No abstract

No abstract

Abstract:A numerical model was developed for simulating groundwater-wetland interactions and contaminant transport. The model calculates transient hydraulic head and a transient free surface in a two-dimensional, heterogeneous domain, with variable and transient boundary conditions (infiltration, evapotranspiration, surface water), and water and contaminant fluxes across the aquifer-wetland interface. Contaminant transport is also simulated, with contaminant sources located at the free surface, wetland, or within the saturated domain. The model was applied to assess groundwater-wetland interactions and the transport of septic-system-derived contaminants at Point Pelee, Ontario, Canada. The model successfully simulated the field observations of groundwater flow and contaminant plumes. Where the barrier bar is narrow, the seasonal reversal in the direction of groundwater flow is caused by differences in the elevation of the water surface of Lake Erie and that of the marsh. This, in turn, induces the contaminants to oscillate between movement towards the lake during the winter and towards the marsh during the summer. Hence, contaminant plumes are bimodal in shape. Where the barrier bar becomes wider, the lake and the marsh have less effect, and hence contaminants move in one direction, along the principal direction of groundwater flow towards the marsh.

Groundwater flow regimes adjacent to coastal wetlands of the Great Lakes are highly transient and vary among different types of coastal wetlands. Groundwater flow is controlled by (1) the physiography of the land adjacent to the wetland, (2) the relative elevations of the wetland and the lake, as they fluctuate over time, and (3) the amount of infiltration and evapotranspiration that occurs at the land and wetland. Groundwater from the mainland adjacent to a wetland will flow towards and discharge into the wetland throughout the year. In a spit that partially protects a wetland from a lake, the source of groundwater is precipitation and snowmelt; not water from the lake or wetland. Here, groundwater continually flows from either side of a central groundwater divide towards the lake or wetland, with the elevation of the lake only affecting the rate of groundwater drainage. Because barrier bars completely separate a lake from a wetland, the elevation of the lake and wetland are different. When the barrier bar is narrow, the resultant hydraulic gradient across the barrier bar causes groundwater flow to oscillate between flowing towards the lake during the fall and winter and towards the wetland during the spring and summer. But as the width of the barrier bar increases, the impact of the lake and wetland diminish relative to the amount of precipitation and snowmelt infiltrating into the barrier bar. Thus, the groundwater flow regime is characterized by a central groundwater divide with groundwater on either side continuously flowing towards the lake and wetland throughout the year. Intradunal wetlands are actually several small wetlands within a series of relic beach ridges and parabolic dunes. Groundwater flow regimes here are highly variable and transient with flow adjacent to different wetlands, and at different times of the year, exhibiting continuous flow to a wetland, oscillating direction of flow, and lateral migration of the groundwater divide. However, these groundwater flow patterns are caused by precipitation and evapotranspiration within the wetland complex and not by fluctuating lake levels.

Density‐induced advection of organic vapors has been demonstrated to be a significant transport process in granular porous media if the permeability is sufficiently high such that density‐induced advection dominates over vapor diffusion. In the context of partially saturated fractured media, the presence of a network of open fractures may lead to rapid rates of density‐induced advection of the vapors in the fractures, even though the matrix may have low permeability. To investigate the various factors which affect vapor migration in discretely fractured porous media, a two‐dimensional finite element model has been developed whereby the porous matrix is represented by rectangular elements and the fractures are represented as one‐dimensional line elements which are superimposed onto the rectangular grid. The cross‐sectional model includes the processes of advection due to density and pressure gradients, and vapor and aqueous diffusion in both the fractures and the porous matrix. Phase partitioning between the vapor and aqueous phases and the aqueous and solid phases is assumed to be at equilibrium. Results from simulations involving a single vertical fracture indicate that density‐induced advection decreases in importance as the fracture aperture decreases. In this case, there appears to be a critical fracture aperture, above which density‐induced advection is the dominant process, and below which vapor diffusion dominates. Simulations that include a network of horizontal and vertical fractures show the importance of matrix properties such as air porosity and water content. An increase in the matrix air porosity results in a higher storage capacity for vapor‐phase contaminants and allows more diffusion to occur from the fractures to the matrix, whereas a higher matrix water content results in an increase in the degree of phase partitioning between the vapor and the water phases. Although these two processes act to retard the migration of the vapor plume, they can be problematic in the context of vapor extraction in fractured porous media. The reason for this is that these processes tend to increase the amount of contaminant mass that exists in the low‐permeability matrix. As a result, vapor extraction in fractured porous media can be very difficult.