Wind‐Induced Versus Plume‐Induced Inter‐Basin Exchange—Resolving Causal Influences in Plume‐Lake Modeling

Abstract Bubble‐plumes are commonly used in lake restoration to alleviate problems associated with hypolimnetic hypoxia. In lakes with multiple basins separated by sills, bubble‐plumes have been used locally to boost oxygen levels in individual basins. However, they have the potential to affect inter‐basin exchange leading to unexpected results. Our goal is to assess the relative importance of natural versus plume forcing as drivers of exchange. This is critical to evaluate the field‐scale performance of bubble‐plume systems. We hypothesize that the contribution of bubble‐plumes as drivers of inter‐basin oxygen transport depends on the depths of detrainment and maximum plume rise relative to sill level (plume geometry). To test this hypothesis, 1D integral bubble‐plume models coupled to a 3D‐hydrodynamic model are applied to simulate the performance of an oxygenation system installed in 1990 in the north basin of Amisk Lake, Canada. Sources of uncertainty in bubble‐plume modeling, associated with model assumptions and parameter values (structural and parametric uncertainty), are systematically analyzed, and their effect on plume‐structure and inter‐basin exchange predictions are quantified. The effects of plume forcing on exchange rates and patterns are only significant as the equilibrium depth rises above the sill. For the prevailing conditions in the study case and the most widely accepted plume model, this occurs with low probability. More plausibly (for most parameter combinations), the plume injects oxygen below the sill, the oxygenated water being transported by internal waves between basins. Solid conclusions on the dominant drivers of large‐scale transport arise in attribution studies accounting for model uncertainty.