Sediment resuspension in a shallow lake.
Eu Gene Chung, Fabian A. Bombardelli, and S. Geoffrey Schladow

Many shallow lakes around the world have increasing loads of nutrients, heavy metals, and other toxic substances [Linge and Oldham, 2002; Romero et al., 2002]. Sediment-water interactions in those lakes become increasingly important with time, since the bed sediments act as repositories for the added nutrients and toxic substances, which may eventually be released into the overlying water column. The upper layers of the bed sediments of shallow lakes participate in the exchange of substances with the water column through physical, chemical and biological processes and, very importantly, via sediment resuspension and horizontal sediment transport [Garcıa, 1999]. These upper layers can thus influence the cycling of nutrients, heavy metals and organic micropollutants in shallow lakes and reservoirs for a long time [Blom et al., 1992].

The prevalence of sediment resuspension in shallow, wind-exposed lakes has resulted in the development of many parametric models attempting to describe the relation between the hydrodynamics of water bodies and sediment resuspension rates [Sheng and Lick, 1979; Aalderink et al., 1985; Luettich et al., 1990; Hamilton and Mitchell, 1996; Admiraal et al., 2000; Sanford and Maa, 2001; Mian and Yanful, 2004; Cozar et al., 2005]. The rate of sediment resuspension (or entrainment rate of sediment into suspension) is quantified by Φ_E, the vertical flux of solid particles close to the bottom. Dividing the entrainment rate by w_s, the terminal particle fall velocity in quiescent fluid, yields E_s, the nondimensional coefficient for entrainment of bed sediment into suspension [Garcia and Parker, 1991, 1993; Parker, 2004; F. A. Bombardelli and M. H. Garcia, Numerical simulation of wind-induced resuspension of bed sediment in sallow lakes, paper presented at the International Water Resources Engineering Conference, American Society of Civil Engineers, Seattle, Washington, 1999]. When the entrainment process is in equilibrium [Admiraal et al., 2000; Parker, 2004], E_s = c_ae, where c_ae is the equilibrium near-bed sediment concentration (generally expressed in volume of sediment per unit volume of fluid, and measured at a certain distance a from the bottom).

Most of the available models to compute E_s or Φ_E agree upon establishing a dependence of the entrainment rate on the bottom shear stress induced by the motion of the attendant fluid, and also on the local sediment characteristics, usually through a given measure of the sediment size [Sanford and Maa, 2001]. Aalderink et al. [1985] put forward the following relation for sediment resuspension: Φ_E ~ w^K, where w indicates the wind speed at 10 m above the surface and K is an exponent which varies with the wind speed. Mehta et al. [1982], Raudkivi [1998] and Sanford and Maa [2001] reviewed a large number of formulae for the erosion rate of cohesive sediments, which can be subsumed under general expressions shown in Appendix A (see equation (A1)). There is also a vast body of formulations to compute the entrainment rate of sediment in open-channel flows [Garcıia and Parker, 1991; Garcıa, 1999]. Most of the relations for E_s are of the following type:

E_s ~ Φ_E ~ t_^Pb ~ u_*^2P ~ w^K

where t_b and u_* are the bed shear stress and the wall friction (shear) velocity due to skin friction, respectively, and P and K are empirical exponents. These relations for open-channel flows have been obtained under steady state, equilibrium conditions (or, moderate disequilibrium conditions), for essentially unimodal, noncohesive sediment particles, and for a uniform distribution of the shear stress in space. One such relation is the expression by Garcia and Parker [1991, 1993] for noncohesive sediment, also presented in Appendix A (see equation (A2)).

On the basis of a thorough review of the literature on the subject of sediment resuspension, the relations coming from open-channel flows have not been tested in lakes. In spite of the inherent differences between the boundary layers in open-channel and lake flows, this is somewhat surprising, especially considering the fact that these expressions have been derived mainly using dimensional analysis. Moreover, it turns out that the formulations for the computation of the sediment entrainment rate can produce widely disparate results under the same set of conditions, which adds to the inherent complexity of understanding sediment-related processes in lakes.

On the other hand, to observe sediment resuspension, sediment concentrations are usually measured close to the bottom, under the hypothesis that quasi-equilibrium conditions are prevalent. This measurement is undertaken either continuously, by using beam transmissometers or optical backscatter (OBS) sensors [Gloor et al., 1994; James et al., 1997; Jin and Wang, 1998; Weyhenmeyer et al., 1995], or in an integrated way by using sediment traps [Evans, 1994; Jurg, 1996; Kozerski, 1994; Lindstrom et al., 1999; Rosa et al., 1983].

In this paper, the results of a 4-month field experiment on sediment resuspension in a large, shallow, hypereutrophic lake, the Salton Sea in southern California, are presented. The field measurement program included the observation of currents and waves using a Nortek acoustic wave and current (AWAC) profiler, the observation of water temperatures using a thermistor chain, the use of OBS sensors for the measurement of turbidity, a surrogate for suspended sediment concentration [Gippel, 1989; Suk et al., 1998; Cozar et al., 2005], and the use of meteorological data. The paper reports one of the first field campaigns in lakes using the AWAC, whose signal strength has not been investigated in detail to date, and the correlation of a set of variables coming from different sources during a relatively long period of time (section 2.1; cf. Table 1). In section 3.1, the contribution of different forcing mechanisms (waves, currents and surface seiches) to the sediment resuspension in the lake is quantified, and the signal strength of the AWAC is used to address the vertical distribution of sediment in the water column. Section 3.1 also puts forward novel relations between the AWAC’s backscatter intensity and turbidity. Finally, section 3.2 presents the modeling of sediment entrainment into suspension in the Salton Sea by introducing an extension of the Garcia and Parker formula.