Three-dimensional modeling of ground-airborne transient electromagnetic responses of typical models based on the finite difference approach

Ground-airborne transient electromagnetic (GATEM) methods combine the advantages of ground TEM and airborne TEM, which can simultaneously fulfill the requirements of detection depth and work efficiency. In this study, the GATEM responses of typical models were numerically modeled using the finite-difference time-domain (FDTD) approach. The traditional FD approach of TEM uses the upward-continuation boundary condition at the surface of the earth but does not model complex terrains with an air layer. In this study, the FDTD approach was used to directly mesh and calculate the air layer. Specifically, based on the quasi-static conditions of the electromagnetic field and the frequency spectral features of the transmitting current waveform, the choice method for obtaining the fictitious permittivity in the air layer was studied, and the stability problem in the calculation of the air layer was addressed. The convolutional perfectly matched layer boundary conditions were used at the air and underground boundaries to improve the accuracy of the solution in the late field. Moreover, the results of this approach were compared with the analytical solution and the results of other algorithms, and the accuracy and reliability of the approach were verified. Next, this approach was used to perform three-dimensional forward simulations on the GATEM responses of typical models such as the 3-D body model, rugged topography model, and low-resistivity overburden model. The modeling results show that the GATEM response is significantly affected by flight altitude in the early stage, whereas the late response is less affected by the flight altitude. The sensitivity of the GATEM response was smaller than that of the ground TEM. Signal reversal is observed in the ∂Bz/∂t response curve under the influence of a hill. However, the influence of the hill can be treated as a background field in the late period. The ∂Bz/∂t curve observed above the goaf is higher than that without the goaf, but the difference between them will be reduced due to the influence of the low-resistivity overburden and measurement noise.