| The main purpose of gravity and magnetic exploration is to study the density and magnetic characteristics of subsurface geological bodies by measuring their gravity and magnetic fields.Through corresponding interpretations,this method can help achieve the goals of studying deep tectonics,regional tectonics,as well as exploring resources and energy.Gravity and magnetic exploration offers numerous advantages,including relatively economical and practical data acquisition,large exploration depth,and rapid acquisition of surface observation information.With the continuous development of gravity and magnetic data observation via air,satellite,ground,and well,the applications of gravity and magnetic data will continue to expand in the future.As a result,gravity exploration and magnetic exploration remain the two most fundamental and commonly used means of geophysical exploration.The interpretation of gravity and magnetic data is primarily divided into three parts:processing,conversion,and inversion of gravity and magnetic data.The purpose of processing and conversion is to obtain the anomaly type of the inverse geological body and eliminate interference to highlight target anomaly information.Derivative calculations of gravity and magnetic data often correspond to the boundaries of geological tectonic units or the location of faults.Additionally,extension processing can highlight deep tectonic features through upward extension and regional tectonic features through downward extension.Calculating the polarization of magnetic anomalies is another example,which eliminates the influence of oblique magnetization of the geomagnetic field to obtain a more prominent correlation between the magnetic body and magnetic field.However,the interpretation of gravity and magnetic anomalies requires further enhancement and closer quantification.Using physical inversion tools for gravity and magnetic data is essential to clarify the deep characteristics,physical distribution,and size of geological bodies.Three-dimensional physical inversion of gravity and magnetic data is an important aspect of the quantitative interpretation of these data.This process not only determines the geometric position of anomalies but also quantitatively calculates physical parameter shape,volume,size,spatial distribution of anomalies,and other information.This information provides a crucial basis for subsequent geological interpretation work.In the early days,gravity and magnetic inversions were limited by computer computing power,mainly focusing on morphological inversions.However,with the improved computer computing power,three-dimensional physical inversion for gravity and magnetic data has gradually become the mainstream direction.The development of 3D physical inversion methods for gravity and magnetic data aims to address two problems:the improvement of vertical and lateral resolution of 3D physical inversion on target geological bodies,and the reduction of multi-resolution in3D physical inversion of gravity and magnetic data.This paper builds on previous research to further investigate these two problems.The main results are summarized below.（1）Study of gravity and magnetic law draping recognition capability for a given observation accuracy.In this paper,we take the sphere（point source model）and the infinitely extended horizontal cylinder（line source model）as examples from the theory of gravity and magnetic field orthorectification,and consider the relationship between the magnitude of gravity and magnetic force and the observation accuracy under the given observation accuracy to study the vertical recognition ability,so as to eliminate the influence of the anomalous background field,improve the reliability of the research results,and study the vertical recognition ability of gravity and magnetic force on the isolated anomaly.It also compares and analyzes the relationship between the strength of gravity and magnetism on the vertical recognition ability of isolated anomalies,so that the appropriate measurement method can be selected in advance,which is a certain guidance for the early exploration.（2）Study on the lateral discrimination ability of gravity and magnetic method for a given observation accuracy.Taking the sphere（point source model）and infinitely extended horizontal cylinder（line source model）as examples,the theoretical lateral resolution of gravity and magnetic anomalies and their order derivatives（presented as theoretical lateral resolution coefficients）as well as the actual lateral resolution under the given observation accuracy are studied from the gravity and magnetic orthorectification theory,which can provide relevant guidance for the construction design in the actual exploration.（3）Research on gravity and magnetic 3D property inversion methods based on unstructured grid segmentation of vertical recognition and lateral resolution.By using the research results of gravity and magnetic vertical recognition capability and lateral resolution capability,we can provide theoretical guidance for reasonable subsurface space dissection,and reduce the number of grid dissections while taking into account the change of grid dissection sparsity,so that the subsurface space can be dissected effectively and in accordance with the basis,and the deep and shallow dissection units can have equivalent contributions.This method can be applied to the gravity and magnetic 3D property inversion to reduce the"skin effect"caused by the rapid decay of the kernel function with depth change.Without adding any depth weighting function to the model term,the inversion results of unstructured grid segmentation can resist the effect of the decay of the kernel function with depth,which effectively mitigates the"skin effect"and improves the computational efficiency of the gravity and magnetic 3D inversions.（4）Research on gravity and magnetic 3D physical properties inversion method based on threshold pruning technique.An improved model weighting function is introduced,which can overcome the"skin effect"and suppress the"tailing effect"to a certain extent,based on which the gravity and magnetic 3D inversion method based on the threshold pruning technique is proposed.By setting the threshold size,the first inversion result less than the threshold value is assigned to 0,and only the inversion result greater than or equal to the threshold value is retained,and the trimmed property distribution is forwarded to obtain the fitted anomaly,and the observed anomaly is subtracted from the fitted anomaly to obtain the remaining anomaly.Until the number of iterations reaches the preset maximum number or the remaining anomaly is less than or equal to the observed accuracy,the iteration is stopped and the final physical distribution is output.（5）Study on the sparse inversion method of gravity and magnetic properties based on L0-norm.The regularized sparse inversion based on the L0 parametric constraint can usually obtain the model solution with sharp boundaries,but the inversion process is easy to cause the"overfocus"status,and there are various methods to select the model weighting function to match the focusing function.In this paper,we design three typical synthetic models（synthetic models with different burial depths,synthetic models with inclined slab,and synthetic models with vertical superposition）,and compare them with no model weighting constraint,model term constraint with depth weighting function introduced,sensitivity matrix weighting constraint introduced,unstructured grid dissection inversion,focusing constraint with depth weighting and minimum support function introduced,and focusing constraint with sensitivity matrix weighting and minimum The inversion results of the focusing constraint of the support function,through the study shows that the gravity and magnetic method of three-dimensional physical inversion method using the sensitivity model weighting matrix combined with the minimum support function has a good application effect.The paper begins by addressing the inherent resolution limitations of gravity and magnetic data（vertical recognition ability and lateral resolution ability）.Building on this foundation,it then examines the problem of unstructured grid profiling to mitigate the skinning effect in inversion.Additionally,the paper explores new method techniques（threshold pruning technique,L0-norm sparse optimization）aimed at enhancing the inversion resolution and reducing the multiresolution of inversion solutions.Through model testing and practical data processing,the paper confirms the accuracy and efficacy of the proposed method in improving the low resolution and serious multiresolution issues inherent in 3D physical inversion of gravity and magnetic data. |
