Abstract

Seismic wavelet estimation is very important in the processing and interpretation of seismic data. The noise present in seismic data deviate the wavelet estimates and thereby mislead the processing and interpretation. We present Spatial Singular Spectrum Analysis (SSSA) and Multivariate Wavelet (MW) based de-noising accompanied with generalized inversion based wavelet estimation to surmount the above problems. We apply the algorithm on a three layer model synthetic data contaminated with Gaussian white noise. Our results suggest that the usage of SSSA or combined usage of SSSA and MW is robust scheme for de-noising. The seismic wavelet and reflectivity series estimated from noisy synthetic data using the proposed methods are matching well with the original seismic wavelet and reflectivity series data. Thus, we conclude that the SSSA or combined SSSA and MW denoising of seismic data is essential as a pre filtering for precise wavelet estimation via generalized inversion method.