CN113484920B - Two-dimensional structured inversion method for frequency domain electromagnetic sounding data - Google Patents
Two-dimensional structured inversion method for frequency domain electromagnetic sounding data Download PDFInfo
- Publication number
- CN113484920B CN113484920B CN202110943184.5A CN202110943184A CN113484920B CN 113484920 B CN113484920 B CN 113484920B CN 202110943184 A CN202110943184 A CN 202110943184A CN 113484920 B CN113484920 B CN 113484920B
- Authority
- CN
- China
- Prior art keywords
- data
- electric field
- inversion
- frequency
- imaging
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V3/00—Electric or magnetic prospecting or detecting; Measuring magnetic field characteristics of the earth, e.g. declination, deviation
- G01V3/38—Processing data, e.g. for analysis, for interpretation, for correction
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/30—Assessment of water resources
Abstract
The invention discloses a two-dimensional structured inversion method for frequency domain electromagnetic sounding data, which comprises the following steps: acquiring frequency domain horizontal electric field data along the direction of the measuring line; adopting a space distribution process by using an electrostatic field formula to simulate abnormal electric field distribution; frequency division information extraction is carried out on electric field abnormal data in an abnormal electric field; carrying out electric field abnormal information classification pre-judgment on the single frequency data after the extraction treatment; performing local optimization inversion within a defined range on the anomalies of each pre-determined electric field; performing joint probability imaging on the peripheral nodes and the grid nodes; carrying out fusion processing on inversion imaging results of high-frequency, medium-frequency and low-frequency points; designing a structured resistivity initial model according to the multi-frequency fusion imaging result; and providing the structured initial model to a two-dimensional inversion program to obtain a final electrical distribution section result. The technical scheme of the invention aims to solve the problems of inaccurate positioning of the induction source by electromagnetic field imaging and distortion influence of the induction source on observed data.
Description
Technical Field
The invention belongs to the field of geophysical exploration, and particularly relates to a frequency domain electromagnetic sounding data two-dimensional structured inversion method based on an abnormal body recognition technology.
Background
Electromagnetic field inversion and imaging interpretation methods have been developed to different extents in recent years, and the understanding of electromagnetic fields of complex geological structures is increased. However, there are still problems faced in the inversion and imaging of actual data. For example, three-dimensional electromagnetic field simulation in approximate inversion is still limited to regular and simpler models, or one-dimensional or two-dimensional formal interpretation alternatives are adopted, otherwise simulation difficulty, calculation amount and cost remain major problems. In the imaging inversion, although continuous imaging in space can be achieved, higher requirements are put on data acquisition precision and acquisition density, and the exploration cost is greatly improved. For these reasons, there is a desire to develop new inversion imaging methods.
In the processing of frequency domain electromagnetic sounding actual measurement data, a strategy of correcting first and then inverting resistivity is generally adopted for static effects caused by shallow electrical non-uniformities. Technically, the application of tensor impedance decomposition techniques can be effectively corrected and can suppress other disturbances (Groom, r.w., bailay, R, C,1989, zhao Guoze, shang Ji, et al 1996, wei Sheng, wang Guying, et al 1996,F.E.M.Lilley,1998,G.W.McNeice andA.G.Jones 2001). Based on the recognition that static displacement is "high frequency noise" in cross section, spatial filtering and electromagnetic array methods (EMAP) can overcome the static displacement effect more effectively (Wang Guying et al 1990, 1992).
However, the correction of the static effect is to manually erase the objectively existing electrical non-uniformity, so that the processing ensures the relative accuracy of inversion results of the electrical distribution in the middle and deep parts of the section, but destroys the high resolution characteristic of the method itself on the distribution of the shallow earth medium. In addition, static effect can be caused, namely, not only the shallow part non-uniform body, but also the electric non-uniform body positioned in the section and in the deep part is shown by simulation experiments, the distortion influence can be caused on observed data, only the affected frequency band is shifted backwards, the simple full-frequency band translation influence characteristic is not shown any more! Therefore, various static correction methods developed in the past are no longer applicable to the treatment of these distortion effects.
Disclosure of Invention
Aiming at the problems of inaccurate positioning of an induction source and distortion influence of the induction source on observed data in the electromagnetic field imaging in the prior art, the invention provides a two-dimensional structured inversion method of frequency domain electromagnetic sounding data based on an abnormal body recognition technology.
In order to realize accurate positioning of the induction source and effectively eliminate adverse effects on inversion results caused by the accurate positioning, the invention adopts the following technical scheme:
a two-dimensional structured inversion method for frequency domain electromagnetic sounding data comprises the following steps:
s1, acquiring frequency domain horizontal electric field data along the direction of a measuring line, and then converting the frequency domain horizontal electric field data into a data form which is sensitive to the spatial position of an induction source only;
s2, performing spatial distribution processing by adopting an electrostatic field formula to simulate abnormal electric field distribution;
s3, frequency division information extraction is carried out on electric field abnormal data in the abnormal electric field;
s4, carrying out electric field abnormal information classification pre-judgment on the single frequency data after the extraction processing;
s5, performing local optimization inversion within a defined range on the anomalies of each pre-judging electric field to obtain the spatial positions of the distribution of the induction sources;
s6, after the main body control grid of the induction source distribution is determined, performing joint probability imaging on peripheral nodes and grid nodes;
s7, fusing the inversion imaging results of the high-frequency, medium-frequency and low-frequency points;
s8, designing a structured resistivity initial model according to a multi-frequency fusion imaging result;
and S9, providing the structured initial model to a two-dimensional inversion program, and performing iterative correction to obtain a final electrical distribution section result.
Preferably, the frequency domain horizontal electric field data is AMT data or CSAMT data.
Preferably, the data in S1 is converted into:
wherein N is the number of the observed data; t [. Cndot.]Representing a collaborative normalization transformation on a set of data; d (D) i And (3) representing the observation data of the ith measuring point after transformation, wherein y is a data set.
Preferably, in S2, the abnormal electric field is simulated by the theory of electrostatic field, which is expressed as follows:
wherein y is o For observing the horizontal coordinate of the point, y i 、z i Respectively the horizontal coordinate and the vertical coordinate of the ith induction source, lambda i For corresponding induced line charge density E y Epsilon is the horizontal component of the electric field 0 Is a dielectric constant.
Preferably, S4 specifically includes:
s41, counting the abnormal number of the electric fields, and counting the abnormal positions, peak values and bilateral curve symmetry or slow and steep conditions of each electric field;
s42, judging the electrical interface form causing the abnormality of each electric field based on the characteristic of each electric field abnormality curve;
s43, preliminarily determining basic configuration characteristics for inversion according to the pre-judgment of the electrical interface morphology.
Preferably, S5 specifically includes:
s51, according to the judgment and measurement of the SS4 on the abnormity of each electric field, defining the range of data to be fitted for each abnormal inversion;
s52, defining an imaging range of local optimization inversion according to the judgment and measurement of the electric field anomalies by SS 4;
s53, searching an optimal solution of the selected inversion basic configuration by fitting measured data in a specified observation section in a defined ground electric space imaging range, and simultaneously, accurately positioning the space distribution position of the induction source by accepting the solution in an allowable deviation range;
s54, after inversion imaging processing is completed on the electric field anomalies, an induction source spatial distribution control grid reflected by the single-frequency data is obtained, and therefore local accurate imaging processing on induction source distribution is completed.
Preferably, S6 specifically includes:
s61, designing a two-dimensional grid of an imaging area, which comprises the following steps: the horizontal coordinates of each column node and the depth of each row node of the grid;
s62, starting from a first node at the upper left corner of the imaging area, scanning each node of the two-dimensional grid in sequence from left to right and from top to bottom until the last node at the lower right corner of the imaging area; at the current node i of the scan, heuristically setting an induction source, respectively performing forward modeling on the intensities of the induction source in three conditions of-1, 0 and 1, respectively solving fitting differences between three modeling results and measured data, and taking the absolute value with the smallest value to determine m i,i The value, S, is calculated as i ：
Wherein M represents the total number of spatial positions for which a probability value is to be calculated; p (P) i A probability value exists for the induction source at the ith location; t [. Cndot.]Representing a collaborative normalization transformation on a set of data; n is the total number of observed data, D j The observation data of the j-th measuring point; m is m i Model vectors used for forward modeling; f (F) j [m i ]The value of the abnormal response at the jth measuring point when the induction source is at the ith spatial position; s represents the dataset in brackets, S i Then it is the i-th element thereof;
s63, simulating measured data when the abnormal inversion solutions of the electric fields exist simultaneously for grid nodesThe S is calculated according to the situation i Values, i.e. S of all trellis nodes i The absolute values are the same;
s64, calculating the S at all the nodes i And (3) carrying out normalization transformation on the value data set to realize joint probability imaging calculation of peripheral nodes and grid nodes in a ground electric space.
Preferably, the electrical interface in SS42 is in the form of a vertical line segment, an inclined line segment, or a curved line segment.
Preferably, in S8, the multifrequency fusion imaging result is an inversion result at each frequency point data control grid, which is a reflection of the existence position of the induction source; and judging the relative level of the earth medium resistivity at two sides of the electrical interface according to the positive and negative of the induced charges, and then completing the electrical assignment operation of different parts in the whole ground electric section according to geological, physical property, logging and seismic data, so as to complete the design of the initial structured resistivity model.
The two-dimensional structured inversion method for the frequency domain electromagnetic sounding data has the beneficial effects that:
according to the invention, actually measured horizontal electric field data are derived along the direction of a measuring line and converted through normalization processing, and MT finite element forward modeling experiments prove that the converted data are only sensitive to the spatial position of an induction source in a ground electric space, namely the structural characteristics reflecting the electrical anomaly distribution of a ground medium are highlighted, and the characteristics lay a foundation for the research of an electrical anomaly identification technology from an anomaly electric field.
The invention adopts the strategy of simulating the distribution of the abnormal electric field by adopting the electrostatic field theoretical formula, compared with the numerical simulation method such as finite element, the invention greatly reduces the workload of simulation calculation and greatly reduces the difficulty of developing further research on the abnormal electric field.
According to the frequency division extraction technology of electric field abnormal information, the high-frequency observation data is subtracted by the low-frequency observation data, so that the electric abnormal distribution structure information reflected by the low-frequency data only can be obtained, inversion and imaging processing are carried out on the extracted low-frequency data, the resolution capability of deep electric field abnormal distribution can be greatly improved, and the knowledge of industry personnel on the real resolution capability of a frequency domain electromagnetic sounding method can be possibly completely updated through development of the research.
The method carries out classification prejudgment on the single frequency data electric field abnormity information and carries out local optimization inversion technology of a defined range on each prejudgment electric field abnormity, and can realize accurate positioning on the spatial distribution of induction sources, thereby realizing local high-resolution imaging on a main body control grid, and being a key point of greatly improving the transverse resolution capability of an abnormal electric structure.
According to the invention, after the main body control grid of induction source distribution is determined by the inversion, the grid node scanning function of the periphery is calculated, and finally the grid node scanning function is combined with the grid node to image, so that the research on the spatial distribution condition of the induction source in the single frequency data exploration depth range is completed, and the imaging result of the periphery of the grid mainly plays a role of setting off a background of the induction source distribution.
In order to obtain the research result of the full depth of the section (the maximum depth reflected by the data from the ground surface to the lowest frequency), the inversion imaging results of all frequency points such as high frequency, medium frequency, low frequency and the like are required to be fused, namely the imaging results of all frequency point data are overlapped.
Based on the full-depth imaging result of the spatial distribution of the induction source, the invention mainly judges the relative height of the earth medium resistivity at the two sides of the electrical interface according to the positive and negative of the induction charge, and then completes the design work of the initial model of the structured resistivity by assisting with the known data of geology, physical property, logging, earthquake and the like in the work area.
On the basis of determining the abnormal electric field distribution rule, the invention solves the problem that the current electromagnetic field imaging cannot accurately and spatially position the induction source by a series of techniques; meanwhile, by imaging the distribution condition of induction sources with different depths in the section, a structural resistivity model design scheme is provided, so that a foundation is laid for fundamentally and positively solving the adverse effect on inversion results caused by observed data distortion caused by abnormal bodies with different depths.
The invention fully extracts the resolution potential of the frequency domain electromagnetic sounding method while respecting all information contained in measured data, and applies more objective and more reasonable structured electrical distribution information to the two-dimensional resistivity iterative inversion process, thereby providing more ideal inversion processing results for geological interpretation work, and providing a step of great practical value!
Drawings
FIG. 1 is a technical roadmap of a two-dimensional structured inversion method of frequency domain electromagnetic sounding data based on an outlier recognition technology;
FIG. 2 is a graph of a single rule vertical electrical interface model and its simulation results versus analysis;
FIG. 3 is a schematic cross-sectional series contrast diagram of a ground model including three non-uniformities and simulation results thereof after electric field anomaly frequency division extraction information;
FIG. 4 is a schematic diagram of three sets of charge distribution, and superimposed information of two anomalies identified by the electric field anomaly simulation curve and inversion-probability imaging method, and the result of the spatial distribution imaging of the induction sources;
fig. 5 is a series of contrast graphs of two resistivity profiles obtained by performing conventional inversion on the simulation result of the theoretical model in fig. 3, and the inversion result obtained by performing multi-frequency point inversion imaging fusion processing on the spatial distribution of the induction source by adopting the method and performing two-dimensional iterative correction on the designed structured electric model.
Detailed Description
The invention is further described below with reference to the accompanying drawings:
FIG. 1 is a technical roadmap of a two-dimensional structured inversion method of frequency domain electromagnetic sounding data based on an outlier recognition technology. The implementation of the invention comprises the following steps:
s1, acquiring frequency domain horizontal electric field data along the direction of a measuring line, and then converting the frequency domain horizontal electric field data into a data form which is sensitive to the spatial position of an induction source only, namely
Wherein N is the number of the observed data; t [. Cndot.]Representing a collaborative normalization transformation on a set of data; y is a data set; d (D) i The observation data representing the i-th measurement point after transformation, the value of which belongs to the set [ -1,1]A kind of electronic device. The frequency domain horizontal electric field data can be AMT data of a natural source, or CSAMT, time-frequency electric method and wide area electric method data of an artificial source.
The meaning of making such a conversion to measured data is illustrated in a theoretical model. FIG. 2 shows a graph (a) of a fault model with a vertical electrical interface, wherein the resistivity of the high-resistance layer is 1000 Ω.m, and the low-resistance layer is 10 Ω.m and 100 Ω.m, respectively, for MT finite element forward modeling; the graph (b) is a calculation result of the real part Eyr of the horizontal component of the electric field in the observation period t= 10.666667s, and the graph shows that the electric difference relationship between the magnitude of the electric field value and the electric property in the model is obvious, that is, the smaller the resistivity of the low-resistance layer is, the smaller the observed electric field value is, and the fluctuation amplitude of the curve is flattened; FIG. (c) is dE yr The two curves are basically coincident, which means that the transformed data is independent of specific resistivity value of the low-resistance layer, i.e. the transformed data is basically the same as long as the pattern of the electrical distribution is unchanged, which means that the transformed data is sensitive to only the spatial position where the induced charges exist and basically independent of the intensity of the induced charges, which is the characteristic of the data, and the study of the abnormal body recognition technology is greatly facilitated.
S2, adopting a space distribution process by using an electrostatic field formula to simulate abnormal electric field distribution
Because the magnetotelluric field is quasi-steady, the distribution characteristics of the underground electromagnetic field can be described by electrostatic fields when the observation frequency is low enough in a certain range of electricity. In view of this, the present invention simulates an abnormal electric field by the theory of electrostatic field. The formula used is as follows:
wherein y is o For observing the horizontal coordinate of the point, y i 、z i Respectively the horizontal coordinate and the vertical coordinate of the ith induction source, lambda i For corresponding induced line charge density E y Epsilon is the horizontal component of the electric field 0 Is a dielectric constant.
In order to examine how the approximation effect of the abnormal electric field calculated by the above formula to the finite element corresponding simulation result, the present invention makes the following experiments: on a vertical electrical interface of a fault model (see fig. 2 (a)), an induction source with a linear charge density of 1 (on a discontinuous electrical interface from low resistance to high resistance of an electric field, the induction linear charge is positive) is arranged, a calculation result of the induction source is shown as a 2 nd curve in fig. 2 (d), a 1 st curve is a simulation result corresponding to a finite element, and the two are good in consistency. Through the experiment, the feasibility and effectiveness of simulating an abnormal electric field by using a formula in an electrostatic field under a certain condition are proved. Obviously, in the process of carrying out inversion imaging on the induction source by using abnormal field data, a simple analytical formula is used for simulating an abnormal field, so that the calculation workload is greatly reduced, and the inversion imaging speed is greatly improved.
S3, frequency division information extraction is carried out on electric field abnormal data in abnormal electric fields
For researching the spatial distribution condition of induction sources with different depths, observation data with different frequencies are sequentially used, and in order to improve the longitudinal resolution capability of an imaging result, a frequency division extraction technology is adopted for electric field anomaly information, namely high-frequency observation data is subtracted from low-frequency observation data, so that anomaly electrical structure information reflected by only the low-frequency data (the high frequency and the low frequency are relatively speaking) can be obtained. FIG. 3 is a diagram of (a) a model of ground electricity comprising three non-uniformities; (c) The electric field anomaly simulation result shows that the reflection of the electric interface between the two vertical electric surfaces of the superficial non-uniform body is clear and strong, the reflection of the middle-deep non-uniform body is distinguishable, and the reflection of the deep non-uniform body is blurred. (d) The graph is a result of frequency division extraction of electric field anomaly information, and the graph shows that the reflection of the extracted middle-frequency band, low-frequency band information pair and deep anomaly is basically clear and displayed to the same extent as the reflection of the high-frequency band on the shallow anomaly. Therefore, inversion and imaging processing are carried out on the extracted middle-frequency band and low-frequency band data, so that the resolution capability of middle and deep ground electricity abnormal distribution can be greatly improved.
S4, performing electric field abnormal information classification pre-judgment on the single frequency data after the extraction processing
S4-1, analyzing and counting the abnormal number of the electric fields, and calculating the positions, peak values and bilateral curve symmetry or slow and steep conditions of the abnormal electric fields;
s4-2, judging possible electrical interface forms causing the abnormality of each electric field, such as forms of a vertical line segment, an inclined line segment, a curved line segment and the like, based on the analysis of the abnormal curve characteristics of each electric field;
s4-3, for complex electrical interface distribution situations, feedback judgment and measurement can be carried out on the interface form in a man-machine interaction mode; firstly, a basic electrical interface configuration, such as an inclined line segment or a curved line segment, can be selected from a software system interactive panel, an abnormal electric field is simulated by arranging induced charges on the interface, and then whether the selected configuration is reasonable or not can be roughly judged through the comparison analysis of a simulated curve and an actual measured curve, namely the coincidence condition of the curve forms of the simulated curve and the actual measured curve; if the deviation is larger, S such as interface morphology modification, abnormal electric field simulation, curve comparison analysis and the like can be edited and modified, and repeated man-machine interaction operation is carried out until feedback judgment and measurement of the complex electric interface distribution morphology are completed; and finally forming a basic electrical interface configuration design through the feedback judgment process so as to provide the local optimization inversion of the subsequent defined range;
s4-4, preliminarily determining basic configuration characteristics for inversion, such as configurations of points, line segments, bending curves and the like, according to the pre-judgment of the electrical interface morphology.
As shown in fig. 4, wherein graph (a) is a theoretical model comprising three groups of shallow, medium, and deep charge distributions; the graph (b) is a simulated abnormal electric field curve, according to analysis on curve morphology, two electric field anomalies can be distinguished firstly, and finally, point-shaped points can be selected as basic configuration of inversion imaging according to symmetry of curves at two sides of a peak value. Of course, the curve also includes weak electric field abnormality information caused by deep charge combination, and the weak information is processed by the frequency division extraction technology.
S5, carrying out local optimization inversion in a defined range on the anomalies of each pre-determined electric field to obtain the spatial position of the distribution of the induction source
S5-1, according to the judgment and measurement of the anomaly of each electric field by SS4, defining the range of data to be fitted for each anomaly inversion, for example, for a punctiform configuration, a single-side radius with an anomaly peak value as the center can be designated;
s5-2, according to the judgment and measurement of the SS4 on the abnormity of each electric field, defining the imaging range of local optimization inversion, for example, for a punctiform configuration, the imaging range can be determined to be right under each abnormal peak value and the like;
s5-3, searching an optimal solution of the selected inversion basic configuration by fitting measured data in a specified observation section in a defined ground electric space imaging range, and simultaneously, accurately positioning the space distribution position of the induction source by accepting the solution in an allowable deviation range;
s5-4, after inversion imaging processing is completed on the electric field anomalies, an induction source spatial distribution control grid reflected by the single-frequency data can be obtained, and therefore local accurate imaging processing on induction source distribution is completed.
As shown in fig. 4 (d), the white area in the graph is the locally accurate imaging result of the determined induction source spatial distribution control grid through local optimization inversion of a defined range, and fig. 4 (c) is the fitting condition of the inversion result to measured data, namely, the abnormal combination of shallow and middle charges in the inversion result is resolved and processed through the research of single-frequency data.
S6, after the main body control grid for determining induction source distribution, performing joint probability imaging on peripheral nodes and grid nodes
Specifically, the method comprises the following steps:
s6-1, designing a two-dimensional grid of the imaging area according to the actual measurement data of the frequency domain sounding or the forward modeling result, wherein specific parameters comprise horizontal coordinates of nodes in each column of the grid, depths of nodes in each row and the like.
S6-2、For the peripheral nodes, starting from the first node at the upper left corner of the imaging area, each node of the two-dimensional grid is scanned in turn in the order from left to right and from top to bottom until the last node at the lower right corner of the imaging area. At the current node i of the scan, heuristically setting induction sources, respectively performing forward modeling on three conditions of intensities of-1, 0 and 1 (respectively representing negative charge, no charge and positive charge), respectively solving fitting differences between three modeling results and measured data, and taking the smallest absolute value to determine m i,i Value (m when heuristic charge is 0 and 1) i,i =1; otherwise m i,i = -1), S is calculated as follows i ：
Wherein M represents the total number of spatial positions for which probability values are to be calculated; p (P) i The probability value exists for the induction source at the ith position, and the size of the probability value is between-1 and 1; t [. Cndot.]Representing a collaborative normalization transformation on a set of data; n is the total number of observed data, D j The observation data of the j-th measuring point; m is m i Model vectors for forward modeling, in which there is typically only the ith component m i,i The value is 1 or-1, and the values of the rest components are 0; f (F) j [m i ]The value of the abnormal response at the jth measuring point when the induction source is at the ith spatial position; s represents the dataset in brackets, S i Then it is the i-th element thereof.
S6-3, calculating the S by adopting the fitting condition of the abnormal inversion solutions of the electric fields to the measured data when the abnormal inversion solutions of the electric fields exist simultaneously for the grid nodes i Values, i.e. S of all trellis nodes i The absolute values are the same;
s6-4, calculating the S at all the nodes i And (3) carrying out normalization transformation on the value data set to finish joint probability imaging calculation of peripheral nodes and grid nodes in a ground electric space.
For the theoretical model shown in fig. 4 (a), the effect of the combined imaging is shown in fig. 4 (d), i.e. the imaging result gives very accurate spatial localization for both shallow and deep section charge combinations.
S7, fusing the inversion imaging results of the high-frequency, medium-frequency and low-frequency points
After the processing of the two S5 and S6 is completed on each single frequency data, in order to obtain the imaging result of the full depth ground electric space, the inversion-probability imaging result of each frequency point such as high frequency, intermediate frequency and low frequency needs to be fused, and in the simplest case (that means that the intermediate frequency data is excluding the high frequency information, the low frequency data is excluding the intermediate frequency information, etc.), the imaging result of each frequency point data only needs to be overlapped and synthesized.
As shown in the theoretical model in fig. 3 (a), in the simulated section diagram shown in fig. 3 (d), the observation period T is 0.008333s, 0.100000s and 0.888890s respectively, after imaging is completed sequentially, the effect of fusing the imaging results of the three frequency points is shown in fig. 5 (c), and it can be seen from the figure that the multi-frequency fusion result gives relatively accurate spatial positioning to the distribution of the vertical electrical interfaces of the three non-uniform bodies.
S8, designing a structured resistivity initial model according to the multi-frequency fusion imaging result
The multi-frequency fusion imaging result, which is based on the inversion result of the data control grid of each frequency point, is an accurate reflection of the existence position of the induction source, the relative height of the earth medium resistivity at two sides of the electric interface can be judged according to the positive and negative of the induction charge, and the known geological, physical, logging, earthquake and other data in a research area can be aided to finish the electric assignment operation of different parts in the whole ground electric section for the next time, so that the design work of the structured resistivity initial model has not been paid enough attention to the processing of the frequency domain electromagnetic sounding data, and is one of the main reasons that the inversion result of the electric method data is generally difficult to meet geological interpretation requirements.
The structural resistivity model design process comprises the following steps:
(1) For a theoretical model, the background resistivity value of the ground electric space can be given according to the first high-frequency point data of the measured apparent resistivity curve which is not basically influenced by the static effect; for an abnormal body distribution area, firstly, drawing the distribution profile of the abnormal body distribution area through a screen modeling software system according to the accurate identification result of the multi-frequency inversion-probability imaging pair non-horizontal electrical interface; finally, whether the abnormal body is a low-resistance body or a high-resistance body is judged according to the polarity of the induced charges on the electrical interface (namely, positive charges are accumulated at the interface when the abnormal body enters the high-resistance body from a low-resistance background along the direction of the measuring line, and negative charges are accumulated on the interface), and the electric interface is filled with relatively low-resistance or high-resistance electrical values, so that the design work of the initial model of the structured resistivity is completed.
(2) In the processing of actual production data, according to various mastered data such as geology, physical properties, well logging, earthquakes and the like in a research area and a neighboring area, main stratum in the detection depth range of the research area is subjected to arrangement and analysis work of the physical property data of each stratum representative of lithology, when the collected data are insufficient, a certain amount of field sample collection and laboratory physical property test data are supplemented to combine the results such as average thickness and the like counted by each main stratum in the area, and the construction work of an intra-area layer background ground electric model is completed; then, based on the multi-frequency inversion-probability imaging result, the given accurate position of the distribution of the induction source, and according to the polarity of the induction charge, judging the relative height of the earth medium resistivity at the two sides of the electrical interface; and finally, the background stratum and the abnormal body are sequentially drawn out through a screen modeling software system, and corresponding electrical values are filled, so that the design work of the initial model of the structured resistivity in the whole section is completed for the next time.
And S9, providing the structured initial model to a two-dimensional inversion program, and performing iterative correction to obtain a final electrical distribution section result.
As shown in the theoretical model of fig. 3 (a), fig. 3 (b) shows TM mode apparent resistivity simulation results, it can be seen that each of the inhomogeneities has a serious distortion effect on the measured data, the effect of the shallow inhomogeneity appears as a full band, and the band affected by the middle and deep inhomogeneities is delayed, and two results are usually obtained by performing inversion processing using the data containing the distortion effect, see fig. 5 (a) and (b), and the two results have a better lateral resolution on the inhomogeneities, so that the problem is that the reflection on the anomaly is shallow in depth, small in scale, or the phenomenon of hanging a "noodle" cannot cover the bottom, so that the two inversion results cannot meet the requirements of actual production. The data processing strategy provided by the invention is adopted to provide the structured resistivity initial model designed and completed according to the figure 5 (c) for a two-dimensional NLCG inversion program, the final processing effect is shown in the figure 5 (d), and the graph shows that the processing result gives near perfect display on the distribution of three abnormal bodies of shallow, medium and deep.
According to the two-dimensional structured inversion method for the frequency domain electromagnetic sounding data, frequency domain horizontal electric field data along the direction of a survey line are obtained and then converted into a data form which is sensitive to the space position where induced charges exist only; comparing the theoretical model of the same induction source distribution with the calculation result of the electrostatic field formula through MT finite element forward modeling, and confirming that the distribution of an abnormal electric field can be effectively simulated by using the electrostatic field theoretical formula; the research on the spatial distribution condition of induction sources with different depths is carried out by using observation data with different frequencies, and in order to improve the longitudinal resolution capability of an imaging result, a frequency division extraction technology is adopted for electric field abnormality information; to improve the transverse resolution, the electric field abnormality information is judged, and reasonable prediction is given to the distribution range and the form of the induction source; on the basis of prejudging the space distribution characteristics of the induction sources, carrying out local optimization inversion of a defined range on each electric field abnormality, namely imaging a specified space range by fitting main body characteristic data of each abnormality, and only receiving inversion results within allowable deviation, thereby obtaining accurate space distribution results of the induction sources; according to the distribution condition of induced charges in the multi-frequency fusion imaging result, the relative height of the earth medium resistivity at two sides of the electrical interface can be judged, and then according to known geological, physical, logging, earthquake and other data in a research area, the electrical filling work of different parts in the whole section can be completed next time; after the structured resistivity model is designed, the structured resistivity model can be used as an initial model and provided for a two-dimensional inversion program, and a final imaging result can be obtained through iterative correction. The method can solve the problem that the processing effect is seriously affected due to the inclusion of the static effect in the frequency domain electromagnetic sounding data, changes the previous thought of correcting and then inverting the static effect, fully exploits the resolution potential of the frequency domain electromagnetic sounding method by the proposal of the abnormal body recognition technology, and thus makes significant contribution to the promotion of the progress of the data processing technology of the method.
The above description of the embodiments of the present invention has been provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and all the inventions using the inventive concept are protected as long as various changes are within the technical spirit of the present invention defined and defined by the appended claims.
Claims (9)
1. The two-dimensional structured inversion method for the frequency domain electromagnetic sounding data is characterized by comprising the following steps of:
s1, acquiring frequency domain horizontal electric field data along the direction of a measuring line, and then converting the frequency domain horizontal electric field data into a data form which is sensitive to the spatial position of an induction source only;
s2, performing spatial distribution processing by adopting an electrostatic field formula to simulate abnormal electric field distribution;
s3, frequency division information extraction is carried out on electric field abnormal data in the abnormal electric field;
s4, carrying out electric field abnormal information classification pre-judgment on the single frequency data after the extraction processing;
s5, performing local optimization inversion within a defined range on the anomalies of each pre-judging electric field to obtain the spatial positions of the distribution of the induction sources;
s6, after the main body control grid of the induction source distribution is determined, performing joint probability imaging on peripheral nodes and grid nodes;
s7, fusing the inversion imaging results of the high-frequency, medium-frequency and low-frequency points;
s8, designing a structured resistivity initial model according to a multi-frequency fusion imaging result;
and S9, providing the structured initial model to a two-dimensional inversion program, and performing iterative correction to obtain a final electrical distribution section result.
2. The two-dimensional structured inversion method of frequency domain electromagnetic sounding data according to claim 1, wherein the frequency domain horizontal electric field data is AMT data or CSAMT data.
3. The two-dimensional structured inversion method of frequency domain electromagnetic sounding data according to claim 1, wherein the data in S1 is converted into:
wherein N is the number of the observed data; t [. Cndot.]Representing a collaborative normalization transformation on a set of data; d (D) i And (3) representing the observation data of the ith measuring point after transformation, wherein y is a data set.
4. The two-dimensional structured inversion method of frequency domain electromagnetic sounding data according to claim 1, wherein the abnormal electric field is simulated by using an electrostatic field theory in S2, and the method is expressed as follows:
wherein y is o For observing the horizontal coordinate of the point, y i 、z i Respectively the horizontal coordinate and the vertical coordinate of the ith induction source, lambda i For corresponding induced line charge density E y Epsilon is the horizontal component of the electric field 0 Is a dielectric constant.
5. The two-dimensional structured inversion method of frequency domain electromagnetic sounding data according to claim 1, wherein S4 specifically comprises:
s41, counting the abnormal number of the electric fields, and counting the abnormal positions, peak values and bilateral curve symmetry or slow and steep conditions of each electric field;
s42, judging the electrical interface form causing the abnormality of each electric field based on the characteristic of each electric field abnormality curve;
s43, preliminarily determining basic configuration characteristics for inversion according to the pre-judgment of the electrical interface morphology.
6. The two-dimensional structured inversion method of frequency domain electromagnetic sounding data according to claim 1, wherein S5 specifically comprises:
s51, according to the judgment and measurement of the electric field anomalies in the S4, defining the range of the data to be fitted for the anomaly inversion;
s52, defining an imaging range of local optimization inversion according to the judgment and measurement of the electric field anomalies in the S4;
s53, searching an optimal solution of the selected inversion basic configuration by fitting measured data in a specified observation section in a defined ground electric space imaging range, and simultaneously, accurately positioning the space distribution position of the induction source by accepting the solution in an allowable deviation range;
s54, after inversion imaging processing is completed on the electric field anomalies, an induction source spatial distribution control grid reflected by the single-frequency data is obtained, and therefore local accurate imaging processing on induction source distribution is completed.
7. The two-dimensional structured inversion method of frequency domain electromagnetic sounding data according to claim 1, wherein S6 specifically comprises:
s61, designing a two-dimensional grid of an imaging area, which comprises the following steps: the horizontal coordinates of each column node and the depth of each row node of the grid;
s62, starting from a first node at the upper left corner of the imaging area, scanning each node of the two-dimensional grid in sequence from left to right and from top to bottom until the last node at the lower right corner of the imaging area; at the current node i of the scan, heuristically setting an induction source, respectively performing forward modeling on the intensities of the induction source in three conditions of-1, 0 and 1, respectively solving fitting differences between three modeling results and measured data, and taking the one with the smallest absolute value to determineM is fixed i,i The value, S, is calculated as i ：
Wherein M represents the total number of spatial positions for which a probability value is to be calculated; p (P) i A probability value exists for the induction source at the ith location; t [. Cndot.]Representing a collaborative normalization transformation on a set of data; n is the total number of observed data, D j The observation data of the j-th measuring point; m is m i Model vectors used for forward modeling; f (F) j [m i ]The value of the abnormal response at the jth measuring point when the induction source is at the ith spatial position; s represents the dataset in brackets, S i Then it is the i-th element thereof;
s63, calculating the S by adopting the fitting condition of the abnormal inversion solutions of the electric fields to the measured data when the abnormal inversion solutions of the electric fields exist simultaneously for the grid nodes i Values, i.e. S of all trellis nodes i The absolute values are the same;
s64, calculating the S at all the nodes i And (3) carrying out normalization transformation on the value data set to realize joint probability imaging calculation of peripheral nodes and grid nodes in a ground electric space.
8. The method of two-dimensional structured inversion of frequency domain electromagnetic sounding data according to claim 5, wherein the electrical interface in SS42 is in the form of a vertical segment, an inclined segment, or a curved segment.
9. The two-dimensional structured inversion method of frequency domain electromagnetic sounding data according to claim 1, wherein in S8, the multi-frequency fusion imaging result is an inversion result at each frequency point data control grid, which is a reflection of the existence position of the induction source; and judging the relative level of the earth medium resistivity at two sides of the electrical interface according to the positive and negative of the induced charges, and then completing the electrical assignment operation of different parts in the whole ground electric section according to geological, physical property, logging and seismic data, so as to complete the design of the initial structured resistivity model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110943184.5A CN113484920B (en) | 2021-08-17 | 2021-08-17 | Two-dimensional structured inversion method for frequency domain electromagnetic sounding data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110943184.5A CN113484920B (en) | 2021-08-17 | 2021-08-17 | Two-dimensional structured inversion method for frequency domain electromagnetic sounding data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113484920A CN113484920A (en) | 2021-10-08 |
| CN113484920B true CN113484920B (en) | 2023-05-19 |
Family
ID=77945584
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110943184.5A Active CN113484920B (en) | 2021-08-17 | 2021-08-17 | Two-dimensional structured inversion method for frequency domain electromagnetic sounding data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113484920B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116047617B (en) * | 2023-03-10 | 2023-06-27 | 中国地质科学院地球物理地球化学勘查研究所 | Method and device for identifying geological features between wells |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1989000705A1 (en) * | 1987-07-20 | 1989-01-26 | University Of Utah Research Institute | Multi-array borehole resistivity and induced polarization system |
| WO2007126481A2 (en) * | 2006-04-06 | 2007-11-08 | Exxonmobil Upstream Research Company | Method for obtaining resistivity from controlled source electromagnetic data |
| WO2007130205A2 (en) * | 2006-05-04 | 2007-11-15 | Exxonmobil Upstream Research Company | Time lapse analysis with electromagnetic data |
| CN104422970A (en) * | 2013-09-04 | 2015-03-18 | 中国石油天然气集团公司 | Amplitude-preserving static correcting method for electromagnetic data |
| CA2937199A1 (en) * | 2015-07-23 | 2017-01-23 | Jose Manuel Martinez Ortega | Apparatus for airborne and ground electromagnetic prospecting and method thereof |
| CN109541695A (en) * | 2018-11-16 | 2019-03-29 | 吉林大学 | Artificial field source frequency domain electric-force gradient far-zone apparent resistivity fast imaging method |
| CN110348568A (en) * | 2019-07-16 | 2019-10-18 | 山东科技大学 | A kind of deep Mined-Out Areas method suitable for strong electromagnetic area |
| CN110488365A (en) * | 2019-09-11 | 2019-11-22 | 湖南五维地质科技有限公司 | A kind of multipolarization telluric electromagnetic sounding method |
| CN110989002A (en) * | 2019-11-07 | 2020-04-10 | 吉林大学 | High-precision data interpretation method for shallow low-resistance abnormal body of ground-air time domain electromagnetic system |
| CN111103628A (en) * | 2020-01-14 | 2020-05-05 | 桂林理工大学 | Two-dimensional inversion method and device for electric field data in magnetotelluric (TE) polarization mode |
| CN111103627A (en) * | 2020-01-14 | 2020-05-05 | 桂林理工大学 | Two-dimensional inversion method and device for electric field data by magnetotelluric (TM) polarization mode |
| CN111323830A (en) * | 2020-01-14 | 2020-06-23 | 东华理工大学 | Joint inversion method based on magnetotelluric and direct-current resistivity data |
| CN111352163A (en) * | 2020-03-03 | 2020-06-30 | 吉林大学 | Magnetotelluric depth sounding-based static effect correction method and system |
| CN111965712A (en) * | 2020-10-21 | 2020-11-20 | 国网江西省电力有限公司电力科学研究院 | Method for correcting static effect of controllable source audio magnetotelluric method |
| CN113126172A (en) * | 2020-01-16 | 2021-07-16 | 中国石油天然气集团有限公司 | Static displacement correction method and device |
| CN114265124A (en) * | 2021-12-30 | 2022-04-01 | 成都理工大学 | Unfavorable geologic body positioning method based on time domain transient electromagnetic probability inversion |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7299131B2 (en) * | 2004-12-17 | 2007-11-20 | Baker Hughes Incorporated | Induction resistivity imaging principles and devices in oil based mud |
| US9110183B2 (en) * | 2006-12-06 | 2015-08-18 | Technoimaging, Llc | Systems and methods for remote electromagnetic exploration for mineral and energy resources using stationary long-range transmitters |
| US10705240B2 (en) * | 2017-05-11 | 2020-07-07 | Saudi Arabian Oil Company | Capacitive electromagnetic formation surveillance using passive source |
-
2021
- 2021-08-17 CN CN202110943184.5A patent/CN113484920B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1989000705A1 (en) * | 1987-07-20 | 1989-01-26 | University Of Utah Research Institute | Multi-array borehole resistivity and induced polarization system |
| WO2007126481A2 (en) * | 2006-04-06 | 2007-11-08 | Exxonmobil Upstream Research Company | Method for obtaining resistivity from controlled source electromagnetic data |
| WO2007130205A2 (en) * | 2006-05-04 | 2007-11-15 | Exxonmobil Upstream Research Company | Time lapse analysis with electromagnetic data |
| CN104422970A (en) * | 2013-09-04 | 2015-03-18 | 中国石油天然气集团公司 | Amplitude-preserving static correcting method for electromagnetic data |
| CA2937199A1 (en) * | 2015-07-23 | 2017-01-23 | Jose Manuel Martinez Ortega | Apparatus for airborne and ground electromagnetic prospecting and method thereof |
| CN109541695A (en) * | 2018-11-16 | 2019-03-29 | 吉林大学 | Artificial field source frequency domain electric-force gradient far-zone apparent resistivity fast imaging method |
| CN110348568A (en) * | 2019-07-16 | 2019-10-18 | 山东科技大学 | A kind of deep Mined-Out Areas method suitable for strong electromagnetic area |
| CN110488365A (en) * | 2019-09-11 | 2019-11-22 | 湖南五维地质科技有限公司 | A kind of multipolarization telluric electromagnetic sounding method |
| CN110989002A (en) * | 2019-11-07 | 2020-04-10 | 吉林大学 | High-precision data interpretation method for shallow low-resistance abnormal body of ground-air time domain electromagnetic system |
| CN111103628A (en) * | 2020-01-14 | 2020-05-05 | 桂林理工大学 | Two-dimensional inversion method and device for electric field data in magnetotelluric (TE) polarization mode |
| CN111103627A (en) * | 2020-01-14 | 2020-05-05 | 桂林理工大学 | Two-dimensional inversion method and device for electric field data by magnetotelluric (TM) polarization mode |
| CN111323830A (en) * | 2020-01-14 | 2020-06-23 | 东华理工大学 | Joint inversion method based on magnetotelluric and direct-current resistivity data |
| CN113126172A (en) * | 2020-01-16 | 2021-07-16 | 中国石油天然气集团有限公司 | Static displacement correction method and device |
| CN111352163A (en) * | 2020-03-03 | 2020-06-30 | 吉林大学 | Magnetotelluric depth sounding-based static effect correction method and system |
| CN111965712A (en) * | 2020-10-21 | 2020-11-20 | 国网江西省电力有限公司电力科学研究院 | Method for correcting static effect of controllable source audio magnetotelluric method |
| CN114265124A (en) * | 2021-12-30 | 2022-04-01 | 成都理工大学 | Unfavorable geologic body positioning method based on time domain transient electromagnetic probability inversion |
Non-Patent Citations (3)
| Title |
|---|
| 基于异常体识别的电磁概率成像方法及应用分析;张翔;《2021年中国地球科学联合学术年会论文集》(第1期);全文 * |
| 基于自适应矢量有限元法的三维大地电磁正反演研究;秦策;《中国博士学位论文全文数据库基础科学辑》(第1期);全文 * |
| 支持向量机在物探方法预测含水层含水量中的应用;侯征;《中国优秀硕士学位论文全文数据库基础科学辑》(第3期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113484920A (en) | 2021-10-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mauriello et al. | 3D geoelectric tomography and archaeological applications1 | |
| Mauriello et al. | Principles of probability tomography for natural-source electromagnetic induction fields | |
| CN107742015A (en) | DC lasering electric method Three-dimensional Numerical Simulation Method based on any dipole dipole device | |
| CN102937721A (en) | Limited frequency tomography method for utilizing preliminary wave travel time | |
| CN110007357B (en) | Aviation TEM and aviation MT joint inversion method | |
| CN113484920B (en) | Two-dimensional structured inversion method for frequency domain electromagnetic sounding data | |
| Di Maio et al. | Quantitative interpretation of multiple self-potential anomaly sources by a global optimization approach | |
| Cella et al. | High-resolution geophysical 3D imaging for archaeology by magnetic and EM data: the case of the iron age settlement of Torre Galli, Southern Italy | |
| Papadopoulos et al. | Electrical resistivity tomography for the modelling of cultural deposits and geomophological landscapes at Neolithic sites: a case study from Southeastern Hungary | |
| Galley et al. | Geophysical inversion for 3D contact surface geometry | |
| Feng et al. | Quantitative imaging for civil engineering by joint full waveform inversion of surface-based GPR and shallow seismic reflection data | |
| CN109597136A (en) | A kind of mine total space transient electromagnetic data processing method | |
| Zhang et al. | The electrical conductivity structure of the Tarim basin in NW China as revealed by three-dimensional magnetotelluric inversion | |
| CN115292890A (en) | Site soil pollutant concentration three-dimensional space prediction method based on multi-source auxiliary data development | |
| Başokur et al. | Object-based model verification by a genetic algorithm approach: Application in archeological targets | |
| Oh et al. | Interpretation of controlled-source electromagnetic data from iron ores under rough topography | |
| Gündoğdu et al. | Characterization of the bridge pillar foundations using 3d focusing inversion of DC resistivity data | |
| Jiang et al. | Quasi-2D block inversion of large-scale surface nuclear magnetic resonance profile data using a laterally constrained model | |
| CN109655910A (en) | The two-parameter full waveform inversion method of Ground Penetrating Radar based on phasing | |
| Zhao et al. | U-net-based pseudoseismic imaging for the short-offset transient electromagnetic method | |
| CN113406707A (en) | Magnetotelluric multi-scale and multi-time-period detection method | |
| Zhou et al. | An intelligent MT data inversion method with seismic attribute enhancement | |
| Wen et al. | Seismic structure-constrained inversion of CSAMT data for detecting karst caves | |
| CN115951416A (en) | Earth electromagnetic sounding data retrieval method, apparatus, equipment and storage medium | |
| CN114488327A (en) | Horizontal magnetic field and vertical magnetic field combined measurement method in well based on ground base point |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |