Research And Application Of Wavelet Multiscale Analysis In Gravitational Field Separation

At present,gravity exploration is still widely used to explore minerals,resources and study the distribution of the earth's internal structures.In the data processing of measured gravity anomalies,the separation of gravity anomalies needs further in-depth study and improvement,because the geological information of the earth's internal structure caused by gravity anomalies on the ground surface is complex and difficult to match.Generally speaking,we regard gravity anomaly on the ground surface as the superposition of regional anomaly and local anomaly.Wavelet transform has attracted the attention of scholars in many fields since its appearance,and many scholars in China have also introduced wavelet into gravity data processing.This paper introduces the basic principles of gravity exploration,construction techniques,gravity correction and anomaly separation methods.And the principles of Fourier transform,short-time Fourier transform,wavelet transform,mallat algorithm and wavelet multiscale analysis.For the multi-sphere model,the analytic continuation method is used to separate the gravity anomaly.the upward continuation will amplify the regional anomaly and the downward continuation will amplify the local anomaly.Separating anomalies by vertical derivation will amplify local anomalies,and secondorder derivation will have a better effect on amplifying local anomalies.Using wavelet multiscale method to separate anomalies,five kinds of wavelets are selected respectively,and the standard mean square deviation of approximation result value and regional anomaly value is taken as the error parameter to determine the best decomposition layer,and the best decomposition layer and the best wavelet are determined.After upgrading the optimal wavelet scheme,multiscale analysis is carried out,and it is found that the effect is slightly better than that of the initial wavelet.In cuboid model,lifting wavelet is used to conduct multiscale separation of gravity anomalies in the model,and the optimal decomposition layer number is determined.At the same time,the meaning of correlation coefficient is studied,and the abnormal curve of a certain section is separated,and then the correlation coefficient is calculated,which is better than the unseparated correlation coefficient.Finally,through lifting wavelet multiscale four-layer separation of gravity data in a certain work area in Jiaodong,the distribution of abnormal bodies in different depths is obtained,which improves the interpretation accuracy of gravity exploration data and further confirms the application effect wavelet multiscale analysis technology in gravitational field separation.