INTERPRETING LANDFORMS FROM REMOTELY SENSED IMAGERY TO INFER SUBSURFACE PROPERTIES

Seismic, acoustic, and electromagnetic sensor networks are critical components of the Army’s new battlefield initiatives. Since seismic and acoustic signals follow curving ray paths, they provide essential non-line-of-sight detection and tracking capabilities. However, this attribute also makes these signals highly sensitive to environmental and meteorological conditions. This sensitivity must be considered when deploying sensor networks, in analyzing their performance, and adapting each sensor to local geological conditions. Changes in near-surface geologic boundary conditions and material properties have first order effects on surface wave modes, which are the primary signals of interest. Our primary thrust is thus to define the logic of and develop a preliminary working model for predicting near-surface conditions.

We will develop an inferential geologic model using simple geological rules of thumb, remotely sensed imagery, and GIS databases. The model will extract geophysical parameters from relational databases to develop three-dimensional geologic “block diagrams” for seismic/material-properties models that are used to calibrate seismic sensors. The overall scientific objectives are to: 1) define procedures for rapid landform discrimination from remotely sensed imagery; 2) use these data to infer the geologic and material properties of the landforms; 3) produce a GIS-based conceptual geologic model to define three-dimensional seismic property distribution based on the inferred geologic properties; and 4) develop a working prototype model for site tests in areas where the geology is already known and remotely sensed data are available to evaluate the geologic and geophysical data output.

This project has just begun and this poster discusses the project status and future plans.