Regional and Residual Gravity Anomaly Separation Using Singular Spectrum based Frequency Filtering Methods: A Case Study of Shallow Subsurface Modeling from Nagpur, India

Abstract

We present here the application of a single and multichannel singular spectrum analysis (SSA)-based data filtering algorithm for regional and residual gravity field separation. Initially, the single-channel SSA method was tested on synthetic gravity profile data generated using Gravity and Magnetics Modelling System (GM-SYS), a gravity modeling software, for a sedimentary basin structure corrupted with Gaussian white noise and compared with the convnetional Fast Fourier transform (FFT)-based wave number (non-data adaptive) and discrete wavelet transform (semi-data-adaptive) filtering approaches. The comparative results show that the residual filtering of gravity data using the SSA method produces less/minimal artifacts compared to FFT-based wave number filtering and wavelet filtering methods. Following the synthetic examples, we further employed multichannel SSA-based frequency filtering on 2D gridded gravity data from Umred, Nagpur district, Maharashtra, India, to separate the residual anomaly for modelling the shallow subsurface structures. The modelled depth sections of two mutually perpendicular profiles suggest the presence of alluvium, Lameta, Kamthi and Barakar formations with varying thicknesses. The formations identified from the models show a clear match with the local geology and nearby borehole information. We suggest that the single and multichannel SSA based frequency filtering are robust approaches for separating regional and residual components from gravity anomalies.