Effects Of Spatial Data Density, Sensor Noise And Position Errors On Uxo And Clutter Target Parameters From Inversions Of Alltem Data

ALLTEM is a multi-axis electromagnetic induction system designed for unexploded ordnance (UXO) applications. It uses a continuous triangle-wave excitation and provides good late-time signal-to-noise ratio (SNR) especially for ferrous targets. Inversions of field data acquired in survey (moving platform) mode over the Army’s UXO Calibration Grid and Blind Test Grid at the Yuma Proving Ground (YPG), Arizona in 2006 produced polarizability moment values for many buried UXO items that were reasonable and generally repeatable for targets of the same type buried at different orientations and depths. In 2007 we finished construction of a test stand that allows us to collect data with varying spatial data density and accurate automated position control. We have studied the behavior of physics-based nonlinear inversions of ALLTEM test stand data as a function of spatial data density, sensor SNR, and position error. These studies have been performed as part of our effort to develop quantitative confidence levels for our inversions. A high confidence level in inversion-derived target parameters will be required when a target is declared to be harmless scrap metal that may safely be left in the ground. Unless high confidence can be demonstrated, regulators will likely require that targets be dug regardless of any “no-dig” classifications produced from inversions, in which case remediation costs would not be decreased. Multi-axis transmitter (Tx) and receiver (Rx) systems such as ALLTEM provide a richer data set from which to invert for the target parameters required to distinguish between clutter and UXO. Our inversions are more tolerant of sensor noise and position error than has been reported for single-axis, single-element data inversions and thus higher confidence in calculated target parameters can be achieved. This is especially important in the case of moving platform survey mode field data.