Regional-residual anomalies and noise separation from gravity field using singular value decomposition method

Potential field data is the assembled of effects of all underground sources. Computing regional-residual anomaly is a critical step in modeling and inversion in the gravity method. Existence quantitative noise in corrected gravity data is unavoidable. In this paper, we present a novel separation method based on a Singular Value Decomposition (SVD) analysis of gravity dataset. With the SVD, a matrix of bouguer gravity data can be decomposed to a series of eigenimages. The number of required eigenimages or threshold for the reconstruction of the regional and residual (local) anomalies maps and noise distribution map from bouguer anomaly is determined based on the derived singular values by SVD. To reconstruct the data set by eigenimages may lose negligible information. We have considered which this value is equivalent with the mean of the variance of the resulted matrixes by eigenimages. The efficiency of the Singular Value Decomposition method was tested with the noisy synthetic gravity data of a hybrid model of the sphere as a local anomaly and deep-seated sloping plane as a regional anomaly. The separation results are satisfactory. The proposed method was applied on gravity field dataset of the Qom area, Iran.