REGIONAL SCALE GROUNDWATER FLOW: THE IMPORTANCE OF FLUID DENSITY AND DIMENSIONALITY

The conceptual model of groundwater flow put forward in the initial papers by Jozsef Toth considered systems with fresh water. However, the deep pore fluids in the crystalline rock of the Canadian Shield and the sedimentary sequences of the Michigan Basin are characterized by total dissolved solids (TDS) concentrations that can exceed 200 g/l. The impact of these high density fluids on regional groundwater flow was investigated in two case studies using the model FRAC3DVS. In the first case study, an analysis of a 5,734 km2 watershed situated on the Canadian Shield was conducted. The second case study investigated flow in the part of the Michigan Basin in south-western Ontario. The 19,000 km2 domain included stratigraphic units from the Precambrian to the Devonian.

For the Canadian Shield setting, the groundwater flow system was assessed by exploring its sensitivity to topography, spatially variable matrix permeability distributions, pore water salinity and the dissipation of elevated initial pore pressures from the last glacial period. Steady-state groundwater flow analyses indicate that piezometric heads in all model layers are highly correlated to the complex surface topography. Flow changes from zones of groundwater recharge to discharge over relatively short distances. This variation coupled with a rock permeability that decreases significantly with depth results in a groundwater system where shallow flow to a depth of 10s of meters dominates the overall water balance. The flow in the deeper rock is not regional. Groundwater was not predicted to underflow the major rivers and their tributaries.

The analysis of the Michigan Basin supports the fact that the dense, high TDS groundwater that occurs in the Silurian and deeper formations is millions of years old with transport being diffusive. Groundwater flow of the low TDS pore fluids in the Devonian units and surficial drift is controlled by topography, agreeing with the conceptual model put forward by Jozsef Toth.

For both case studies, the impact of fluid density is to create a more active shallow zone with lower TDS, and a sluggish deeper zone with high TDS. This study has preserved the complexity of the watershed-scale groundwater system; flow interpretations that are based on geometrically simplified conceptual models may yield physically implausible results.