基于逐减随机震源采样法的频率域二维黏滞声波方程全波形反演

doi:10.13278/j.cnki.jjuese.201606305

摘要/Abstract

摘要： 全波形反演方法利用叠前地震波场的运动学和动力学信息重建地下速度结构，具有揭示复杂地质背景下构造与岩性细节信息的潜力。然而，巨大的计算量是阻碍其发展的一个瓶颈问题。为此，研究者们提出了震源编码技术来减少计算量，但是此方法在模型更新过程中会引进随机串扰噪声，降低反演结果准确性。所以，在保证计算精度的情况下，本文提出了采用逐减随机震源采样的方法来高效计算全波形反演问题。笔者将此方法应用于频率域二维黏滞声波波动方程全波形反演，开始了在频率域进行随机震源采样类方法的研究，计算过程中共使用了依次增大的8个频率段；并应用Overthrust模型来验证此类随机震源采样法的正确性。实验结果表明：基于逐减随机震源采样法的反演结果与实际Overthrust模型的拟合误差为0.065 65，而应用基于全部震源的全波形反演方法得到的反演结果与实际Overthrust模型的拟合误差为0.064 64，两者差别不大；但计算用时由740 min减少到291.2 min，即计算效率提高了2.54倍。为了更好地确定方法的有效性，将其应用于Marmousi模型进行试算。模型试算结果表明：基于逐减随机震源和基于全部震源得到的反演结果与实际Marmousi模型的拟合误差分别为0.080 12和0.078 97，相差不大；但计算用时由1 218.9 min减少到274.4 min，计算效率提高了4.44倍。综上，在保证反演精度的情况下，基于逐减随机震源采样法的频率域全波形反演方法大大减少了计算量，具有不可替代的计算优势，并且没有引进随机串扰噪声。

关键词: 逐减随机震源采样法, 频率域, 二维黏滞声波方程, 全波形反演, 无记忆拟牛顿算法, L-BFGS

Abstract: The full waveform inversion (FWI) method utilizes kinematic and dynamic information of pre-stack seismic data to rebuild underground velocity structure. However, the large amount of calculation is a bottleneck problem for its development. Therefore, Researchers proposed the techniques of the source-encoding in order to reduce calculation amount, but this method would introduce random crosstalk in model updates. The paper uses decreasing random shot subsampling method to invert full waveform effectively in the case of guaranteeing the calculation accuracy. The paper applied the method into frequency-domain full waveform inversion of 2D viscous acoustic wave equation, and research the type of random shot sampling method in the frequency domain. We totally use eight gradually increased frequencies in FWI process, and apply overthrust model to demonstrate the validity of such random shot sampling method. Through the experiment,we can see that the fitting error between inversion results based on decreasing random shot subsampling method and the actual overthrust model is 0.065 65, and the fitting error between inversion results based on all the shots and the actual overthrust model is 0.064 64. Namely the inversion results of the two methods have little difference. However, the calculated time reduces from 740 min to 291.2 min. Namely the computational efficiency increased 2.54 times. In order to better determine the validity of the method, we apply it to Marmousi model. The fitting error between inversion results based on decreasing random shot subsampling method and the actual Marmousi model is 0.080 12, and the fitting error between inversion results based on all the shots and the actual Marmousi model is 0.078 97. Namely the inversion results of the two methods have little difference. However, the calculated time reduces from 1 218.9 min to 274.4 min. Namely the computational efficiency increased 4.44 times. In conclusion, under the condition of the inversion accuracy, the full waveform inversion based on the decreasing random shot sampling method in frequency domain greatly reduces the amount of calculation. Namely it has irreplaceable computing advantage. Therefore there is no introduction of random crosstalk noise.

Key words: decreasing random shot subsampling method, frequency-domain, 2D viscous acoustic wave equation, full waveform inversion, memory-less quasi-newton algorithm, L-BFGS

中图分类号:

P631.4

冯晅, 鲁晓满, 刘财, 周超, 金泽龙, 张明贺. 基于逐减随机震源采样法的频率域二维黏滞声波方程全波形反演[J]. 吉林大学学报(地球科学版), 2016, 46(6): 1865-1873.

Feng Xuan, Lu Xiaoman, Liu Cai, Zhou Chao, Jin Zelong, Zhang Minghe. Frequency-Domain Full Waveform Inversion of 2D Viscous Acoustic Wave Equation Using Decreasing Random Shot Subsampling Method[J]. Journal of Jilin University(Earth Science Edition), 2016, 46(6): 1865-1873.

参考文献

[1] Herrmann F J. Randomized Sampling and Sparsity: Getting More Information from Fewer Samples[J]. Geophysics,2010,75(6):WB173-WB187.
[2] Milton A, Trickett S, Burroughs L. Reducing Ac-quisition Costs with Random Sampling and Multidimensional Interpolation[J]//SEG Technical Program Expanded Abstracts, 2011, 31(1): 52-56.
[3] 杨庆节,刘财,耿美霞,等.交错网格任意阶导数有限差分格式及差分系数推导[J].吉林大学学报(地球科学版), 2014, 44(1): 375-385. Yang Qingjie, Liu Cai, Geng Meixia, et al. Staggered Grid Finite Difference Scheme and Coefficients Deduction of Any Number of Derivatives[J]. Journal of Jilin University (Earth Science Edition), 2014, 44(1): 375-385.
[4] Tarantola A. Inversion of Seismic-Reflection Data in the Acoustic Approximation[J]. Geophysics, 1984, 49(8): 1259-1266.
[5] Sirgue L, Pratt R G. Efficient Waveform Inversion and Imaging: A Strategy for Selecting Temporal Frequencies[J]. Geophysics, 2004, 69(1): 231-248.
[6] 邓武兵,韩立国,李翔,等.L-BFGS和SPGL1在全波形反演中的联合应用[C]//中国地球物理学会第二十七届年会. 长沙: 中国地球物理学会, 2011: 714. Deng Wubing, Han Liguo, Li Xiang, et al. Appply the Combination of L-BFGS and SPGL1 to Full-Waveform Inversion[C]//The 27th Annual Meeting of Chinese Geophysical Society. Changsha: Chinese Geophysical Society, 2011: 714.
[7] 苏超,周辉,林鹤.时间域声波全波形反演及GPU加速[C]//中国地球物理学会第二十八届年会. 北京:中国地球物理学会,2012: 486. Su Chao, Zhou Hui, Lin He.Acoustic Full Waveform Inversion in Time-Domain and Its Acceleration by GPU[C]//The 28th Annual Meeting of Chinese Geophysical Society. Beijing: Chinese Geophysical Society, 2012: 486.
[8] 龙桂华,李小凡,张美根,等.频率域黏弹性声波透射波形速度反演[J].地震学报, 2009, 31(1): 32-41. Long Guihua, Li Xiaofan, Zhang Meigen, et al. Vis-coacoustic Transmission Waveform Inversion for Velocity Structure in Space-Frequency Domain[J].Acta Seismologica Sinica, 2009, 31(1): 32-41.
[9] 刘璐,刘洪,张衡,等.基于修正拟牛顿公式的全波形反演[J].地球物理学报, 2013, 56(7): 2447-2451. Liu Lu, Liu Hong, Zhang Heng, et al. Full Waveform Inversion Based on Modified Quasi-Newton Equation[J].Chinese Journal of Geophysics,2013, 56(7): 2447-2451.
[10] 成景旺,顾汉明,刘春成,等.频率域反射波全波形速度反演[J].地球科学:中国地质大学学报, 2013, 38(2): 391-397. Cheng Jingwang, Gu Hanming, Liu Chuncheng, et al. Full Waveform Inversion for Velocity Structure from Reflected Wave Seismic Data in the Frequency-Domain[J]. Earth Science: Journal of China University of Geoscience, 2013, 38(2): 391-397.
[11] Li X, Aravkin Y A, van Leeuwen T. Fast Rando-mized Full-Waveform Inversion with Compressive Sensing[J]. Geophysics, 2012, 77(3): A13-A17.
[12] Morton S A, Ober C C. Faster Shot-Record Depth Migration Using Phase Encoding[C]//1998 SEG Annual Meeting. New Orleans: Society of Exploration Geophysicists, 1998: 1131-1134.
[13] Krebs J, Anderson J, Hinkley D, et al. Fast Full-Wavefield Seismic Inversion Using Encoded Sources[J]. Geophysics, 2009,74(6): WCC177-WCC188.
[14] Ben-Hadj-Ali, Operto H S, Virieux J. An Efficient Frequency Domain Full Waveform Inversion Method Using Simultaneous Encoded Sources[J]. Geophysics, 2011, 76(4): R109-R124.
[15] Gao F, Atle A, Williamson P. Full Waveform In-version Using Deterministic Source Encoding[C]//2010 SEG Annual Meeting. Denver: Society of Exploration Geophysicists, 2010: 1013-1017.
[16] Schuster G T, Wang X, Huang Y, et al. Theory of Multisource Crosstalk Reduction by Phase-Encoded Statics[J]. Geophysical Journal International, 2011, 184(3): 1289-1303.
[17] Ben-Hadj-Ali H, Operto S, Virieux J. An Efficient Frequency Domain Full Waveform Inversion Method Using Simultaneous Encoded Sources[J]. Geophysics, 2011, 76(4): R109-R124.
[18] Choi Y, Alkhalifah T. Source-Independent Time-Do-main Waveform Inversion Using Convolved Wavefields: Application to the Encoded Multisource Waveform Inversion[J]. Geophysics, 2011, 76(5): R125-R134.
[19] Choi Y, Alkhalifah T. Application of Multi-Source Waveform Inversion to Marine Streamer Data Using the Global Correlation Norm[J]. Geophysical Prospecting, 2012, 60(4): 748-758.
[20] Moldoveanu N. Random Sampling: A New Strategy for Marine Acquisition[J]. SEG Expanded Abstracts, 2010(1): 51-55.
[21] Ha W, Shin C. Efficient Full Waveform Inversion Using a Cyclic Shot Sampling Method[C]//2012 SEG Annual Meeting. Las Vegas: Society of Exploration Geophysicists, 2012: 1-5.
[22] Wang C, Yingst D, Brittan J, et al. Fast Multi-Parameter Anisotropic Full Waveform Inversion with Irregular Shot Sampling[C]//2014 SEG Annual Meeting. Denver: Society of Exploration Geophysicists, 2014: 1147-1151.
[23] Bilinskis I. Digital Alias-Free Signal Processing[M]. Chichester: John Wiley & Sons, Ltd, 2007.
[24] Pratt R G, Worthington M H. Inverse Theory App-lied to Multi-Source Cross-Hole Tomography: Part 1:Acoustic Wave-Equation Method[J]. Geophysical Prospecting, 1990, 38(3): 287-310.
[25] Liu C, Gao F X, Feng X, et al. Memory-Less Quasi-Newton(MLQN) Method for 2D Acoustic Full Waveform Inversion[J]. Exploration Geophysics, 2015, 46(2): 168-177.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed