Objective  To address the challenge of detecting aquifers outside the roadway excavation contour, this study combined the close-range detection via roadway-borehole transient electromagnetic (TEM) method with the wavefield transform. This will help simultaneously receive detection signals in front of the mining face and outside the roadway excavation contour while also highlighting the boundary information of aquifers. Moreover, to obtain a pseudo-wavefield featuring both reasonable kinematic characteristics and high sensitivity to changes in the electrical structures of strata, this study proposed a self-adaptive algorithm of inverse wavefield transform, thereby improving the accuracy and efficiency of the inverse wavefield transform of roadway-borehole TEM data.

Methods The TEM wavefield transform faces some inherent limitations, such as a large dynamic range of the kernel function and high ill-posedness of the inverse transform equation. To address these challenges, this study, using the analytical solutions of the kernel function of wavefield transform over an infinite interval, proposed a method to automatically determine the interval of integration for inverse wavefield transform based on parameters such as the acquisition time of TEM data. The proposed method can minimize the dynamic range of the kernel function. Furthermore, using the precise integral method, this study developed an algorithm for solving inverse wavefield transform equations based on a self-adaptive step size of integration and iteration process, thus converting the solving of the ill-conditioned system of equations into solving an integral stably. For aquifers outside the roadway excavation contour, their TEM response signals were determined through forward modeling using the finite volume method. Accordingly, the pseudo-wavefield records of their vertical magnetic field components of TEM data were obtained. Moreover, the inverse wavefield transform of measured TEM data from boreholes was also conducted using the same algorithm.

Results  The results indicate that the obtained pseudo-wavefield records exhibited reasonable kinematic characteristics. The reflected wave trajectories in the pseudo-wavefield records were reasonably similar to the time-distance curves in the single-shot seismograms. Regarding numerical accuracy, the back substitution results of the pseudo-wavefield data were highly consistent with the forward modeling-derived or measured data, with the maximum relative errors all below 10%.

Conclusions  Under the roadway-borehole observation mode of TEM data, the proposed self-adaptive algorithm of inverse wavefield transform enables the acquisition of pseudo-wavefield records with clear physical significance and reliable numerical accuracy. Furthermore, the obtained pseudo-wavefield is highly sensitive to the aquifer boundaries with significant electrical contrasts, providing robust data support for further pseudo-wavefield imaging.