Ps. Hill et Arm. Nowell, COMPARISON OF 2 MODELS OF AGGREGATION IN CONTINENTAL-SHELF BOTTOM BOUNDARY-LAYERS, J GEO RES-O, 100(C11), 1995, pp. 22749-22763
Comparison of two models of aggregation and disaggregation of fine sed
iment in continental-shelf bottom boundary layers delineates condition
s under which these processes can be modeled simply. Model 1 predicts
the evolution of the particle size distribution based on layer-average
d mean sediment concentrations. Model 2 avoids the simplifying assumpt
ions of model 1 by representing suspension dynamics with a Monte Carlo
simulation of the joint probability density function of number concen
tration in the various particle size classes. In contrast to model 1,
this technique bases aggregation rates on local concentrations and exp
licitly treats the effect of concentration correlations on aggregation
rates. Each model produces a time series of particle size distributio
n in a steady, horizontally uniform, nondepositing, continental-shelf
bottom boundary layer. Comparisons are made for suspensions that are i
nitially polydisperse and monodisperse, for shear velocities of 0.01 a
nd 0.005 m s(-1), for fractal dimensions of 1.92 and 2.4, for initial
layer-averaged sediment concentrations of 1, 2.5, 10, and 25 x 10(-3)
kg m(-3), and for sticking efficiencies of 0.1 and 1.0. In general, mo
del 1 is accurate for suspensions that are initially polydisperse and
do not develop strong vertical structure. Results suggest that as long
as maximal settling velocity does not exceed roughly 0.6u, vertical
gradients in sediment concentration do not degrade substantially the a
ccuracy of layer-averaged mean models for concentrations typical of th
e continental shelf. When initial suspensions are monodisperse and/or
vertical gradients are large, transport-limited aggregation can slow o
verall aggregation rates of model 2 relative to model 1, or important
interactions can occur between particle types that both show strong ve
rtical structure, speeding overall aggregation rates in model 2. In th
e future the effects on the accuracy of layer-averaged mean models of
short- and long-term unsteadiness, erosion, and deposition will be exp
lored.