2006 Philadelphia Annual Meeting (22–25 October 2006)

Paper No. 13
Presentation Time: 6:00 PM-8:00 PM

DRESSING THE EMPEROR: THE ROLE OF THREE-DIMENSIONAL INFORMATION VISUALIZATION SOFTWARE IN THE DEVELOPMENT OF THREE-DIMENSIONAL HYDROGEOLOGIC MODELS


MEDINA, Cristian R.1, OLYPHANT, Greg A.1 and LETSINGER, Sally L.2, (1)Geological Sciences, Indiana University, Center for Geospatial Data Analysis, 1001 East Tenth Street, Bloomington, IN 47405, (2)Center for Geospatial Data Analysis, Indiana University, Indiana Geological Survey, 611 North Walnut Grove, Bloomington, IN 47405, crmedina@indiana.edu

The goal of this research is to develop a model that describes the saturated and unsaturated groundwater flow in Berrien County, Michigan (1,350 km2), an area containing a complex sequence of glacio-lacustrine deposits. Stone and others (2001) mapped the morphosequences in Berrien County at a scale of 1:24,000, which includes georeferenced structure contours for 20 individual units. We have developed a methodology to translate this detailed morphostratigraphy into a solid three-dimensional geologic model, and then into a three-dimensional block of data that can be used as input to a finite-difference groundwater-flow model. Letsinger and others (2006) describe the process of using geographic information system software to convert the structure contours into georeferenced raster layers that describe each unit. At this stage of the reconstruction, only the bounding surfaces between the units are defined. In order to stack the units in vertical space using customized computer code, a “virtual well field” (regularized two-dimensional array of points) samples each x-y location in each of the 20 rasterized data layers. Units that are intersected from the top bounding surface (surface topography) to the bottom bounding surface (bedrock surface) are then identified. The result of this step is a vector (one-dimensional array) at each virtual well location that describes the elevation of each morphostratigraphic unit boundary intersected at that location. However, at this stage, the model is essentially a regularized three-dimensional point cloud, and three-dimensional information visualization software (3DIVS) is then utilized to generate a solid geologic model by interpolating the vertical geologic “samples” throughout the model domain. A finite-difference grid (“brickpile”) at the chosen resolution of the groundwater-flow model is then generated from the solid geologic model using data-processing functions of the 3DIVS.