Noise attenuation for multi-sensor streamer data via cooperative de-noising
Can Peng, Hongzheng Jin, Ping Wang · 2014
Summary The benefits of multi-sensor seismic data (which includes both pressure records and particle velocity or acceleration records) for de-ghosting and interpolation strongly depend on the quality of the accelerometer data. Typical raw accelerometer records are heavily contaminated by noise and are much noisier than corresponding pressure data. Therefore, jointly processing pressure data with the accelerometer data without noise removal can actually degrade the overall results. We propose a new method to attenuate noise in the accelerometer records using information from the pressure data; it is based on the acceleration data converted from the pressure data (equivalent acceleration). It entails cooperative de-noising in a high angular resolution complex wavelet transform domain between the raw accelerometer records and the converted accelerometer data.