The flow of rivers into lakes: Experiments and models
Thesis
This study considers the motion of river water flowing into a lake. The study has two parts: field observations of an interflowing river in Lake Iseo (Italy); and laboratory experiments on basins fed by an underflow. In both, observations are compared to existing and new models. There is significant uncertainty in predicting contaminant transport from inflows to lakes. Reducing the uncertainty contributes both to understanding lake ecosystems and protecting water resources. In field observations, the progressive stages of the inflow were interpreted as a laterally falling plunge region with little mixing, an underflow region with substantial mixing, and finally an intrusion. A simple theory for a laterally falling plunge region was developed, giving the ratio of offshore extent of the plunge region to river width as equal to the initial densimetric Froude number. This theory agreed with the observed plunge location better than existing models. In laboratory experiments, the evolving density stratification in a basin fed by underflowing gravity currents was measured. The transport pathway of the inflowing fluid was visualised with dyes. Gravity currents underflowing uniform ambient uid over an intermediate range of source Reynolds number (Re_s) at constant Richardson number (Ri) had entrainment coefficients (E) that varied linearly with Res. The values for E were within the previously observed range, but were lower than previously observed for the Ri used here, because Re_s was lower here than in previous studies. Dye pulses showed that gravity currents in stratified ambients did not only entrain, as occurs when in uniform ambient fluid, but also effluxed fluid over the whole depth of the stratified ambient. Analytical expressions were developed for the evolving stratification generated by efflux from the gravity current. The new efflux model predicted the observed stratifications more closely than existing entrainment models. The efflux model could better match observations by including the continued mixing that occurs within the stratified ambient and including the undiluted sublayer observed within the gravity current. Future work should quantify efflux from gravity currents and incorporate the efflux mechanism into full lake basin simulations. Given the deviation of existing entrainment models from observations shown here, a better model of gravity current efflux would improve modelling of lakes and a variety of buoyancy driven flows, from building ventilation to ocean circulation.