论文部分内容阅读
Microseismicnoise attenuation is very important for seismic data analysis and interpretation,especiallyin the complex 3D anisotropic media.We solve 3D elastic wave equation with VTI media and applynoncausal regularized nonstationaryautoregression (NRNA) in f-x-y domain for noise attenuation.NRNA consider that the central trace can be predicted by all around this trace which can adaptively estimate seismic events of which slopes vary in 3D space from all directions in 3D seismic cube.A shaping regularization technology is added to make the nonstationaryautoregression problem realizable in mathematics with high computational efficiency.A synthetic data example demonstrates that f-x-y NRNA can be effective in suppressing random noise and improve trace-by-trace consistency.