GSA Connects 2022 meeting in Denver, Colorado

Paper No. 38-1
Presentation Time: 1:40 PM

PROGRESS TOWARDS AN UPDATABLE MEDIUM-SCALE BEDROCK TOPOGRAPHY MAP OF ILLINOIS


PHILLIPS, Andrew1, CARRELL, Jennifer2 and BALIKIAN, Riley2, (1)Illinois State Geological Survey, University of Illinois at Urbana-Champaign, Champaign, IL 61820, (2)Illinois State Geological Survey, Prairie Research Institute, University of Illinois at Urbana-Champaign, 615 E Peabody Dr., Champaign, IL 61820

The existing statewide small-scale (1:500,000) bedrock topography map of Illinois was modeled in 1994. We are updating this with results of more recent large- and medium-scale (1:24,000 – 1:100,000) mapping, an ever-growing digital database of water and engineering wells, digital bedrock soils data, lidar surface elevations for areas with bedrock exposure, and maps from surrounding states. Our initial concept was to carry results of the large-scale mapping via contours while largely relying on an unsupervised classification of wells for unmapped areas. The unsupervised classification of well logs provides ~75,000 acceptable bedrock top records out of ~150,000 locations, although more are expected with additional filtering. The well dataset includes significant uncertainty where elevations are assigned based on unverified location information provided by the source.

Using a mixture of contour and point data with streamlines to constrain hydrology is conceptually pleasing within an ANUDEM model, a spline-based model we have used successfully for many years for large-scale mapping of bedrock terrains across Illinois. However, our small-scale models have proven less satisfactory most notably because of wild oscillations along boundaries where contours break, very different results in areas with only point data compared to areas with only contour data, and treatment of bullseyes caused by the uncertain well dataset. We also have recognized seemingly unresolvable local inconsistencies where new mapping of adjacent areas occurred based on apparently different conceptual models of the bedrock topography.

We plan to resolve these issues through both modified data treatments and modeling approaches. New mapping will be carried as point datasets derived from gridded datasets, and we continue experiments with parameter tuning in ANUDEM, natural neighbor interpolation of point data, Empirical Bayesian Kriging, and separating physiographic regions of the bedrock surface for targeted treatments. We work towards a “living” statewide bedrock topography map of Illinois that can readily incorporate new data as it becomes available and can be used to provide on-demand small-scale maps.