Badlands Bombing Range Research Articles

Comparison of Performance of Airborne Magnetic and Transient Electromagnetic Systems for Ordnance Detection and Mapping

Magnetic and electromagnetic data collected by helicopter boom-mounted systems at three different sites permit direct comparison of the systems as to their suitability for buried ordnance detection, mapping, and discrimination. Airborne boom-mounted magnetic systems are at a more advanced stage of development than their electromagnetic counterpart. However, in basaltic terrain, transient electromagnetic systems have proved capable of detecting buried ordnance, whereas magnetic systems may fail to detect ordnance altogether. Magnetic systems use passive sensors and these can be distributed along the boom structure such that dense data can be collected with sensors spaced 1–2 m apart over a broad swath, up to 12-m wide. The ORAGS-TEM electromagnetic system, having only two receivers, must rely on interleaved flight lines to obtain data of a spatial density approaching that of the airborne magnetic systems. The total magnetic fields from unexploded ordnance decays at 1/R3 for total field and its gradient at 1/R4. This permits adequate signal-to-noise levels to be easily attained for larger ordnance types at survey heights up to seven meters. Active electromagnetic fields decay at between1/R4 and 1/R6, depending on ordnance type and sensor geometry, and this constrains current electromagnetic systems to practical survey altitudes of less than three meters. Tests at the Badlands Bombing Range indicate that, in some circumstances, the signal-to-noise for the airborne electromagnetic system exceeds that of airborne magnetic systems, and even ground electromagnetic systems. Because time must be allowed for transmitter current buildup and decay, ORAGS-TEM is not capable of sampling along line at the same spatial density as can magnetic systems. However, the temporal signal decay permits greater opportunity for ordnance discrimination than magnetic measurements.

UXO Time-Constant Estimation from Helicopter-Borne TEM Data

We have developed a new helicopter-borne transient electromagnetic system, known as ORAGS-TEM, which was designed for the detection of unexploded ordnance (UXO) through very low altitude measurements. This system has already achieved considerable success in demonstrations over prepared test grids and a bombing site at the former Badlands Bombing Range (BBR) in South Dakota. UXO ranging in size from [Formula: see text] [Formula: see text] bombs to [Formula: see text] illumination shells and [Formula: see text] [Formula: see text] rocket components were detected by both magnetometer and transient electromagnetic technologies during these trials, conducted in September, 2002. The signal∕noise ratio (SNR) observed in TEM measurements during these trials was high, prompting us to ask, “What degree of UXO discrimination can be achieved through detailed analysis of this airborne TEM dataset?” Given the degree of spatial averaging and sampling limitations imposed by the system’s flight height and speed, we felt that a very detailed analysis of the type performed by Pasion and Oldenburg would not be practical. Instead, we developed an improved transient analysis technique based on the Matrix Pencil Method to improve the accuracy of exponential decomposition of the observed transients. Where target SNR values of 10 or higher were present this method yielded repeatable results that reliably distinguished compact, long-time-constant targets such as bombs and artillery shells from short-time-constant targets such as thin-walled scrap from practice bombs. As system sensitivity and resolution continues to improve, we expect that similar target discrimination methods will become standard data analysis tools.

Determination of the parameters of compact ferro-metallic objects with transforms of magnitude magnetic anomalies

The transforms based on magnitude of the magnetic anomaly (MMA), or absolute value of anomalous magnetic intensity, are relatively insensitive to the direction of the source magnetisation vector and show high centricity. Obtaining them requires calculation of only first-order horizontal derivatives of the component magnetic anomalies X, Y and Z, obtained through transformation of total magnetic anomalies Δ T (TMA). We studied the use of the magnetic transforms based on the MMA of a dipole source and that of their ratios for determining the geometric and physical parameters of compact environmental ferro-metallic sources. Analytical formulas for all the parameters of the magnetic dipole based on calculation of MMAs, their gradients and Laplacian, and their ratios were obtained. The horizontal position of the dipole source is determined by the maximum of the Laplacian of the MMA, which is the transform with the highest centricity. We applied the approach on model data and on magnetic anomalies over environmental sources, represented by unexploded ordnance (UXO) and clutter from the Badlands Bombing Range, USA. The obtained parameter values were compared with those estimated through Euler deconvolution using unprescribed structural index and linear model of the background, and through inversion. The comparison showed that the proposed method produced similar depth results to that of the Euler deconvolution method for the objects with estimated structural index over and around 2.5, i.e., for the sources, which to a great extent can be approximated with a dipole. The inversion for the dipole parameters using as initial parameters the estimated with the MMA transform method ones confirmed the results of the MMA transform method for the dipole coordinates and magnetisation for all items.

Using physics-based modeler outputs to train probabilistic neural networks for unexploded ordnance (UXO) classification in magnetometry surveys

The outputs from a physics-based modeler of magnetometry data have been successfully used with a probabilistic neural network (PNN) to discriminate unexploded ordnance (UXO) from ordnance-related scrap. Cross-validation predictions were performed on three data sets to determine which modeler parameters were most valuable for UXO classification. The best performing parameter combination consisted of the modeler outputs depth, size, and inclination. The cross-validation results also indicated that good prediction performance could be expected. Model outputs from one location at a site were used to train a PNN model, which could correctly discriminate UXO from scrap at a different location of the same site. In addition, data from one site, the former Buckley Field, Arapahoe County, CO, was used to predict targets detected at an entirely different training range. The Badlands Bombing Range, Bull's Eye 2 (BBR 2), Cuny Table, SD. Through careful selection of the probability threshold cutoff, the UXO detection rate obtained was 95% with a false alarm rate of only 37%. The ability to distinguish individual UXO types has been demonstrated with correct classifications between 71% and 95%.