포항 현장자료 적용을 위한 순환 신경망 기반 해수면 기인 다중반사파 억제 알고리즘 개발 = Development of a Recurrent Neural Network Algorithm for the Attenuation of Surface-Related Multiples in Pohang Field Data Applications

다국어 초록 (Multilingual Abstract)

In marine seismic exploration, surface-related multiples create artificial boundaries that hinder accurate geological interpretation, making their removal essential. Conventional multiple removal methods and seismic processing software vary in effectiveness based on physical properties and are sensitive to parameter settings. Thus, this study applied a stacked bidirectional long short-term memory (BiLSTM) model, known for its high performance in seismic noise suppression and signal classification, along with the fractional Fourier transform to attenuate surface-related multiples. The study also analyzed training results based on the range of rotation angles in the fractional Fourier transform (FrFT), and select the optimal angle using root mean squared error and peak signal-to-noise ratio. The optimized network was applied to field data from Pohang, and its performance was compared with seismic processing software, confirming its effectiveness and practical applicability for field data.

국문 초록 (Abstract)

해양 탄성파 탐사에서 해수면 기인 다중반사파는 실제 지층에 없는 경계면을 생성하여 지질학적 해석을 방해하므로 제거가 필수적이다. 기존의 다중반사파 제거 기법과 탄성파탐사 자료처리 소프트웨어를 적용하는 경우, 물리적 특성에 기반하여 탐사환경에 따라 적용 효과가 달라지거나 자료처리 변수에 민감한 영향을 받는다는 한계점이 있다. 본 연구에서는 잡음 억제 및 신호 분류에서 우수한 성능을 보이는 순환 신경망 기반의 stacked BiLSTM 네트워크와 분수 푸리에 변환을 적용하여 해수면 기인 다중반사파를 억제하였다. 또한, 분수 푸리에 변환의 회전 각도 범위에 따른 학습 결과를 분석하고, 평균 제곱근 오차와 최대 신호 대 잡음비를 통해 최적의 회전 각도를 설정하였다. 최적화된 네트워크를 포항 현장자료에 적용하여 탄성파탐사 자료처리소프트웨어 결과와 비교한 결과, 현장자료 적용 가능성을 검증하였다

참고문헌 (Reference)

• 1 Namias, V., "The fractional order Fourier transform and its application to quantum mechanics" 25 : 241-265, 1980
• 2 Almeida, L. B., "The fractional Fourier transform and time-frequency representations" 42 (42): 3084-3091, 1994
• 3 Verschuur, D. J., "Surface-related multiple elimination, an inversion approach" Delft University of Technology 1991
• 4 Siahkoohi, A., "Surface-related multiple elimination with deep learning" 4629-4634, 2019
• 5 Foster, D. J., "Suppression of multiple reflections using the Radon transform" 57 : 386-395, 1992
• 6 Dragoset, W. H., "Some remarks on surface multiple attenuation" 63 (63): 772-789, 1998
• 7 Stewart, P. G., "Solutions for deep water imaging" GeoHorizons 8-22, 2007
• 8 Tian, L., "Seismic spectral decomposition using short-time fractional Fourier transform spectrograms" 192 : 104400-, 2021
• 9 Yilmaz, Ö., "Seismic Data Analysis: Processing, Inversion and Interpretation of Seismic Data" Data, Soc. Expl. Geophys 2001
• 10 Akal, T., "Sea floor effects on shallow-water acoustic propagation, in bottom-interfacing ocean acoustic" Plemum 1980
• 11 Scales, J. A., "SU user’s manual" Theory of Seismic Imaging 217-258, 1995
• 12 이강훈 ; 편석준, "Reverse-time migration using surface-related multiples" 21 (21): 41-53, 2018
• 13 Nam, H. S., "Removal of seabed multiples in seismic reflection data using machine learning" 23 (23): 168-177, 2020
• 14 남호수 ; 임보성 ; 권일룡 ; 김지수, "Removal of seabed multiples in seismic reflection data using machine learning" 23 (23): 168-177, 2020
• 15 RadExPro Seismic Software LLC., "RadExPro 2024.3 User Manual"
• 16 Urick, R. J., "Principles of Underwater Sound" McGraw-Hill 1983
• 17 Peacock, K. L., "Predictive deconvolution : Theory and practice" 34 : 155-169, 1969
• 18 Robinson, E. A., "Predictive decomposition of seismic traces" 22 : 767-778, 1957
• 19 Cheong, S., "Optimal rejection of sea bottom, peg-leg and free-surface multiples for multichannel seismic data on southeastern sea" 12 : 289-298, 2009
• 20 Schlumberger, "Omega Geophysical Data Processing Software User Manual"
• 21 Lokshtanov, D., "Multiple suppression by single channel and multichannel deconvolution in the tau-p domain" SEG 1482-1485, 1995
• 22 Weglein, A. B., "Multiple attenuation : an overview of recent advances and the road ahead" 18 (18): 40-44, 1999
• 23 Hampson, D., "Inverse velocity stacking for multiple elimination" SEG 422-424, 1986
• 24 Berkhout, A. J., "Imaging of multiple reflections" 71 : SI209-SI220, 2006
• 25 Mendlovic, D., "Fractional Fourier transforms and their optical implementation : I" 10 : 1875-1881, 1993
• 26 Saxena, R., "Fractional Fourier transform : A novel tool for signal processing" 85 (85): 11-26, 2005
• 27 Sheriff, R. E., "Exploration Seismology" Cambridge Univ. Press 1995
• 28 Berkhout, A. J., "Estimation of multiple scattering by iterative inversion, Part I : Theoretical considerations" 62 : 1586-1595, 1997
• 29 Kim, Y., "Estimated source wavelet-incorporated reverse-time migration with a virtual source imaging condition" 61 : 317-333, 2013
• 30 정서제 ; 정우근 ; 신성렬 ; 김동현 ; 김재수 ; 변기훈 ; 이다운, "Development of dolphin click signal classification algorithm based on recurrent neural network for marine environment monitoring" 26 (26): 126-137, 2023
• 31 Lee, D., "Denoising sparker seismic data with Deep BiLSTM in fractional Fourier transform" 184 : 105519-, 2024
• 32 Lopez, G. A., "Closed-loop surface-related multiple elimination and its application to simultaneous data reconstruction" 80 : V189-V199, 2015
• 33 Wang, T., "Closed-loop SRME based on 3D L1-norm sparse inversion" 65 : 1145-1152, 2017
• 34 Verschuur, D. J., "Adaptive surface-related multiple elimination" 57 (57): 1166-1177, 1992