Comparison of flight height controlled UAV and ground vector magnetometer survey data covering an urban landfill

Abstract

An autonomous unmanned aerial vehicle (UAV)-mounted 3-axis fluxgate vector magnetometer (FVM) was flown over a Golden, Colorado landfill to map buried ferromagnetic objects and infrastructure. The FVM is picoTesla-scale resolution. The UAV used active global navigation satellite system (GNSS) positioning when flying above a one-meter resolution LiDAR topographic surface, to maintain a slung sensor at ground-clearance heights that averaged 2-5 meters (m) during surveying. Prior to flying the magnetometer, the UAV height settings above the LiDAR surface were calibrated using test flights. The setting that provided a reasonably consistent three meters ground clearance for the slung sensor was then used for the 15 km aeromagnetic survey. All flying was done within visual line of sight (VLOS). Prior to the airborne surveys, the FVM was walked over the landfill on similar 10 m spaced ground survey lines, with the sensor a constant two meters height above ground. The FVM ground and airborne survey datasets are compared after both are calibrated for errors and corrected for sensor movements, and converted to diurnal-corrected total magnetic intensity (TMI) scalar data. As expected, the aeromagnetic data though high resolution, is slightly longer wavelength, lower frequency and lower resolution than the ground data, primarily due to inverse distance attenuation of the magnetic field at the 2-5 m averaged sensor heights flown above the magnetic sources. Another factor may be sample density; the airborne survey was flown at 4 m/s, whereas the ground survey was walked at an average 1 m/s. The ground survey returns an average 4-5 times higher sample density at a constant two-meter height from the 200 Hz sampling FVM. The first gridded data comparison is between the aeromagnetic survey and ground survey upward continuation filtered (UCF) 1 m, so that the FVM survey levels are on the same airborne level average height of 2-5 m. The UCF ground survey retains higher resolution. To increase the resolution of the aeromagnetic data, it was subsequently downward continuation filtered (DCF) both 1 and 2 m to ground survey levels. The subsequent magnetic images show that the DCF aeromagnetic data resolution is now near-parity with the ground survey data. Similar magnetic signatures are identified on both datasets, and their morphologies, amplitudes, and locations are comparable, indicating low level UAV terrain drape surveys could replace ground surveys in some circumstances. The results of the vector data calibration and compensation are also presented in graphical and numeric format. Lastly, an alkali vapor magnetometer (AVM) TMI dataset, walked the prior year at two meters height over the same FVM ground survey lines, is compared with the FVM gridded images as the scalar TMI surveying reference standard.