Adaptive Seismic Denoising Based on the Synchrosqueezed-Continuous Wavelet Transform and Block-Thresholding
Abstract
The Continuous Wavelet Transform (CWT), short time window Fourier transform (STFT), and synchrosqueezed transform CWT (SSCWT) have been applied to single channel seismic data to remove noise from signal and signal from noise. Several theoretical investigations culminated in 4 publications concerning Time-Frequency Representations (TFR) of seismic data and resulted in experimental software that was delivered to interested scientists at AFTAC. In addition, a MatLab software Graphical User Interface and associated inline function, BCseis (for Block Choice Seismic Analysis) has been written for experimental and routine processing of seismic data using hard and soft block thresholding, band-pass filtering, and seismogram decomposition by manipulating a time series CWT. Studies of seismic data were performed for 2 SPE explosions recorded at Mina, NV, at the NVAR array, 12 AFTAC explosions recorded by the IRIS Community Wavefields experiment in northern Oklahoma, and a suspected small nuclear event in North Korea. Use of thresholding techniques significantly improved signal-to-noise ratios for events that can be buried by noise. However, array beamforming ability can be compromised if the resulting signal becomes uncorrelated over the aperture of the array.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2018
- Accession Number
- AD1061053
Entities
People
- Charles A. Langston
Organizations
- University of Memphis