نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشگاه صنعتی شاهرود / شاهرود / ایران
2 دانشگاه صنعتی شاهرود / شاهرو / ایران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
The measured potential field data can be considered as the result of the superposition of the anomalies from sources with various depths. Regional anomalies due to the origin of deep structures and residual anomalies due to the origin of shallow structures form the long and short parts of the total measured field wavelength, respectively. Therefore, one of the most important steps in the potential field data processing is the regional-residual anomalies separation which is used as the basis for inversion and interpretation. The process of separating regional and residual anomalies in potential field data is usually performed in the measured or frequency domain. Methods such as moving averaging, polynomial fitting, and minimum curvature are some of the well-known methods in the potential field separation in the measuring domain. Methods that perform the separation process in the frequency domain have superior performance compared to other methods, making them more common and widely used. Methods such as simple wavenumber filtering, matched filters, preferential filters, and Wiener filters are some of the common methods in the frequency domain to separate regional and residual anomalies. Various researches have shown that the rank of trajectory matrix obtained from measured potential field data depends on the depth of the anomaly source, and the rank of trajectory matrix of the deep sources are lower than that of the shallow sources. In this paper, the spectral analysis of singular values (SSA) was used to reduce the rank of the trajectory matrix obtained from gravity data in order to separate the regional and residual anomalies. Based on the theory of the SSA method, the following method was proposed to separate regional and regional anomalies in 2D gravity data. At the first step, the trajectory matrix is calculated from the Henkel matrices obtained from the measured data. Then, the obtained trajectory matrix is decomposed to eigen triples by employing the SVD and the eigenimages of it are calculated. The optimal value of rank is obtained from the elbow point of the cumulative contribution chart for eigenimages and the trajectory matrix related to regional anomaly is constructed using optimal rank. Finally, the separated regional anomaly is obtained by averaging along anti-diagonals element of the reconstructed trajectory matrix. The efficiency of the proposed method is investigated on both synthetic and real field data examples. Investigating the relationship between the depth of origin of the anomaly and the rank of the trajectory matrix calculated from the measured data showed that there is an inverse relationship between them. The obtained results of synthetic and real data showed that the technique of reducing the rank of the trajectory matrix using SSA can be used as a method of separating anomalies with different depths of origin in potential field data. Also, comparing the results of the proposed method with the results of polynomial fitting and matched filtering methods showed that the proposed method has a better performance in the separation of residual and regional anomalies and can produce better results in environments with high geological complexity.
کلیدواژهها [English]
)