The Research And Application On Several Processing Methods To The Potential-field Data

Geophysics exploration plays an impotent role in solving the geological issues, what’s more, the development of the potential data processing forces on the establishment of more stabilized processing method and more effective interpretation routine which could satisfy the fast progress of the earth science. This paper forces on several potential-field processing methods, mainly including: potential field edge detection method, Potential field rapid image method and Shearlet transform application, aiming at making some improvements on the stability and accuracy to the calculation. Furthermore, it adopts some suitable methods to combine with the effective interpretation routine, which could be the foundation to solve the geological issues effectively.Potential field edge detection technology is set as the irreplaceable position for its advantage in the interpretation to the structure’s feature and distribution. At present, the edge detection can be classified based on the calculation way mainly into: gradient calculation and mathematical statistics calculation. The former owns the rapid development and general application, however, the errors from the deviation calculation and fake denotation are the shortages; another method can avoid the deviation calculation and improve the SNR(Sign to Noise Rate), however, the distortion or migration may appear in the inverse-result. This paper proposes a new edge detection method — Correlation coefficients of multidirectional standard deviations method which be given an elaboration on the basic principle and the choice to the relative parameters here. Based on a system tests on the synthetical models and the comparison with some other edge detection methods, it infers that this method is anti-noisy and owns the simple calculation without the deviation, and the inversion-result characters as high reliability and resolution which can retain the details effectively. Maybe this method can be a reliable choice to the application of the real data.Potential field rapid image method gets the general development in recent years, this technology owns a exciting application—prospect with fast calculation, no pre-process and the stabile result. Here, it gives a detail analysis to large amount experiments with DEXP method(Depth from Extreme Points Method method) and the automatic DEXP method: the former is short in the structural index(SI) whose assignment is subjective with no clear rule; the latter solves the SI issue, however, the “Trapezium line” disturbs the effect of the inversion, and the threshold causes new problem, including subjective definition and no equitability. Therefore, this paper proposes Constant Adding Method to make a progress. The improved DEXP method shows satisfactory results in the experiments, what’s more, the paper gives some discussions on the choice of derivation, and it infers that the suitable choic to derviation is beneficial to balance the various signals from different geological bodies. Further, it makes a comparison among the improved DEXP method, Migration method and the Correlation Imaging Method with the synthetic experiments, in order to get more discussions on the stability on depth inversion, resolution on adjacent models inversion and the fault inversion.Shearlet Transform belongs to the multi-scale transform with characterize of simpler mathematical algorithm, better directional sensitivity and sparseness, it also can represent the anisotropy. With the chacterize of Shearlet transform and potential field data, this paper tries to apply the transform on the potential field data processing, including: Data denoising and Edge detection. It describes the details of the denoising based on Shearlet transform, and the models’ experiments shows its feasibility, with the comparison with some other methods in visual and quantification, the denoising method based on Shearlet transform can be taken as a effective and satisfactory way to denoise the potential field data. What’s more, this paper proposes an edge detection method based on the Shearlet transform—Coefficients Weighted Edge detection method which based on the Shearlet transform and mathematical statistics calculation. With the syntactical models’ experiments and comparison with some other edge detection methods, the method shows visible advantages in the state that high frequency noise and low frequency regional signal are both existing, which reflects the good equitability in the edge detection inversion. It infers that the method is valuable to expand to the real data processing.Wudalianchi volcanic cluster is a typical characterize in interior with single genesis, the geological condition is complex here, so that the research on the relative geological issues is so meaningful. Based on the research above about the processing methods to the potential field data, it adopts the suitable methods to process the real gravity anomaly and magnetic anomaly, one hand, the processing supplies the proof to the methods’ availability; on the other hand, it may enrich the geophysics information about this area. In this part, it combines the information subtracted by the suitable processing methods and the relative geological information, which realizes the analysis on the characteristic of gravity and magnetic field, the residual field and the vertical second derivative results. What’s more, it gets the initiative inference on the structural distribution and lithology distribution, which may be helpful to the future solution to the relative geological issues.