Parallel computations for the simulation of seismic waves on the basis of the additive Schwartz method

Authors

  • M.A. Belonosov Trofimuk Institute of Petroleum Geology and Geophysics of SB RAS https://orcid.org/0000-0001-6527-7906
  • S.A. Solovyev Trofimuk Institute of Petroleum Geology and Geophysics of SB RAS
  • V.A. Cheverda Trofimuk Institute of Petroleum Geology and Geophysics of SB RAS
  • K. Kostov Schlumberger Moscow Research Center
  • G.V. Reshetova The Institute of Computational Mathematics and Mathematical Geophysics of SB RAS (ICM&MG SB RAS)

Keywords:

Laguerre transform, definite operator, additive Schwartz method, parallel computations, scalability

Abstract

The Laguerre time transform for the elastic wave equations leads to a definite spatial operator independent of a separation parameter. This allows one to perform parallel computations on the basis of Schwartz alternations using a domain decomposition with overlapping. On each alternation step, the resulting system of linear algebraic equations in each subdomain is solved independently, so one can use a direct solver on the basis of the LU-decomposition. Since the spatial operator is independent of separation parameters, this decomposition can be performed only once and be saved in RAM for each elementary subdomain to use for all right-hand sides. This approach is implemented and the corresponding software for high performance computers with hybrid parallel architecture is developed. A number of numerical results illustrating the analysis of scalability are discussed.

Author Biographies

M.A. Belonosov

S.A. Solovyev

V.A. Cheverda

K. Kostov

Schlumberger Moscow Research Center
• Head of Department

G.V. Reshetova

References

  1. Вишневский Д.М., Лисица В.В., Чеверда В.А. Комбинирование конечно-разностных схем для моделирования волновых процессов в упругих средах, содержащих анизотропные слои // Сибирский журнал вычислительной математики. 2012. 15, № 5. 175-181.
  2. Конюх Г.В., Михайленко Б.Г. Применение интегрального преобразования Лагерра при решении динамических задач сейсмики // Тр. ИВМиМГ СО РАН. Сер. Математическое моделирование в геофизике. 1998. № 5. 106-123.
  3. Костин В.И., Лисица В.В., Решетова Г.В., Чеверда В.А. Конечно-разностный метод численного моделирования распространения сейсмических волн в трехмерно-неоднородных разномасштабных средах // Вычислительные методы и программирование. 2011. 12, № 2. 85-93.
  4. Левченко В.Д. Асинхронные алгоритмы как способ достижения 100
  5. Мацокин А.М., Непомнящих С.В. Метод альтернирования Шварца в подпространствах // Изв. высших учебных заведений. 1985. 29, № 10. 61-66.
  6. Решетова Г.В., Чеверда В.А. Использование преобразования Лагерра для построения идеально подходящих поглощающих слоев без расщепления // Математическое моделирование. 2006. 18, № 10. 91-101.
  7. Суетин П.К. Классические ортогональные многочлены. M.: Наука, 1974.
  8. Chan T., Mathew T.P. Domain decomposition // Acta Numerica. 1994. 3. 61-143.
  9. Colella P., Bell J., Keen N., Ligocki T., Lijewski M., van Straalen B. Performance and scaling of locally-structured grid methods for partial differential equations // Journal of Physics: Conference Series. 2007. 78. 012013.
  10. Collino F., Tsogka C. Application of the PML absorbing layer model to the linear elastodynamic problem in anisotropic heterogeneous media // Geophysics. 2001. 66, N 1. 294-307.
  11. Fossen H., Hesthammer J. Structural geology of the Gullfaks Field, northern North Sea // Geological Society. London. Special Publications. 1998. 127. 231-261.
  12. Gander M., Haloern L., Nataf F. Optimized Schwarz methods // Proc. of 12th Int. Conf. on Domain Decomposition Methods. Chiba, 2001. 15-27.
  13. Gould N.I. M., Hu Y., Scott J.A. A numerical evaluation of sparse direct solvers for the solution of large sparse, symmetric linear systems of equations. Technical Report RAL-TR-2005-005. Rutherford Appleton Laboratory. Chilton, 2005.
  14. Graves R.W. Simulating seismic wave propagation in 3D elastic media using staggered-grid finite differences // Bull. of the Seismological Society of America. 1996. 86, N 4. 1091-1106.
  15. Karypis G., Kumar V. A fast and high quality multilevel scheme for partitioning irregular graphs // SIAM J. on Scientific Computing. 1998. 20, N 1. 359-392.
  16. Korneev V.I., Kostin V.I., Kovalevsky V.V., Malyshkin V.E., Tcheverda V.A., Vishnevsky D.M. Modelling of seismic waves propagation for 2D media (direct and inverse problems) // Lecture Notes in Computer Sciences. Vol. 1277. 1997. 350-357.
  17. Kostin V.I., Lisitsa V.V., Reshetova G.V., Tcheverda V.A. Simulation of seismic waves propagation in multiscale media: impact of cavernous/fractured reservoirs // Lecture Notes in Computer Sciences. Vol. 7133. 2012. 54-64.
  18. McCully J. The Laguerre transform // SIAM Review. 1960. 2, N 3. 185-191.
  19. Mikhailenko B.G., Mikhailov A.A, Reshetova G.V. Numerical viscoelastic modeling by the spectral Laguerre method // Geophysical Prospecting. 2003. 51. 37-48.
  20. Nepomnyashchikh S.V. Domain decomposition methods // Radon Series Comput. Appl. Math. 2007. 1. 81-159.
  21. Pissarenko D., Reshetova G., Tcheverda V. 3D finite-difference synthetic acoustic log in cylindrical coordinates: parallel implementation // J. of Computational and Applied Mathematics. 2010. 234, № 6. 1766-1772.
  22. Plessix R.E. A Helmholtz iterative solver for 3D seismic-imaging problems // Geophysics. 2007. 72, N 5. SM185-SM194.
  23. Schenk O., Gartner K., Fichtner W. Efficient sparse LU factorization with left-right looking strategy on shared memory multiprocessors // BIT. 2000. 240, N 1. 158-176.
  24. Schenk O., Gartner K. Sparse factorization with two-level scheduling in PARDISO // Proc. of the 10th SIAM Conf. on Parallel Processing for Scientific Computing. Portsmouth, 2001.
  25. Schenk O., Gartner K. Two-level scheduling in PARDISO: improved scalability on shared memory multiprocessing systems // Parallel Computing. 2002. 28. 187-197.
  26. Virieux J. P-SV wave propagation in heterogeneous media: velocity-stress finite-difference method // Geophysics. 1986. 51, N 4. 889-901.
  27. Zahradnik J., Priolo E. Heterogeneous formulations of elastodynamic equations and finite-difference schemes // Geophysical Journal International. 1995. 120, N 3. 663-676.
  28. Zhang Z., Zha H., Simon H. Low-Rank approximation with sparse factors: basic algorithms and error analysis // SIAM J. of Matrix Analysis and Applications. 1999. 23, N 3. 706-727.

Published

07-11-2012

How to Cite

Белоносов М., Соловьёв С., Чеверда В., Костов К., Решетова Г. Parallel Computations for the Simulation of Seismic Waves on the Basis of the Additive Schwartz Method // Numerical Methods and Programming (Vychislitel’nye Metody i Programmirovanie). 2012. 13. 525-535

Issue

Section

Section 1. Numerical methods and applications

Most read articles by the same author(s)