrock topography, control stream-groundwater exchange. The role of these drivers depends on the flow situation. During dry hydrologic conditions, the bedrock topography controls the flow from the stream into the groundwater, while with increasing wetness, upslope-footslope connectivity and precipitation control the near-stream groundwater flow direction.In order to predict solute transport in stream corridors, mathematical models are widely used, but the estimation and interpretation of their parameters is difficult. This is because combinations of different parameter values can give the same model performance (identifiability problem). Chapter 2 proposes a novel iterative modelling approach that combines random sampling with global and dynamic identifiability analysis to simulate tracer breakthrough curves obtained for different discharge conditions. The results demonstrate that the method is able to obtain identifiable model parameters, which has not been systematically achieved in previous studies. The analysis also shows how calibrating parameters without assessing their identifiability can lead to an uncertain prediction of solute transport in the stream corridor. The new method, therefore, improves the process interpretation of the parameters.While past in-stream experiments of solute transport were mostly conducted during low flow conditions, Chapter 3 presents results from 31 in-stream tracer experiments carried out in the Weierbach over three hydrologic years that comprise both low flow and high flow conditions. The spatially dense groundwater data from Chapter 1 are used to estimate the extent of the groundwater zone receiving streamwater during each experiment, while the iterative modelling approach from Chapter 2 allows robust estimation of the model parameters for each tracer experiment. The analyses show that the streamwater-groundwater exchange has a decreasing influence on water and solute transport in the stream corridor with increasing discharge. This is because of the relatively lower localised water losses from the stream channel to the adjacent groundwater. Model parameter interaction increases with discharge, due to the dominance of advection-dispersion parameters over transient storage parameters.Overall, the thesis advances our understanding of hydrological and solute transport processes in stream corridors. The research highlights the role of both morphological and hydrological factors in stream corridor processes, and the dynamics of near-stream water flow directions during the hydrologic year. The results reported in this thesis pave the way for a holistic understanding of water movement through the stream corridor and contribute to accurate model predictions under different hydrologic conditions.