Differentiation and Recognition of Overlapped Layers in Seismic Data Using Discrete Wavelet Transform

Abstract

Oil well jobs are highly dependent on earth layers position and their composition. So it is really important to interpret seismic data in such a good way that there are fewer problems. One major problem in seismic data interpretation is layers overlapping. Thin and close layers with similar properties may overlap cause of large signal length, thin layer thickness, and small distance between the layers. The goal was to separate overlapped layers and showing their real thickness and achieved by operating discrete wavelet transform (DWT) on the seismic data. DWT applied on 2D seismic section to overcome the problem. DWT uses high and low band-pass filters which separate signal in to high (details) and low (approximations) frequencies, and then by applying threshold makes some suspicious coefficients zero (which cause noise or overlapping). Inverse transform builds the final data in which the layers may not overlapped again. Final results provide proper locations for drilling jobs, and so determine pay zone thickness and usefulness in reservoir calculations and well completion. The results indicate the positive effect of discrete wavelet transform on seismic data on one southern Iranian field.

Practical application of a safe human-robot interaction software

“…There are many feature extraction methods according to the type and purpose of the signal. For example, the mel frequency cepstral coefficients (MFCC) method has been used with speech data (Bingol and Aydogmus, 2018), while discrete wavelet transform (DWT) has been used in the analysis of seismic data (Mohammad Panah et al , 2015). Feature extraction methods such as MFCC are based on Fourier series; however, Fourier series cannot produce the desired results for short-term sudden changes.…”