Sensitivity functions of airborne EM systems over layered models

Tuesday, 11 June 2019: 13:10
Davie West Building, DW103 (Florida Atlantic University)
Anders Vest Christiansen, Aarhus University, Aarhus, Denmark, Line Meldgaard Madsen, Aarhus University, Geoscience, Aarhus C, Denmark and Esben Auken, University of Aarhus, Aarhus, Denmark
Abstract:
Time-domain (TEM) and fre­quency-domain (FEM) methods with many different measurement systems and different transmitter-receiver configurations can be used for investigation of the near subsurface. By studying sensitivity func­tions of the different systems, it is possible to compare the configurations and asses their advantages and shortcomings.

During the last decades, advancement within airborne TEM systems have made it possible to acquire early time data, which makes it possible to get at good resolution of the near-surface geophysical properties. The research question here is to investigate how good the near surface resolution is and compare it across systems and with layered models.

We present a simple method to compute sensitivity functions for different measurements systems in both time- and frequency-domain over layered models and with inclusion of all system characteristics such as Tx-Rx geometry, waveforms and filters. For a small volume in the Earth, the sensitivity is given by the partial derivative of the measured electric fields (E-fields) with respect to the conductivity of the volume. The sensitivity can be positive or neg­ative, meaning that the measured response will either increase or decrease with an increase of the con­ductivity and also meaning that the sensitivity contribution from this particular volume adds or subtracts to the total sensitivity.

To make quantitative representations of the sensitivity functions in 1D and 2D, the 3D sensitivity model is integrated along the relevant axes. These can further be interpreted in terms of footprint in both lateral and vertical directions.

In the figure below the 2D sensitivity and the 1D sensitivity with depth are shown for a homogeneous halfspace. The sensitivities are normalized by the maximum sensitivity at the given time. The 1D sen­sitivities are integrated from the absolute value of the 3D sensitivities to avoid that positive and negative values cancel each other out.

At least three different EM systems will be compared in the presentation, namely: 1) An airborne TEM system (SkyTEM), 2) an airborne FEM system (Resolve), and 3) a ground-based TEM system (tTEM), to have a ground for comparison against best-case resolution from the surface. Sensitivity functions and footprints for layered models will be presented as well.

Figure: 2D sensitivity distributions for the SkyTEM (a) and tTEM (b). The images shown are for 5 µs over a 30 ohmm halfspace. The images are normalized individually. c) displays the 1D sensitivity at 5 µs (black) and at 100 µs (grey) for tTEM (full line) and SkyTEM (dashed line)