Geophysical Potential Field Anomaly Separation Method With Optimal Mother Wavelet and Spatial Locating Multiresolution Analysis (MRA)

Abstract

Gravity or magnetic field, say geophysical potential field, is the superposition of gravity or magnetic effects of all geological bodies of different depths, scales, and forms. The signal, say anomaly, caused by the target geological body must be separated from the measured potential field before used for inversion and interpretation. The classical separation methods based on multiresolution analysis (MRA) have two problems. One is how to choose the optimal mother wavelet to separate the anomalies. Another is how to separate the anomalies better when the spectrum of different geological bodies aliased each other in scale. For the first problem, we propose a quantitative evaluating indicator, sparse index (SI), to help us choose the optimal mother wavelet for the particular separation task. For the second problem, we separate the residual anomalies based on MRA with spatial locating, using the wavelet coefficients within particular spatial and scale arrange to reconstruct the regional and residual anomaly, respectively. The results of three experiments, including separating the magnetic anomalies of a 2D and a 3D geology model, show that our new approach combining optimal mother wavelet and spatial locating MRA separation is superior to the classical wavelet transform method and can improve the accuracy of separation result obviously.