New Insights into the Anomaly Genesis of the Frequency Selection Method: Supported by Numerical Modeling and Case Studies

The frequency selection method (FSM) is the further development of the audio frequency telluric electricity method (TEFM). However, debate continues regarding the mechanisms leading to anomaly genesis. Therefore, the present study intends to explore this using 2D forward modeling of magnetotelluric (MT) sounding and practical applications of FSM on three Chinese case studies in karst and granitic settings. In the first stage, the profile curves and pseudo-section of apparent resistivity (ρs\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\rho }_{s}$$\end{document}) and horizontal electric field component (Ey)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${(E}_{y})$$\end{document} in MT mode are obtained by forward calculation. As a result, the static shift in ρs\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\rho }_{s}$$\end{document} is observed over the near-surface inhomogeneities, as documented in the literature. Additionally, the profile curves of Ey\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${E}_{y}$$\end{document} showed an obvious static shift in the rectangular coordinate system (i.e., the curve rises with the increase in frequency) which is a well-known phenomenon. The pseudo-sections of Ey\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${E}_{y}$$\end{document} also showed static shift characteristics at the horizontal position above the anomaly, referred to as the “noodles phenomenon.” The FSM results were obtained from case studies related to groundwater and low-resistivity clay-filled karst body identification. The ΔV\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta V$$\end{document} section curves and pseudo-section showed a significant low potential, and a “noodles phenomenon,” respectively, above the low-resistivity anomalous body. These abnormal characteristics of ΔV\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta V$$\end{document} are the basis for delineating the horizontal position of the groundwater aquifer by applying FSM. It is concluded that the anomaly of FSM is the reflection of the static shift in MT, and hence, the FSM can be categorized as a “static shift method.” Therefore, this inspired us to investigate whether the static shift feature of the surface Ey\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${E}_{y}$$\end{document} component could be utilized to explore near-surface geological bodies such as clay-filled or water-filled cavities.


Introduction
The frequency selection method (FSM) is a general name given to numerous similar methods proposed by Chinese scholars in the early 1980s successively, such as the electric pulse method of a natural field (Yang, 1982), stray current method (or audio earth potential method) (Yang, 1982), the frequency selection method of telluric current (Han et al., 1985), sound frequency geoelectric field method (Xin, 1982), natural low-frequency electric field method (Lin et al., 1983), telluric electrical field frequency selection method (Li, 1991), audio frequency telluric electricity method (Lian et al., 1991), interference electric field method (Bao et al., 1994), natural alternating field method (Luo, 1994), and underground magnetic fluid detection method (Zhou et al., 2009). These methods work on the same equipment detection principle and measuring methods and are therefore generally called FSM. FSM is naturally the further development and application of audio frequency telluric electricity method (TEFM). The field operations of the FSM and audio magnetotelluric (AMT) method are the same; however, FSM measures the horizontal components of the electric field at different frequencies of electromagnetic signals on the earth's surface.
At present, the advancement in geophysical instrumentation has expanded the operating frequencies to the range of 10-5000 Hz. The potential electrode spacing is usually set at 10 or 20 m as an implementing section method. Therefore, the instruments have the advantages of being portable, having a simple operation, and high efficiency. Since the 1980s, the development of FSM, including instrumentation and applications, has been flourishing. In 1976, B. H. Liang proposed the electric pulse of the natural field method and applied it to explore groundwater (Yang, 1982). Yang (1982) proposed the stray current method and carried out theoretical research on the basis of experiments. Xin (1982) proposed the audio geoelectric field method, and the relative SDD-1 acoustic geodetic field instruments were further popularized. Lin et al. (1983) proposed a natural low-frequency electric field method. They modeled the distribution law of the anomaly of frequency selection on the vertical dike model according to the propagation law of the magnetotelluric electromagnetic field. Han and Wu (1985) proposed a frequency selection method for the telluric electricity field and developed the DX-1 frequency selection instrument for the telluric electricity field by referencing the magnetotelluric instrument. Their instrument has been well promoted and applied in China in the past. Recently, Han and Han (2020) compiled their practical application examples over years into a monograph. Bao (1994) proposed the interfering electric field method and developed the corresponding instrument, implementing indoor simulation experiments and a considerable number of field measurements. Zhou et al. (2009) proposed an underground magneto fluid detection method and developed an underground magneto fluid detector and portable underground water source detector. Some other geologists mentioned some other name concepts in their published articles (Liang et al., 2016;Luo, 1994). The FSM has already acquired successful applications in the exploration of shallow groundwater resources and hazards associated with mine water (Farzamian et al., 2019;Singh et al., 2022;Song et al., 2021;Yang et al., 2020aYang et al., , 2020b. However, the nature of these concepts still relates to FSM according to the equipment they used.
Since the introduction of FSM, numerous studies have been carried out with the main focus on the development and application of instruments (Cheng and Bai, 2014;Yang et al., 2020a), especially in groundwater resources and disaster studies, using various types of instruments. Nevertheless, there is limited theoretical research on this method. One possible reason is the complexity of natural sources and factors of human structures measuring shallow natural electromagnetic source fields. Additionally, manual reading of an instrument was complicated before the advent of an intelligent frequency selector. The electric field components measured by the pointer-deflecting instrument include only a few limited frequencies, which inhibits an understanding of the profile curve. Yang et al. (2017Yang et al. ( , 2020a focused more on the theoretical research of FSM and completed some relevant studies driven by long-term practical applications. Though some simple relative geological models can simulate abnormal curves similar to the real ones, many aspects of FSM studies, such as the size and ranges of the anomaly, seem to achieve unsatisfactory results. Magnetotellurics (MT) is a passive-source electromagnetic method with a field source, proposed by A. N. Tikllonov and L. Cagnird in the early 1950s. Because of its great exploration depth, MT has been widely applied to oil and gas fields, coal mines, metal mines, karst water structures, crustal lithosphere structure, and other aspects (Yang et al., 2019). The MT forward method can be realized by the integral equation method, finite difference, or finite element method, and the numerical solution is obtained by the approximation of a differential or integral equation and the solution of linear equations. The finite element method is more advantageous in MT forward calculation than the other two methods. Coggon (1971) firstly realized the importance of electromagnetic forward simulation of finite element methods. Subsequently, some experts applied the MT forward method by using rectangular elements, triangular elements, mixed elements, and unstructured triangulation elements, respectively. Chen (1981) and Hu et al. (1982) implemented MT forward modeling research. Xu (1994) further studied finite element and mesh portioning. Chen et al. (2000) proposed the finite element direct iteration algorithm. Tan et al. (2003) introduced the biconjugate stable gradient method based on previous studies. Zhou et al. (2021) studied forward and inversion 2.5-D electromagnetic methods in the frequency domain. Vatankhah et al. (2022) developed a joint 3D forward and inversion method of gravity and magnetic data and applied to an untapped economic minerals area. After years of in-depth research, 2D and 3D finite element forward electromagnetic methods have attained some 970 T. Yang et al. Pure Appl. Geophys. stability and have become one of the significant means of studying electromagnetic issues. Static shift is unavoidable in magnetotelluric observation, which is often confused with interference and so suppressed or eliminated, especially in deep-target studies (Di et al., 2019;Hu et al., 2017;Xiong et al., 2021). Just as in the initial period of seismic exploration, Rayleigh waves were always treated as interference signals (Yang & He, 2013). Meanwhile, many studies have been carried out on the formation mechanism of static shift and its relationship with the presence of shallow anomalous bodies. Liu et al. (2018) also presented one viewpoint of using static excursion distortion law to detect shallow anomalous bodies. According to the propagation feature of electromagnetic fields in a horizontal uniformly layered medium, the direction of the current field is parallel to the interface having no accumulated charge. However, charges can be accumulated on the interface where the presence of the surface or subsurface single inhomogeneous body can cause distortions in the observed electric field on the surface (Fig. 1). Similarly, the MT sounding curve will also have a static shift, attributed to the low-frequency characteristics of the magnetotelluric field. In the case where the scale of electrical inhomogeneity is much larger than the wavelength of electromagnetic waves, the observed changes in the apparent resistivity curve and phase curve are not considered as static shifts (Huang et al., 2006). This paper discusses the main causes of abnormal results obtained from the frequency selection method (FSM) of detecting a telluric current based on the causes and features of the static shift of the magnetotelluric (MT) sounding. The objectives are achieved by simulating the static shift of MT using the finite element method based on Maxwell's equation. Unlike previous studies only considering the static shift feature of apparent resistivity and phase curves, we further focused on analyzing the characteristics of surface electric field components. Additionally, three case studies on karst and granitic geological settings have been considered for the validation of numerical findings. The outcomes of the present study will improve future applications of FSM through a better understanding of its physical nature.

Cause of Static Shift
An interface will accumulate charge when current flows through the interface of an inhomogeneous body. According to Gauss's law, the continuity equation of current, and approximation under quasistatic conditions, we can deduce the surface charge density q s on the conductive medium surface as follows: where r 1 and r 2 are the conductivity of the inhomogeneous body and the surrounding rock, respectively, E n1 and E n2 are the strength of a normal electric field at the interface of the body and rock, and e 0 is the dielectric constant. Though q s is small for the field, its effect on the electric field is negligible, which is the physical cause of the static displacement. The influence of accumulated charge on the surface can be detected by an instrument when the skin depth is much larger than the size of the inhomogeneous body, e.g. the situation shown in Fig. 1. MT is always affected by the small inhomogeneous body at the surface or subsurface as exploring deep geological bodies. According to previous research, the secondary electric field generated by the accumulated charge has a positive relationship with the same phase of the primary field, independent of frequency simultaneously. Only the transverse magnetic (TM) field mode is affected under strict two-dimensional geological conditions, while both TM and transverse electric field (TE) modes are affected in three-dimensional conditions. In addition, static displacement is also related to underground resistivity, the position of the electric field electrode, the length of the electric measuring dipole, and the size ratio of the inhomogeneous body.

Boundary and Variational Problem
Any electromagnetic problem can satisfy Maxwell equations. Forward simulation studies a steady-state field problem under the conditions satisfied by the natural electromagnetic method. Given that the angular frequency is x and the time dependence is e Àixt , MT responses can be described by Maxwell's equations: where E and H are electric and magnetic fields, respectively, l, r and e are magnetic permeability, electrical conductivity and permittivity, respectively, i is the imaginary unit, and i 2 ¼ À1.
Given that the underground electrical structure is two-dimensional and the strike is along the x axis, the y axis is perpendicular to the x axis and horizontally to the right, and the z axis is vertically downward (Fig. 2). As the electromagnetic field propagates vertically downward into the medium in the form of plane waves, Eqs. (2) and (3) are expanded according to components, and two independent polarization modes (TE and TM) can be obtained. Since FSM generally only measures the horizontal electric field component along the survey line, this study only discusses the TM polarization mode. The equations in the TM mode are as follows: where H x in Eq. (4) satisfies the following partial differential equation.
o oy , and k ¼ ixl, the variational problem corresponding to 2D MT forward modeling solved by the finite element method is as follows: where X represents whole computational domain, AB and CD are, respectively, upper and lower bounds in Fig. 2, k is wave number, and k ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi if the displacement current neglected.

2D Finite Element Method
In order to improve the accuracy of the simulation, the computational domain is discretized as rectangular elements. Biquadratic interpolation is adopted in this rectangular element, namely, each element takes a total of eight points at the four vertices and four midpoints of the four edges. Through the unit analysis of unit e, the unit integrals of the three integrals in the first equation in Eq. (6) can be separated into In the above equations, T represents the transposition of the matrix. The specific formulas of coefficient matrices K 1e , K 2e , and K 3e can be found in Xu (1994) and Liu et al. (2009). Extend K 1e , K 2e , and K 3e into a matrix composed of all nodes, that is, sum all elements and obtain By taking Eq. (10) and making it equal to zero, a system of linear equations can be obtained as follows: where K is an overall stiffness matrix. Then the solution of the linear equations, namely u of each node, can be obtained by substituting the upper boundary value of Eq. (6). After calculating the u value of each node (i.e., H x of TM polarization mode), we can solve the partial derivative of ou oz along the vertical direction of the surface by using the difference method: where H x1 represents the field value of the surface, H x2 , H x3 , and H x4 are the field values at the first three isometric grid nodes below the surface, respectively, l is the vertical distance from the first node to the fourth nodes. At last, the E y value can be calculated by the second formula in Eq. (4) along the horizontal direction. Meanwhile, the impedance Z TM , the apparent resistivity q TM a , and the impedance phase u TM of TM polarization mode can be calculated by the following formulas:

Solution of Linear Equations and Mesh Generation
Two-dimensional finite element forward calculation of MT ultimately comes down to solving symmetric large-scale spare and ill-posed linear equations with complex coefficients. The K in Eq. (11) is positive definite. These equations can be solved by many methods such as singular values decomposition, the Newton method, the conjugate gradient (CG) method, etc. This study adopted the BICGSTAB algorithm without completely LU decomposition, which has the advantages of fast, high precision and good stability.
Only the non-zero elements of the sparse matrix are stored in the solving process. In addition, the square grid can decrease the error of calculation near the boundary through the trail calculation of the uniform half-space model. Therefore, we use the square grid as much as possible to expand the modeling calculation area on the premise of ensuring the calculation accuracy and fully utilizing the storage space of the computer.

Comparison with Analytic Solutions
First, a layered medium model is used to verify the correctness of the above 2D MT modeling code. Parameters q i and h i (i = 1, 2, 3) represent the resistivity and thickness of each layer of medium, respectively, in Fig. 3a. The entire computational domain, namely AB 9 AD, is an area of 6 9 3 km 2 and is discretized by the grid size (d y and d z ) of simulation fixed as 50 m. Sixteen frequencies were used in the logarithmic scale from 1 to 1000 Hz for Vol. 180, (2023) New Insights into the Anomaly Genesis of the Frequency Selection Method 973 calculation or simulation. The analytic solutions of the layered medium model can be obtained from the literature (Xie et al., 2021). The solid line and dots represent the analytic solutions and simulation results of the layered medium model, respectively. Both results are consistent, which indicates the correctness of the 2D MT modeling code.

Simulation Analysis
Applying the aforementioned 2D finite element simulation theory of MT, we can carry out the forward calculation using 2D geological and geophysical models. A forward model of a two-layered horizontal stratified medium with one clay-filled karst anomaly in the substrate layer was chosen (Fig. 4). The entire computational domain constitutes an area of 6 9 3 km 2 and is discretized into 120 9 60 = 7200, namely AB is 6 km, and AD is 3 km. The grid size (d y and d z ) of simulation is fixed as 50 m, and 53 frequencies are used (f = 10.^(-4:0.125:2.5) Hz). Figure 5 shows the results of the model 1 in terms of q s and E y , and their corresponding pseudosections. For convenience, only the results of three frequencies (1, 10, and 100 Hz) are drawn in Fig. 5a, which reflects a downward shift in the q s curve with the increase in frequency, a typical characteristic of static shift. The E y profile curves in Fig. 5b display good differentiation in the Cartesian coordinate system, and their curve forms at the three frequencies remained unchanged, rising significantly with the increase of frequency. Meanwhile, the relatively lowpotential anomaly directly above the abnormal body also increased significantly. Figure 5c shows the obvious ''noodles phenomenon'' (or ''hanging noodles phenomenon'') caused by the static shift. The q s isoline stretches downward at the position corresponding to the abnormal body. The lower the frequency, the more pronounced the stretching. This phenomenon has been documented in the literature, and is expected to be eliminated or suppressed in the measured data of MT, controlled source audio magnetotelluric (CSAMT), and wide-field electromagnetic method (WFEM) (Lei et al., 2017;Li & He, 2021;Tournerie et al., 2007). The pseudo-section of the E y component in Fig. 5d also shows a similar phenomenon, but the stretching direction of the isoline is upward and becomes more obvious with the increase in frequency. At the same time, the magnitude of the anomaly is not as apparent as the q s isoline. In the FSM application, the frequencies are primarily selected in the audio range. Compared with MT, the overall working frequencies of FSM are higher; for example, the frequencies of the pointer DX-1 type of frequency selector are set to only five, namely 14. 6, 71.8, 161, 262, 327, and 783 Hz, and the intelligent PQWT-TC300 operates at 40 frequencies in the range of 12-5000 Hz. For another, the main exploration targets of the selector are the relatively shallow media; for instance, the apparent depth of PQWT-TC150 and TC300 inversion is 150 m and 300 m, respectively. In addition, the size of the target body detected by FSM is generally relatively small. The target of most underground explorations in a karst environment is a fissure, fault, or other geological structures, whose size is generally not as large as that of model 1 (Fig. 4). Therefore, another shallow buried water channel model is adopted and discussed in the following.
The next model is a water channel model constituting two horizontal layers, upper loose sediment and a substrate granite layer. The water channel is supposed to be in the upper layer, semi-filled with air (60,000 X m) and water (60 X m). The entire computational domain ABCD is 50 9 30 m 2 and discretized into 100 9 60 = 6000, having a grid size of 0.5 m (Fig. 6). The calculated frequencies are 40 frequencies in the range of 12-5000 Hz, the same as can be detected by the TC300 frequency selector instrument.
The results of model 2 are shown in Fig. 7 and are similar to those presented in Fig. 5. The FSM often uses three frequencies (25, 67, and 170 Hz) to scan underground sections. The respective q s profile curves of these three frequencies (Fig. 7a) are quite similar to the curves of model 1, as they shift downward with the increase in frequency, but the waveform of the curves remains unaffected. The static shift in Fig. 7b-d is still very obvious, and the ''noodles phenomenon'' is more prominent. The profile curve shows a local maximum at y = 0, which may be the result of the high resistance (i.e., air) filling the upper part of the channel. The ordinate in Fig. 7c and d is the pseudo-depth of approximate inversion based on the skin depth formula of electromagnetic waves. The pseudo-depth is the same as that of the real-time result inversion method adopted by the field instrument.
It is can be concluded from the two considered theoretical models that the static shift in the horizontal electric field component E y profile curve generated by the natural electromagnetic field at the surface is an obvious phenomenon in the Cartesian coordinate system, but the relationship between the lifting direction of the profile curve and the frequency is opposite to that of the q s profile curve. Meanwhile, the static shift is still obvious in the E y pseudo profile, but the stretching direction of the E y isoline is opposite to that of q s . The static shifts of q s and E y curves are all caused by the presence of near-surface inhomogeneity. Therefore, this inspired us to investigate whether the static shift feature of the surface E y component could be utilized to explore near-surface geological bodies such as clay-filled or water-filled cavities.

Case Studies
Three perspective sites in karst and granitic environments are chosen using different instruments capable of measuring different frequency ranges for comparison with the numerical findings.
In accordance with the case studies on the application of FSM, the present study applied the method using three frequencies in data acquisition in Lengshuijiang City (27 o 40 0 45''N and 111 o 26 0 27''E). Geologically, the considered region is carboniferous limestone exposed on the surface. Data were acquired using TC300 equipment with electrode spacing (MN) and station intervals of 10 m and 5 m, respectively. The drilling position was set at 15 m from the start of the profile represented as a black point in Fig. 8. The water flow rate was about 100 t/day drilled at 100-m depth, while at 150-m depth, water yield exceeded 300 t/day. The main outlet depth is about 120 m. Figure 8 is the curve chart of measured results of the FSM at three frequencies in Lengshuijiang City ( Fig. 9), Hunan Province, China. Due to only a few frequencies being observed at that time, it is difficult to clearly recognize the cause of the anomaly from these curves. For the next study, we adopted a TC150 system, advanced and intelligent equipment capable of measuring multiple frequencies. Figure 10 shows the field measurement photograph and relative results of the TC150 frequency selector in the karst exploration of a limestone mine in Linxiang City. The sampling frequencies are 36 within the range of 20-5000 Hz. The results show the noodles phenomenon possibly emerged from the static shift due to the presence of near-surface water-filled karst voids. The curve features of this pseudo-section are similar to those of the theoretical finding (Fig. 7 vs. Fig. 10b). Abnormal causes of the FSM may be attributed to the static shifts. For the direct evidence to prove the finding, the presence of these voids is varied by three drilling boreholes. The drilling positions are marked as YZ4, ZK2, and ZK1 in Fig. 10b.  Figure 11 shows the results from the application of the TC300 frequency selector exploration for mineral water detection at a site (28 o 29 0 48''N, 112 o 3 0 21''E) in Taojiang County, Hunan Province (Fig. 9). Figure 11a shows plots of the profile curves of 20 frequencies, while Fig. 11b shows the potential difference DV pseudo-section of 40 frequencies. The electrode spacing MN is 10 m, and the station interval is 2 m in this case. The late Caledonian granodiorite was found exposed in the area. As the study area is rural residential, and there are transmission lines crossing, the power of the lines was shut down at the time of data acquisition which inhibits the possible impacts of transmission lines on measurements.
It can be seen from Fig. 11a that a relatively very low-potential anomaly occurred at the 18th m of the survey line, and the ''noodles phenomenon'' appears in Fig. 11b, whose features are similar to that of Fig. 7d. Based on past experiences, this location can be marked as a potential site to find water for groundwater exploration. However, according to the results of the site investigation, the anomaly is located just above an underdrain drainage channel of a small local reservoir (Fig. 12). The channel can be divided into two parts, the rectangular ditch as a main part having dimensions of a top buried depth of  Fig. 11 is caused by the water-semi-filled underdrain.
Therefore, this field work further verified that the static shift is the main reason for the FSM anomaly.

Summary
The present study applied the MT two-dimensional finite element method to develop some theoretical insights into the causes of anomaly formulation in FSM, focusing on the study of TM polarization mode, which is similar to the common field observation method of FSM. We calculated the profile curves and pseudo-sections of the horizontal electric field component E y along the direction of the survey line based on the simulation of apparent resistivity q s profile curves and pseudo-sections. Results show obvious uplift in E y profile curves with the increase in frequency in the Cartesian coordinate system in relation to the presence of near-surface low-resistance anomalies, which is actually caused by the static shift phenomenon of MT. The pseudo-section of E y also shows a significant static shift feature, which is similar to that of q s , but with the isoline stretching in the opposite direction.
It can be concluded from the two considered theoretical models that the static shift in the horizontal electric field component E y profile curve generated by the natural electromagnetic field at the surface is an obvious phenomenon in the Cartesian coordinate system, but the relationship between the lifting direction of the profile curve and the frequency is opposite to that of the q s profile curve.
Further, we discussed the measured results of FSM in karst exploration of limestone mines and the groundwater exploration in limestone and granite areas and analyzed the underground geology conditions, and measured curves and pseudo-sections. It is verified that the anomaly of FSM is naturally caused by the existence of inhomogeneous bodies in the near-surface zone. This demonstrates that the FSM, which has been applied in practice for about 40 years, actually uses the static shift phenomenon of the electromagnetic method. Therefore, the FSM method can be called the ''static shift method.'' FSM is a passive source method, and its practical application is mainly concentrated in hydrogeology and engineering geology exploration within about 300 m depth. The field sources are not only homologous with MT but also possibly associated with surface humanistic electromagnetic signals to some extent. This article innovatively proposed the viewpoint of the ''static shift method,'' which only plays a role in starting a further discussion on this issue. The effective utilization of static shift is like ''turning waste into treasure,'' which is expected to be further studied by more experts and scholars.