Quantifying Turbulent Mixing in Plunging River Inflows: Insights from Field Measurements in Lake Geneva

Hyperpycnal river inflows discharging into lakes or reservoirs will plunge and trigger gravity-driven underflows near the bed. Such underflows are called turbidity currents if the density excess is mainly caused by a high sediment concentration. These underflows can reach the bottom of the lake or, alternatively, detach from the bed and intrude horizontally into the lake waters to form an interflow if a layer of equal density is encountered. As underflows carry a number of constituents fed to them by the river or eroded from the bed, such as sediment, contaminants, nutrients and oxygen, their pathway and final destination have an impact on lake or reservoir water quality. The mixing processes in the plunging region are known to be of dominant importance for the dilution of the inflowing river water by entrainment of ambient water, thereby exerting a primary control on the final intrusion depth of underflows. Understanding and quantifying these processes is therefore key. Until now, the majority of estimations of the mixing extent in the plunging region were made within laterally confined laboratory experiments or via passive tracer methodologies. This study focuses on quantifying plunging mixing from flow velocity measurements in a laterally unconfined river inflow in the field and investigating its dependency on inflow conditions.

Field measurements of the plunging Rhône River entering Lake Geneva were conducted using a boat-towed ADCP along a grid of transects for six inflow conditions characterized by the inflow densimetric Froude number Fr_d. The plunging mixing coefficient E_p, which compares the underflow discharge immediately post-plunging to the initial river inflow discharge, was used to quantify the plunging mixing.

Results indicate that for larger Fr_d values (Fr_d > 4) E_p estimates align with laterally confined lab experiments (E_p = O(0.1)). Conversely, for smaller Fr_d values E_p estimates correlate with field tracer measurements of laterally unconfined inflows (E_p = O(1)). E_p decreases with increasing Fr_d, challenging existing numerical simulations predicting the opposite relationship.

This study offers key insights into turbulent mixing rates associated with hyperpycnal river inflows and highlights the need to incorporate realistic field conditions for accurate modeling of plunging river inflows and intrusion depth.