Controls on Sediment Suspension, Flux, and Marsh Deposition near a Bay-Marsh Boundary

Abstract

The sustainability of marshes adjacent to coastal bays is driven by the exchange of sediment across the marsh-bay boundary, where edge erosion commonly leads to lateral marsh loss and enhanced vertical accretion. The timing and patterns of sediment deposition on salt marshes adjacent to larger bodies of water such as coastal bays, however, differ from those on better-studied tidal creek marshes primarily owing to the importance of wind-waves. We combined field measurements and modeling to examine controls on suspended sediment concentrations and fluxes on a tidal flat (tidal range of 1.2 m) and rates of sediment deposition on the adjacent marsh at a site on the Eastern Shore of Virginia. Suspended sediment concentrations over tidal flats were strongly controlled by waves. Yet, storm winds sufficient to drive large resuspension events often coincided with peak tidal elevations that were too low to flood the marsh, which was oriented away from the wind directions most favorable for storm surge, thereby restricting storm-driven, episodic sediment delivery to the marsh. Winds also drove wide variability in the direction of surface currents near the marsh edge when water depths were high enough to flood the marsh. Nevertheless, our results show that sediment in the upper water column over the tidal flat was effectively transported across the marsh edge during flooding tides. A sediment deposition model developed to investigate the combined effects of vegetation and wave action on depositional patterns predicted that waves displace maximum deposition inland from the marsh edge, consistent with measured deposition at the study site. Marsh deposition was sensitive to inundation frequency as well as the concentration of sediment in water flooding the marsh, underscoring the importance of nontidal controls on water surface elevation, such as meteorological effects (e.g., storm surge) and sea level rise. Whereas short-term increases in marsh inundation enhance deposition, sea level rise that results in deeper average water depths over the tidal flats decreases deposition if marsh elevation is rising in step with sea level.

Appendix

Current-generated bed shear stress, τ_curr, was calculated using the expression:

$$ {\tau}_{curr}={C}_D\rho {u}^2 $$

where ρ = 1020 kg m⁻³ is water density, u is current speed, and C_D is the drag coefficient, estimated as:

$$ {C}_D=g{n}^2/\left({h}^{1/3}\right) $$

where n is the roughness coefficient

$$ n={\left[\frac{\sqrt{8g}}{h^{1/6}}\left(2{\log}_{10}\left(\frac{h}{D_{84}}\right)+1\right)\right]}^{-1} $$

(Hornberger et al. 2014; Lawson et al. 2007), h is water depth, g = 9.81 m s⁻², and D₈₄ is the 84th percentile of the grain size distribution.

Wave-induced bottom orbital velocity, u_b, was calculated as:

$$ {u}_b=\frac{\pi {H}_s}{T\;\sinh (kh)} $$

(Wiberg and Sherwood 2008) and wave-generated bed shear stress, τ_wave, was estimated as:

$$ {\tau}_{wave}=0.5{f}_w\rho {u}_b^2 $$

where

$$ {f}_w=0.04{\left(\frac{u_bT}{2\pi {k}_s}\right)}^{-0.25} $$

(Fredsoe and Deigaard 1992), H_s is significant wave height, T is wave period, k is wave number (2π/L), L is wave length, f_w is the wave friction factor, and k_s = 3D₈₄ is the roughness length scale of the bed. Total bed shear stress was calculated as the sum of wave and current shear stress.

To estimate suspended sediment concentrations, C_s, throughout the full water column, the Rouse equation (Rouse 1937) was applied using 3 grain-size fractions (7μm (w_si = 3x10⁻⁵ m s⁻¹); 25 μm(w_si = 4 × 10⁻⁴ m s⁻¹); 100 μm (w_si = 0.005m s⁻¹))

$$ {C}_{s_i}={C}_a{\left(\frac{z\times \left(h-{z}_a\right)}{z\times \left(h-z\right)}\right)}^{r_i} $$

where r_i =  − w_si/(κu_∗curr) is the Rouse parameter for each grain size fraction, i, w_si is the particle settling velocity for each size fraction, u_∗curr, is current shear velocity, κ is von Karman’s constant (0.41), and z is the height in the water column at which C_si is being estimated. C_a is the reference concentration at the reference height at the level z_a. When turbidity measurements are available, C_a is taken as the suspended sediment concentration estimated from measured turbidity and z_a is the height of the turbidity sensor. When turbidity measurements are not available, we estimated C_a as

$$ {C}_a={C}_{bed}\frac{\gamma S}{1+\gamma S} $$

(Smith and McLean 1977), where S = (τ_b − τ_cr)/τ_cr is the excess shear stress determined from τ_b, the total bed shear stress exerted by waves and currents, z_a = 3D₅₀, D₅₀ is the median grain size, and C_bed = 0.3 is the concentration of sediment in the bed (1.0 – porosity), consistent with a muddy bed (Wheatcroft et al. 2007). Critical shear stress was determined to be τ_cr = 0.07 Pa from a plot of NTU versus total shear stress at site 2 (Online Resource 2). This agrees with values based on erosion rate measurements from Lawson (2004). We set the value of the resuspension coefficient γ = 5e⁻⁴, by scaling the estimated SSC to match the measured SSC. Field and laboratory studies have shown large variation in values of γ, ranging from 10⁻² to 10⁻⁵ (e.g., Smith and McLean 1977; Wiberg and Smith 1983; Sternberg et al. 1986; Hill et al. 1988; Drake and Cacchione 1989).