Abstract
Transient electromagnetic method (TEM) is an effective tool to map the groundwater distribution in tunnel construction. To highlight the structural information of the water-bearing zones, inverse wavefield transform is used to extract pseudo-wavefield from TEM data. To provide accurate and reliable pseudo-wavefield, pseudo-wavefield extraction should tackle the large dynamic range of the kernel function, ill-posed inverse transform function, and self-adaptive solving. In this paper, a series of works had been done to solve the above problems. Based on the property of the wavefield transform and its kernel function, we provided an empirical formula to shorten the integral interval of the inverse wavefield transform. A self-adaptive method combined with the precise integral method (PIM) was conducted to provide accurate equation solving. The data misfit between the modeling data and the fitted data is less than 2.5% in our numerical models. The traces of the pseudo-wavefield highlight the boundary of the low-resistivity layers of different depths and thicknesses. For the field data measured in a constructing tunnel, the reflection in our pseudo-wavefield traces provided massive subsurface electrical interfaces with a data misfit of less than 5%. Combined with the low-resistivity zones in the apparent resistivity results, three water-bearing zones were highlighted and then verified by a drilling test, which confirms the reliability and practicality of our method.