zbMATH — the first resource for mathematics

A computational method for full waveform inversion of crosswell seismic data using automatic differentiation. (English) Zbl 1344.86002
Summary: Full waveform inversion (FWI) is a model-based data-fitting technique that has been widely used to estimate model parameters in Geophysics. In this work, we propose an efficient computational approach to solve the FWI of crosswell seismic data. The FWI problem is mathematically formulated as a partial differential equation (PDE)-constrained optimization problem, which is numerically solved using a gradient-based optimization method. The efficiency and accuracy of FWI are mainly determined by the three main components: forward modeling, gradient calculation and model update which usually involves the gradient-based optimization algorithm. Given the large number of iterations needed by FWI, an accurate gradient is critical for the success of FWI, as it will not only speed up the convergence but also increase the accuracy of the solution. However computing the gradient still remains a challenging task even after the adjoint PDE has been derived. Automatic differentiation (AD) tools have been proved very effective in a variety of application areas including Geoscience. In this work we investigated the feasibility of integrating TAPENADE, a powerful AD tool into FWI, so that the FWI workflow is simplified to allow us to focus on the forward modeling and the model updating. In this paper we choose the limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) method due to its robustness and fast convergence. Numerical experiments have been conducted to demonstrate the effectiveness, efficiency and robustness of the new computational approach for FWI.

86-08 Computational methods for problems pertaining to geophysics
86A15 Seismology (including tsunami modeling), earthquakes
Full Text: DOI
[1] Fichtner, A., Full seismic waveform modeling and inversion, (1995), Springer Heidelberg
[2] Lailly, P., The seismic inversion problem as a sequence of before stack migration, (Conference on Inverse Scattering, (1983), SIAM), 206-220
[3] Tarantola, A., Inversion of seismic-reflection data in the acoustic approximation, Geophysics, 49, 1259-1266, (1984)
[4] Tarantola, A., Linearized inversion of seismic-reflection data, Geophys. Prospect., 32, 998-1015, (1984)
[5] Tarantola, A., Inverse problem theory and methods for model parameter estimation, (2005), SIAM Philadelphia · Zbl 1074.65013
[6] Virieux, J.; Operto, S., An overview of full-waveform inversion in exploration geophysics, Geophysics, 74, (2009), WCC1-WCC26
[7] Chavent, G., Identification of function parameters in partial differential equations, (Identification of Parameter Distributed Systems, (1974), American Society of Mechanical Engineers), 31-48
[8] Plessix, R., A review of the adjoint-state method for computing the gradient of a functional with geophysical applications, Geophys. J. Int., 167, 495-503, (2006)
[9] Alexe, M.; Sandu, A., On the discrete adjoints of variable step time integrators, J. Comput. Appl. Math., 233, 1005-1020, (2009) · Zbl 1177.65098
[10] Liu, Q.; Tromp, J., Finite-frequency kernels based on adjoint methods, Bull. Seismol. Soc. Am., 96, 2383-2397, (2006)
[11] Alexe, M.; Sandu, A., Forward and adjoint sensitivity analysis with continuous explicit Runge-Kutta schemes, Appl. Math. Comput., 208, 2, 328-346, (2009) · Zbl 1159.65071
[12] Gou, T. Y.; Sandu, A., Continuous versus discrete advection adjoints in chemical data assimilation with CMAQ, Atmos. Environ., 45, 28, 4868-4881, (2011)
[13] Mader, C. A.; Martins, R. A; Alonso, J. J.; Der Weide, E. V., Adjoint: an approach for the rapid development of discrete adjoint solvers, AIAA J., 46, 4, 863-873, (2008)
[14] Müller, J.-D.; Cusdin, P., On the performance of discrete adjoint CFD codes using automatic differentiation, Int. J. Numer. Methods Fluids, 47, 89, 939-945, (2005) · Zbl 1134.76431
[15] Sandu, A.; Miehe, P., Forward, tangent linear, and adjoint Runge Kutta methods in KPP-2.2 for efficient chemical kinetic simulations, Int. J. Comput. Math., 87, 11, 2458-2479, (2010) · Zbl 1202.65090
[16] Sandu, A.; Chai, T. F., Chemical data assimilation - an overview, Atmosphere, 2, 3, 426-463, (2011)
[17] Thomas, J. P.; Hall, K.; Dowell, E., Discrete adjoint approach for modeling unsteady aerodynamic design sensitivities, AIAA J., 43, 9, 1931-1936, (2005)
[18] Bischof, C. H.; Bücker, H. M.; an Mey, D., A case study of computational differentiation applied to neutron scattering, (Corliss, G.; Faure, C.; Griewank, A.; Hascot, L.; Naumann, U., Automatic Differentiation of Algorithms: From Simulation to Optimization, Computer and Information Science, (2002), Springer New York, NY), 69-74, (chapter 6)
[19] Cioaca, A.; Alexe, M.; Sandu, A., Second-order adjoints for solving PDE-constrained optimization problems, Optim. Methods Softw., 27, 4-5, 625-653, (2012) · Zbl 1260.49060
[20] Singh, K.; Sandu, A., Variational chemical data assimilation with approximate adjoints, Comput. Geosci., 40, 10-18, (2012)
[21] Talagrand, O., The use of adjoint equations in numerical modelling of the atmospheric circulation, Autom. Differentiation Algorithms Theory Implementation Appl., 169-180, (1991) · Zbl 0761.76077
[22] Carmichael, G. R.; Sandu, A.; Potra, F. A., Sensitivity analysis for atmospheric chemistry models via automatic differentiations, Atmos. Environ., 31, 3, 475-489, (1997)
[23] Hwang, D.; Byun, D. W.; Odman, M. T., An automatic differentiation technique for sensitivity analysis of numerical advection schemes in air quality models, Atmos. Environ., 31, 6, 879-888, (1997)
[24] Kaminski, T.; Heimann, M.; Giering, R., A coarse grid three-dimensional inverse model of the atmospheric transport: 1. adjoint model and Jacobian matrix, J. Geophys. Res., 104, 18, (1999), 535-18,553
[25] Kaminski, T.; Heimann, M., Technical comment: inverse modeling of atmospheric carbon dioxide fluxes, Science, 294, 259, (2001)
[26] Mu, M.; Wang, J. F., A method for adjoint variational data assimilation with physical ‘on-off’ processes, J. Atmospheric Sci., 60, 2010-2018, (2003)
[27] Sambridge, M.; Rickwood, P.; Rawlinson, N., Automatic differentiation in geophysical inverse problems, Geophys. J. Int., 170, 1-8, (2007)
[28] Sandu, A., Solution of inverse ODE problems using discrete adjoints, (Biegler, L.; Biros, G.; Ghattas, O.; Heinkenschloss, M.; Keyes, D.; Mallick, B.; Tenorio, L.; van Bloemen Waanders, B.; Willcox, K., Large Scale Inverse Problems and Quantification of Uncertainty, (2010), John Wiley & Sons), 345-364, (chapter 12)
[29] Sandu, A., Reverse automatic differentiation of linear multistep methods, (Lecture Notes in Computational Science and Engineering, 978-3-540-68935-5, vol. 64, (2008), Springer), XVIII, 370p. 111illus. pp. 1-12 · Zbl 1152.65441
[30] Xiao, Y.; Xue, M.; Martin, W.; Gao, J., Development of adjoint for a complex atmospheric model, the ARPS, using TAF, (Martin Bücker, H.; Corliss, George F.; Hovland, Paul; Naumann, Uwe; Norris, Boyana, Automatic Differentiation: Applications, Theory, and Implementations, Lecture Notes in Computational Science and Engineering, vol. 50, (2005), Springer New York, NY), 263-272 · Zbl 1270.86009
[31] Tan, L.; Brytik, V.; Baumstein, A., Verification of gradient and Hessian computation for full wavefield inversion using automatic differentiation, 81st SEG Annual Meeting Abstract, 2762-2766, (2009)
[32] Liao, W., An accurate and efficient algorithm for parameter estimation of 2d acoustic wave equation, Int. J. Appl. Phys. Math., 1, 96-100, (2011)
[33] L. Hascot, V. Pascual, Tapenade 2.1 user’s guide, Rapport technique 300, INRIA Sophia Antipolis,2004. Available from: http://tapenade.inria.fr:8080/tapenade/index.js.
[34] Yang, D.; Peng, J.; Lu, M.; Terlaky, T., Optimal nearly-analytic discrete approximation to the scalar wave equation, Bull. Seismol. Soc. Am., 96, 1114-1130, (2006)
[35] Yang, D.; Tong, P.; Deng, X., A central difference method with low numerical dispersion for solving the scalar wave equation, Geophys. Prospect., 60, 885-905, (2012)
[36] Berenger, J., A perfectly matched layer for the absorption of electromagnetic waves, J. Comput. Phys., 114, 185-200, (1994) · Zbl 0814.65129
[37] Komatitsch, D.; Tromp, J., A perfectly matched layer absorbing boundary condition for the second-order seismic wave equation, Geophys. J. Int., 154, 146-153, (2003)
[38] Liu, Q.; Tao, J., The perfectly matched layer (PML) for acoustic waves in absorptive media, J. Acoust. Soc. Am., 102, 4, 2072-2082, (1997)
[39] Moghaddam, P.; Keers, H.; Herrmann, F.; Mulder, W., A new optimizaiton approach for source-encoding full-waveform inversion, Geophysics, 78, 125-132, (2013)
[40] Symes, W., Reverse time migration with optimal checkpointing, Geophysics, 72, 5, (2007), SM213-SM221
[41] Giles, M.; Duta, M., Algorithm developments for discrete adjoint methods, Am. Inst. Aeronaut. Astronaut. J., 41, 198-205, (2003)
[42] Sen, M.; Stoffa, P., Global optimization methods in geophysical inversion, (1995), Elsevier Science Amsterdam · Zbl 0871.90107
[43] Tran, K.; Hiltunen, D., Two-dimensional inversion of full waveforms using simulated annealing, J. Geotechnical Geoenvironmental Eng., 138, 1075-1090, (2012)
[44] Davidon, W., Variable metric method for minimization, SIAM J. Optim., 1, 1-17, (1991) · Zbl 0752.90062
[45] Liu, D.; Nocedal, J., On the limited memory method for large-scale optimization, Math. Program. B, 45, 503-528, (1989) · Zbl 0696.90048
[46] Morales, J., A numerical study of limited memory BFGS methods, Appl. Math. Lett., 15, 481-487, (2002) · Zbl 1175.90419
[47] Nocedal, J., Updating quasi-Newton matrices with limited storage, Math. Comp., 35, 773-782, (1980) · Zbl 0464.65037
[48] Liao, W., A computational method to estimate the unknown coefficient in a wave equation using boundary measurements, Inverse Probl. Sci. Eng., 19, 6, 855-877, (2011) · Zbl 1252.65160
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.