U.S. Geological Survey Toxic Substances Hydrology Program--Proceedings
of the Technical Meeting, Colorado Springs, Colorado, September 20-24, 1993,
Water-Resources Investigations Report 94-4015
  
Numerical Simulation of Downward Movement of Solutes During
a Natural-Gradient Tracer Test in Sand and Gravel, Cape Cod, Massachusetts
by
Denis R. LeBlanc (U.S. Geological Survey, Marlborough, Mass.)
and Michael A. Celia (Dept. of Civil Engineering and Operations Research,
Princeton Univ., Princeton, N.J.)
Abstract
A numerical, finite-element, solute-transport model was used to test
the hypothesis that the downward movement of a tracer cloud observed during
a natural-gradient tracer test in sand and gravel on Cape Cod, Mass., was
caused, in part, by density-induced sinking. The tracer solution, which
included the nonreactive tracer, bromide, was 0.1 percent denser than the
ambient ground water. The center of mass of the bromide cloud moved vertically
downward about 3.2 meters during 237 days of transport. The model simulated
density-dependent flow and solute transport along a two-dimensional vertical
section 25 meters high and 136 meters long aligned with the direction of
ground-water flow. Transport of the bromide cloud was simulated for a period
of 237 days divided into 191 time steps. The temporal pattern of recharge
applied to the top boundary of the model was determined from daily precipitation
and estimates of evapotranspiration. On the basis of an analysis of spreading
of the bromide cloud during the tracer test, dispersivity was increased
with time in the simulation asymptotically from 0.05 to 0.96 meters.
The simulated downward movement after 237 days was 2.1 meters (about two
thirds of the observed movement). The simulation showed that density-induced
downward movement was most important during the first 37 days of transport
when the density difference between the ambient ground water and the tracer
cloud was greatest. Earlier work indicated that the difference between the
observed and simulated downward movement may be due, in part, to simulation
in only two dimensions of the three-dimensional flow that occurred around
the tracer cloud as it moved downward through the ambient ground water.
Additional simulations during this and earlier studies showed how the amount
of downward movement is affected by the size and shape of the initial bromide
cloud; aquifer properties, such as dispersivity and anisotropy of hydraulic
conductivity; and the type of boundary used to represent the water table
in the model.
|
|
