HYDROSTRATIGRAPHIC CORRELATION – OR THE LACK OF IT – IN MULTI-SCALE INVESTIGATIONS IN GLACIAL DRIFT

The accuracy and applicability of hydrogeologic assessments of glacial drift depend on the quantity and quality of data support, which are generally insufficient given the complex arrangement of Quaternary aquifer and aquitard deposits. While regional studies are useful in depicting broad stratigraphic relationships, they are not designed for assessing groundwater flow at more focused scales.

Geostatistical methods provide a means of quantifying the vertical and lateral correlation of subsurface units. This study considers the application of two techniques: binary indicators for permeable vs. impermeable categorization, and transition probability geostatistics for handling a larger set of material types. The analysis is focused on a 4.5-county portion of central Minnesota. During Late Wisconsinan glaciation, the region was invaded by five ice lobes of different provenance. Maps of glacial landform/sediment assemblages are available from the Minnesota Department of Natural Resources and additional maps of glacial sediments and stratigraphy have been compiled by the Minnesota Geological Survey. The study area contains over 11,000 logged wells, mostly consisting of water well data. Geostatistical analysis was carried out on well logs grouped according to the specific glacial landform/sediment assemblages. The results identified the relative differences in vertical and lateral correlation among the various geomorphological units. In addition, high-quality rotasonic logs from several locations provide a point of comparison to detailed site-specific correlation relationships.

The results of this study may guide data collection strategies for regional or local studies, assess uncertainty, and support a stochastic approach to groundwater flow modeling in glaciated terrain.