بازسازي داده هاي لرزه اي با نمونه برداري مكاني غيريكنواخت

Reconstruction of spatially nonuniform sampled seismic data

The objective of exploration seismology is to measure accurately reflected wavefields from subsurface interfaces in order to generate an image of the geological formations. More often the seismic data is non-uniformly sampled, i.e. the data is not acquired on an equidistantly spaced grid. This may result in artifacts during data processing that complicate interpretation. Sampling may be nonuniform because of human-related reasons such as: 1. Faulty equipment and positioning errors, 2. It may be due to environmental circumstances, such as cable feathering in marine acquisition induced by ocean currents or 3. Due to inaccessible areas in land acquisition (cities, rivers, and canyons.( The generation of uniformly sampled data from non-uniformly sampled data is called reconstruction and many different methods to reconstruct seismic data have been published over the years. The method used in this paper is called reconstruction of non-uniformly sampled data with least squares Fourier transform. It is based on estimating the Fourier coefficients that describe the non-uniformly sampled data, and once these coefficients have been found the signal can be reconstructed on any suitable grid via an inverse Fourier transformation. The efficiency of the method is evaluated on both real and synthetic seismic. All necessary codes were written in MATLAB environment.