Transport and Mixing Processes in Lakes and Reservoirs - Hydro-Environmental Analysis: Freshwater Environments - page 359

Underlows are common in the fall. Shallow riverine waters cool faster than lake waters and
fall storms cause higher turbidity inlows. In lakes, underlows will tend to follow the deep-
est channel, spreading out and dissipating due to the turbulence generated by bottom shear.
In reservoirs, underlows tend to follow the thalweg, or drowned river channel. Velocities will
decrease as the underlow moves along the bottom, due to the shear between the low and the
bottom. As the current velocities decrease, suspended solids may settle out, causing turbidity
currents to lose their identity. However, in reservoirs with relatively steep bottom slopes, tur-
bidity currents may low downward to the dam, resulting in problems in the design and opera-
tion of outlet structures (Smith 1975). Density currents may also reach the dam but generally
do not pose problems like the sedimentation that occurs when a turbidity current reaches outlet
structures.
If the density of the underlows is greater than the bottom waters, they may remain on the bot-
tom. However, most underlows are less dense than the bottom layer. In such a case, the underlow
separates from the bottom as an interlow (either density or turbidity; Figure 13.10). Interlows are
not affected by the bed shear, as with underlows.
Density and turbidity currents have been reported for many lakes and reservoirs. For example,
Anderson and Pritchard (1951) reported winter underlows and spring overlows, along with sum-
mer and fall interlows of the Colorado River into Lake Mead. Lehn (1965, cited by Wetzel 1975,
2001) observed a rapid attenuation of light in the vicinity of the metalimnion due to the intrusion of
river water with high suspended loads (turbidity currents) into Lake Bodensee, Germany. Ford and
Johnson (1981) reported density and turbidity currents on 10 occasions in DeGray Lake, Arkansas.
In many small lakes and reservoirs, the relative importance of the density and turbidity currents is
undetectable (Smith 1975).
Underlows, interlows, and overlows will typically move down-reservoir as long as additional
riverine inlows of similar characteristics can push it forward. In reservoirs, the formation and
movement of underlows and interlows may be enhanced by subsurface withdrawals. Studies of
DeGray Lake, Arkansas (Martin 1987), suggested that summer midlayer dam releases enhanced
interlows above the thermocline. The movement of materials from anoxic zones at the head of the
reservoir along the interlow contributed to a metalimnetic dissolved oxygen minimum and affected
the quality of reservoir releases.
In general, however, underlows and interlows add oxygen-consuming materials to the more
poorly mixed and less productive regions of the reservoir, thereby contributing to dissolved oxygen
depletion in the bottom waters (Ford and Johnson 1986). Density differences result in an effective
isolation of the riverine waters so that they may become anoxic and chemically reduced. Interlows
and underlows are often a major source of reduced materials to downstream reservoir regions (Nix
1981, 1987; Martin 1987). Alternatively, if the underlows are well oxygenated, they may improve
the bottom water quality. Ford and Johnson (1986) found that cold, oxygenated inlows from Lake
Ouachota, Arkansas, maintain anoxic hypolimnion in downstream Lake Hamilton, although coves
within the reservoir become anoxic.
An estimation of the extent of the initial mixing due to inlows, and of the possible formation
of density or turbidity currents, is often necessary to determine the transport of materials through
lakes and reservoirs, to compute release concentrations, and to compute the impact of the inlows
on ambient quality. Failure to consider the formation of density currents, for example, may result
in an overestimation of the dilution of the inlows and their residence time within the lake or
reservoir. High turbidity interlows in Lake Cumberland, Kentucky, have been observed to essen-
tially “short circuit” the reservoir. That is, these currents have remained relatively intact through
the reservoir so that estimates of release concentrations based on assuming mixing with the res-
ervoir waters would be inaccurate. Similarly, Worden and Pistrang (2003) observed shortcuts in
the Wachusett Reservoir, essentially connecting releases from the upstream Quabbin Lake to the
outlow in a metalimnetic “short circuit” undergoing minimal mixing with the ambient Wachusett
Reservoir water.