A time series of measurements of thermal structure and near-bottom currents in a stratified lake revealed both first and second vertical mode internal seiching. The transition into second-mode seiching occurred after first-mode waves, usually generated by pulses of strong winds, died down. We describe the major features of the second-mode seiche and discuss its effects on mixing in the lake below the thermocline. On the basis of calculations of eddy diffusion coefficients, there is only weak mixing in the interior of the hypolimnion.Scaling arguments suggest that mixing at the lake's edges due to motion of the isotherms is not significant despite the larger amplitudes of the secondmode seiche. The most important mixing region appears to be a turbulent bottom boundary layer driven by seiche currents. However, resuspension of bottom sediments did not occur, probably due to a combination of sediment nature and low current speeds.

Kootenay Lake is a long (100 km), deep (140 m), dimictic lake in southeastern British Columbia. Internal waves were observed over an 18‐month period using three thermistor chains moored along the major axis of the lake. To characterize their seasonal aspects, four types of oscillations are recognized: (1) classical seiches with amplitudes of 10–30 m and periods of 1–2 weeks, (2) intrabasin waves with periods of 2–4 days, (3) large‐amplitude waves resembling nonlinear surges, and (4) high‐frequency waves associated with both the internal surge and large‐amplitude internal seiche motions. The periods of the major components of the internal wave field are determined mainly by morphometry but vary seasonally with stratification. Thus the wind, which not only forces internal waves but can also modify stratification, is important. In large lakes like Kootenay, the thermal structure may vary not only spatially but also temporally at periods comparable to those of the internal waves themselves. Hence the shape and frequency of these waves may be altered as they pass through the lake.

The seasonality of physical structure in a deep, temperate lake (Kootenay Lake) is described in relation to its major river inputs (Kootenay and Duncan Rivers). The lake's volume is 37 km3 and its annual outflow is 25 km3 yr-I, yielding a residence time of about 1.5 years. Water mass distributions are controlled by the interactions of three processes: riverine circulation, mixed-layer dynamics, and internal wave behavior. The riverine circulation is determined by the inflow rate and by the relative density of incoming water and ambient lake water. Mixed-layer structure varies in relation to the combined action of the wind and surface heat flux on both seasonal and synoptic time scales. Internal waves may periodically raise a given stratum of water, such as the riverine layer, to a depth where wind mixing is more intense. Knowledge of these processes, their relative importance, and their time-space variability can aid in understanding the ecology of the lake.An important parameter of any natural or man-made lake is its residence time, defined as basin volume divided by outflow rate T, = V/R. In long residence-time systems, wind and local heat exchange are the principal means of setting water in motion (Mortimer 1974;Hollan and Simons 1978). As residence time shortens, the interaction between a lake and its through-flowing river becomes increasingly important, eventually resulting in the so-called run-of-the-river system in which river inputs completely dominate the distribution of properties. In the middle of this spectrum are many lakes whose dynamics are influenced by river through-flow but modified by other physical mechanisms within the lake. As a category, such lakes are still poorly understood. For example, a review of ecosystem modeling of northern lakes by Fox et al. (1979) contains no reference to the processes of lakeriver interaction" We discuss here the annual cycle of mixing and circulation in Kootenay Lake, British Columbia, and show how the seasonal patterns of circulation and water mass distribution are controlled by the interplay of advective and surface-driven processes.

A description is given of the heat and water balances of Kootenay Lake, British Columbia. Although the lake has a moderately high throughflow (781 m 3 s-•), the heat budget is more in character with a lake that is little influenced by rivers. The discussion focuses upon the role of rivers in the thermal history of lakes; a contrast is drawn between Kootenay Lake and Kamloops Lake, British Columbia, a strongly riverinc lake with a much greater heat budget. It is suggested that the riverinc nature of lakes can be characterized using a renewal time that applies to that region of the lake lying above the base of the river plume. Furthermore, the relative roles of advective and surface effects can be qualitatively described using a ratio of surface buoyancy or wind energy to river production of turbulent energy. These parameterizations may prove to be useful descriptors in a lake classification system based upon riverinc quality.

No abstract