Modeling the seismic wave equation using a staggered grid finite-difference method optimized with a genetic algorithm

Abstract

Simulation of seismic waves is a critical component in the imaging of subsurface structures using actual data, where numerical dispersion remains a challenging task. The finite-difference (FD) approach is popular for solving wave equations because it is simple to implement and requires less memory and computing time due to recursion. However, the staggered grid finite-difference (SGFD) methods have gained popularity due to their improved accuracy and stability. In this study, we introduce an optimization approach using a genetic algorithm (GA) to minimize numerical dispersion. The SGFD coefficients were optimized to reduce numerical errors and improve the accuracy of seismic wave simulations, considering both spatial and temporal domains. Numerical simulations applied to both homogeneous and heterogeneous velocity models demonstrate that the GA-optimized SGFD schemes achieve substantial reductions in dispersion, even with lower-order approximations, when compared to other methods. An important advantage of the proposed method is that it maintains high accuracy while using lower-order approximations, which significantly reduces computational costs. For example, the optimization of 12th-order FD coefficients took approximately 20 s on a standard computer with 64 GB RAM. The findings demonstrate the efficiency of the proposed approach in improving the accuracy and stability of seismic wave simulations, providing a reliable solution for high-resolution seismic imaging.

Keywords

Numerical dispersion / Seismic wavefield / Staggered grid finite-difference method / Genetic algorithm / Modeling / Optimization

Funding

This research was supported by the Ministry of Earth Science (MoES), New  Delhi (grant number: MoES/OSMART/EFC2021 - OM No. 01(01)/PFC-I/2022).

References

[1] Kelly KR, Ward RW, Treitel S, Alford R. Synthetic seismograms: A finite -difference approach. Geophysics. 1976; 41(1):2-27. doi: 10.1190/1.1440605.

[2] Zhi‐Yang W, Hong L, Xiang‐De T, Yang W. Optimized finite-difference operator based on chebyshev auto-convolution combined window function. Chin J Geophys. 2015; 58(2):192-206. doi: 10.1002/cjg2.20166

[3] Jing H, Yang G, Wang J. An optimized time-space-domain finite difference method with piecewise constant interpolation coefficients for scalar wave propagation. J Geophys Eng. 2019L; 16(2):309-324. doi: 10.1093/jge/gxz008

[4] Vanga M, Barman D, Ojha M. An optimized finite difference method to minimize numerical dispersion of acoustic wave propagation using a genetic algorithm. Geophysics. 2022; 87(3):T265-T279. doi: 10.1190/geo2021-0382.1

[5] Patera AT. A spectral element method for fluid dynamics: Laminar flow in a channel expansion. J Comput Phys. 1984; 54(3):468-488. doi: 10.1016/0021-9991(84)90128-1

[6] Komatitsch D, Vilotte JP. The spectral element method: An efficient tool to simulate the seismic response of 2D and 3D geological structures. Bull Seismol Soc Am. 1998; 88(2):368-392. doi: 10.1785/bssa0880020368

[7] Chaljub E, Komatitsch D, Vilotte JP, Capdeville Y, Valette B, Festa G. Spectral-element analysis in seismology. Adv Geophys. 2007; 48:365-419. doi: 10.1016/s0065-2687(06)48007-9

[8] Cristini P, Komatitsch D. Some illustrative examples of the use of a spectral-element method in ocean acoustics. J Acoust Soc Am. 2012; 131(3):EL229-EL235. doi: 10.1121/1.3682459

[9] Bottero A, Cristini P, Komatitsch D, Asch M. An axisymmetric time-domain spectral-element method for full-wave simulations: Application to ocean acoustics. J Acoust Soc Am. 2016; 140(5):3520-3530. doi: 10.1121/1.4965964

[10] Dormy E, Tarantola A. Numerical simulation of elastic wave propagation using a finite volume method. J Geophys Res Atmos. 1995; 100(B2):2123-2133. doi: 10.1029/94jb02648

[11] Eymard R, Gallouët T, Herbin R. Finite volume methods. In: Handbook of Numerical Analysis. Amsterdam: Elsevier; 2000. p. 713-1018. doi: 10.1016/s1570-8659(00)07005-8

[12] Komatitsch D, Erlebacher G, Göddeke D, Michéa D. High-order finite-element seismic wave propagation modeling with MPI on a large GPU cluster. J Comput Phys. 2010; 229(20):7692-7714. doi: 10.1016/j.jcp.2010.06.024

[13] Igel H, Riollet B, Mora P. Accuracy of Staggered 3‐D finite‐ Difference Grids for Anisotropic Wave Propagation. Society of Exploration Geophysicists. 1992. p. 1244-1246. doi: 10.1190/1.1821960.

[14] Finkelstein B, Kastner R. Finite difference time domain dispersion reduction schemes. J Comput Phys. 2006; 221(1):422-438. doi: 10.1016/j.jcp.2006.06.016

[15] Liu Y, Sen MK. Scalar wave equation modeling with time-space domain dispersion-relation-based staggered-grid finite-difference schemes. Bull Seismol Soc Am. 2011; 101(1):141-159. doi: 10.1785/0120100041

[16] Chu C, Stoffa PL. Determination of finite-difference weights using scaled binomial windows. Geophysics. 2012; 77(3):W17-W26. doi: 10.1190/geo2011-0336.1

[17] Tan S, Huang L. An efficient finite-difference method with high-order accuracy in both time and space domains for modelling scalar-wave propagation. Geophys J Int. 2014; 197(2):1250-1267. doi: 10.1093/gji/ggu077

[18] Tam CK, Webb JC. Dispersion-Relation-Preserving finite difference schemes for computational acoustics. J Comput Phys. 1993; 107(2):262-281. doi: 10.1006/jcph.1993.1142

[19] Geller RJ, Takeuchi N. Optimally accurate second-order time-domain finite difference scheme for the elastic equation of motion: One-dimensional case. Geophys J Int. 1998; 135(1):48-62. doi: 10.1046/j.1365-246x.1998.00596.x

[20] Zhang JH, Yao ZX. Optimized finite-difference operator for broadband seismic wave modeling. Geophysics. 2012; 78(1):A13-A18. doi: 10.1190/geo2012-0277.1

[21] Zhang JH, Yao ZX. Optimized explicit finite-difference schemes for spatial derivatives using maximum norm. J Comput Phys. 2013; 250:511-526. doi: 10.1016/j.jcp.2013.04.029

[22] Liu Y. Globally optimal finite-difference schemes based on least squares. Geophysics. 2013; 78(4):T113-T132. doi: 10.1190/geo2012-0480.1

[23] Liu Y. Optimal staggered-grid finite-difference schemes based on least-squares for wave equation modelling. Geophys J Int. 2014; 197(2):1033-1047. doi: 10.1093/gji/ggu032

[24] Yang L, Yan H, Liu H. Optimal implicit staggered‐grid finite‐difference schemes based on the sampling approximation method for seismic modelling. Geophys Prospect. 2015; 64(3):595-610. doi: 10.1111/1365-2478.12325

[25] Miao Z, Zhang J. Simplified implicit finite-difference method of spatial derivative using explicit schemes with optimized constant coefficients based on L 1 norm. Geophysics. 2023; 89(2):T47-T59. doi: 10.1190/geo2023-0246.1

[26] Peng W, Huang J, Shen Y. Reducing the low-wavenumber dispersion error by building the Lagrange dual problem with a powerful local restriction. J Geophys Eng. 2023; 20(4):798-815. doi: 10.1093/jge/gxad047

[27] Brekhovskikh L.M. Waves in Layered Media. Vol. 42. New York:Academic Press; 1960. p. 129.

[28] Manikas TW, Cain JT. Genetic Algorithms vs. Simulated Annealing: A Comparison of Approaches for Solving the Circuit Partitioning Problem; 1996. Available from: https://digitalrepository.smu.edu/engineering_compsci_research/1 Last accessed on 2025 Apr 13].

[29] Santos EPD, Xavier CR, Goldfeld P, Dickstein F, Santos RWD. Comparing genetic algorithms and Newton-Like methods for the solution of the history matching problem. In: Lecture Notes in Computer Science. Berlin: Springer; 2009. p. 377-386. doi: 10.1007/978-3-642-01970-8_37

[30] Smith Julius O. Physical Audio Signal Processing. California: Stan Ford University: 2010. Available from: https://ci.nii.ac.jp/ncid/bb07079768?l=en Last accessed on 2025 Apr 13].

[31] Billette FJ, Brandsberg-Dahl S. The 2004 BP Velocity Benchmark. In: 67th EAGE Conference and Exhibition. Bunnik: European Association of Geoscientists and Engineers; 2005.