Analysis Of Noise Coherence In Airborne Magnetic Gradients For Uxo Detection

Oak Ridge National Laboratory (ORNL) and the US Army Engineering and Support<br>Center, Huntsville (USAESCH) have been developing advanced helicopter platforms for<br>geophysical arrays since 1997. A significant portion of the funding since 1999 has been through<br>the Environmental Security Technology Certification Program (ESTCP). The most recent<br>refinements to the Oak Ridge Airborne Geophysical System (ORAGS) magnetometer array<br>focused on noise reduction techniques including specialized processing methods. In particular,<br>the use and optimal positioning of multiple sensors to reduce coherent noise through gradient<br>measurements has proven very effective.<br>In boom-mounted magnetometer systems, the primary noise source is the helicopter<br>itself. Noise from this source falls into two main categories: maneuver noise and rotor noise.<br>The former is relatively low frequency, while the latter is relatively high frequency. Frequency<br>filters of various types have been used with success, but fall short where the noise is within the<br>bandwidth of the signal. Such is often the case when searching for unexploded ordnance (UXO)<br>and other shallow targets at rapid flight speeds. Under these conditions, helicopter noise can be<br>reduced by making use of the fact that much of it is coherent between sensors.<br>In this paper, we follow up previous theoretical work (Gamey et al, SAGEEP 2002) with<br>field examples from recent ORAGS-VGrad surveys. We examine the coherence of various types<br>of helicopter noise and demonstrate that measured vertical gradients can offer raw signal-tonoise<br>improvements of 4:1 or more over comparable total field measurements.