Development and application of a multifrequency/multicoil semi airborne UAV-FDEM-system with optimized inductive source

In course of the Project ‘FlowCast’ (ESS Programme / Austrian Academy of Sciences, 2019-2023) a semi-airborne UAV frequency domain EM systems has been developed with the aim of achieving high penetration depths with a UAV sensor system. This required an innovative, powerful inductive source that, due to weight considerations, is designed to be stationary on the ground while the receiver is towed by a drone.  Temporal synchronization of transmitter and receiver takes place via GPS time signal, relative geometry using GPS-RTK recording, correction of the orientation of the receiver using IMU module. In the shallow domain the semi-airborne UAV-EM configuration is compared to a full airborne UAV-EMI system. Both systems can be configured very flexibly in multi-frequency and multi-coil operation and can be optimized for various tasks (conductivity range in the subsurface, penetration depth, resolution). The study presents the hardware development (UAV sensor platform, transmitter, receiver) and data processing (preprocessing and inversion) as well as the results of case studies on test areas around Vienna with buried test bodies or natural structures and ongoing work.

An AIR8 medium lifter (Air6 Systems) octocopter with 25 kg total takeoff mass and 10 kg payload serves as the sensor platform. The receiver is an autonomous sensor system that measures and records amplitude and phase of the magnetic component of a low frequency electromagnetic (EM) signal that is emitted by the stationary transmitter. Auxiliary sensors acquire an accurate time tag, the sensor's GPS position, and the orientation of its main axis. The sensor has a mass of 5 kg, provides sufficient robustness for practical field operations, and can be either hand-positioned along a defined profile, or suspended from an airborne platform that positions the device at a defined position and orientation in 3D space. The combination of position-, amplitude- and phase information of both the receiver and the transmitter allow the estimation of sub-surface electrical conductivity during post-processing of recorded data.

The frequencies to be generated by the transmitter are in the audio range. Wire loop diameter ~ 30m, N=3. A Class D amplifier was developed that voltages up a chopped transmitter signal with a switching frequency of 195 kHz via a transformer and then reassembles it. An FPGA (Field Programmable Logic Array) is used to generate the pulse pattern and the sine signals. This creates a purely digital system. This solution enables large currents in the transmitter loop, thus large dipole moments over a wide frequency range. The 3D inversion program developed at KIGAM is based on the vector finite element method for the 3D electromagnetic forward modeling. The secondary field formulation is used to obtain the secondary field from primary field by ground loop source. 3D resistivity distribution is reconstructed from field EM responses using iterative least squares inversion method adopting active constraint balancing algorithm.