APPLICATION OF A MORPHOLOGIC ELEMENT INTERPOLATOR TO MAPPING GEOLOGIC SURFACES

This presentation describes experience using a Morphologic Element Interpolator to model the buried bedrock surface and the stratigraphic contacts of Dakota County, Minnesota. With this interpolator the modeler can input knowledge of geomorphology as well as point observation data. For example, the modeler may specify the geometry of a dendritic erosional surface and the interpolator will produce a surface that fits the observation data with a dendritic surface of the specified geometry. The morphologic descriptions are saved, and can be reused or updated to produce modified or updated products.

The application presented is an extension of the Analytic Element Method of Dr. Otto Strack, implementing point, line, and areal functions. The functions are called morphologic elements because they are assembled to simulate morphologic features. Elements of different type or scale can be superimposed.

A GIS environment was used to visualize the observation data, interpret geomorphology, and define the morphologic elements. Linear elements of specified sharpness were assembled to represent valleys, ridges, terraces, bending, and faults. Input for the interpolator was derived directly from the GIS. The interpolator was used to produce grids and contour lines.

A two step modeling process was used. The first step was to produce conceptual models, in which the only functions were the morphologic elements, and the observation data was only used in fitting. This step was essential for refining the geomorphologic interpretations. These conceptual surfaces matched the interpretations but fit the observation data only approximately.

The final step was to superimpose the actual observation point data as multi-quadric point functions onto the conceptual models. This produced surfaces that fit all of the observations exactly, yet retained the geomorphologic interpretations.

The conceptual model surfaces are relatively smoother, and are useful for groundwater flow modeling because they add less noise to the results.