FUSION OF LITHOSTRATIGRAPHIC, ALLOSTRATIGRAPHIC, AND HYDROSTRATIGRAPHIC APPROACHES TO CHARACTERIZE A GLACIAL AQUIFER SYSTEM
Deterministic architectures may be established using a traditional lithostratigraphic correlation approach to specify the arrangement of texturally similar sediment bodies, or, alternatively, can be interpreted using allostratigraphic approaches to identify chronostratigraphically significant surfaces that divide sediments into genetically-related packages. Site-specific, field-scale groundwater flow and contaminant transport investigations often require additional detail that can be modeled stochastically, using geostatistics, within the larger deterministic framework. Application of geostatistics requires that assumptions of stationarity are not violated; hence the need to work within a sensible deterministic framework at an appropriate scale.
Moreover, the characterization of spatial variability within physically heterogeneous porous media can be confounded by the commensuration of the term ‘heterogeneous’ with the concept of ‘spatial variability’. Consequently, the conceptual terms uniform, nonuniform, homogeneous, and heterogeneous proposed by Greenkorn and Kessler (1969) and applied by Freeze (1975) will be discussed in the context of hybrid stochastic and deterministic hydrostratigraphic modeling and illustrated using an example from the Fort Wayne Moraine west of Ann Arbor, Michigan, USA. In this example, allostratigraphic correlation methods, constrained by static water level and contaminant concentration measurements, have been employed to construct a deterministic hydrostratigraphic architecture that will be supplemented by stochastic modeling of aquifer and aquitard parameters.