베이지안 방식에 의한 지구물리 역산 문제의 접근

A Bayesian Approach to Geophysical Inverse Problems

  • 오석훈 (기상연구소 해양기상지진연구실) ;
  • 정승환 (한국지질자원연구원 탐사개발연구부) ;
  • 권병두 (서울대학교 지구과학교육과) ;
  • 이희순 (인천교육대학교 과학교육과) ;
  • 정호준 ((주)희송지오텍) ;
  • 이덕기 (기상연구소 해양기상지진연구실)
  • Oh Seokhoon (Marine Meteorology & Earthquake Res. Lab/METRI) ;
  • Chung Seung-Hwan (Geophysical Exploration and Mining Division, Korea Institute of Geoscience and Mineral Resources) ;
  • Kwon Byung-Doo (Dept. of Earth Sciences Education, Seoul National University) ;
  • Lee Heuisoon (Dept. Science Education, Inchon Nat'l Univ. of Education) ;
  • Jung Ho Jun (Heesong Geotek, Co. Ltd.) ;
  • Lee Duk Kee (Marine Meteorology & Earthquake Res. Lab/METRI)
  • 발행 : 2002.11.01

초록

본 연구에서는 지구물리 자료의 베이지안 역산을 효과적으로 수행하는 방법에 관해 논의하였다. 베이지안 처리에서 가장 문제가 되는 사전확률분포를 구하기 위해 지구통계학적 방법을 적용하였으며, 사후확률분포의 추정을 위해 MCMC(Markov Chain Monte Carlo) 방법을 적용하였다. 쌍극자배열 전기비저항 탐사 자료의 2차원 역산을 위해 슐럼버저배열 전기비저항탐사 자료와 시추공 자료를 사전 정보로 이용하였으며, 이들 사전정보에 대해 지구통계학적 방법을 적용하여 사전확률분포를 작성하였다. 쌍극자배열 전기비저항 탐사 자료를 최대 우도함수로 하는 사후확률분포는 차원이 매우 높은 적분을 요구하므로, 이를 추정하기 위해 MCMC기술을 적용하였으며, 보다 효율적인 접근을 위해 Gibbs샘플링 방법을 이용하였다. 그 결과 비모수적 방식으로 사후확률분포를 분석함으로써 보다 신뢰성 있는 해를 구할 수 있었으며, 주변화(marginalization)된 사후확률분포를 이용하여 다양한 분석을 적용할 수 있었다.

This study presents a practical procedure for the Bayesian inversion of geophysical data. We have applied geostatistical techniques for the acquisition of prior model information, then the Markov Chain Monte Carlo (MCMC) method was adopted to infer the characteristics of the marginal distributions of model parameters. For the Bayesian inversion of dipole-dipole array resistivity data, we have used the indicator kriging and simulation techniques to generate cumulative density functions from Schlumberger array resistivity data and well logging data, and obtained prior information by cokriging and simulations from covariogram models. The indicator approach makes it possible to incorporate non-parametric information into the probabilistic density function. We have also adopted the MCMC approach, based on Gibbs sampling, to examine the characteristics of a posteriori probability density function and the marginal distribution of each parameter.

키워드

참고문헌

  1. 오석훈, 권병두, 1997, Simulate Annealing을 이용한 전기비저항 자료의 역산: 한국자원공학회지, 34, 285-293
  2. Deutsch, C. V., and Journel, A. G., 1992, GSLIB: Geostatistical Software Library and User's Guide: Oxford University Press, New York.
  3. Duijndam, A. J. W., 1988a, Bayesian estimation in seismic inversion, Part i: Principles: Geophys. Prosp., 36, 878-898
  4. Duijndam, A. J. W., 1988b, Bayesian estimation in seismic inversion, Part ii: Uncertainty analysis: Geophys. Prosp., 36, 899-918.
  5. Gilks, W. R., Richardson, S., and Spiegelhalter, D. J., 1996, Markov Chain Monte Carlo in Practice: Champman & Hall/CRC
  6. Glacken, I., 1996, Change of support by direct conditional block simulation: Master's thesis, Stanford University., Stanford, CA.
  7. Geman, S., and Geman, D., 1984, Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images: IEEE Trans. on Pattern Analysis and Machine Intelligence, PAMI-6, 721-741.
  8. Gomez-Hernandez, J. A., 1991, A stochastic approach to the simulation of block conductivity fields conditional upon data measured at a smaller scale: Doctoral Dissertation, Stanford University, Stanford, CA
  9. Goovaerts, P., 1997, Geostatistics for Natural Resources Evaluation: Oxford University Press, New York.
  10. Gouveia, W. P., 1996, Bayesian seismic waveform data inversion: Parameter estimation and uncertainty analysis: Ph.D. thesis, Colorado School of Mines, Golden, CO
  11. Hastings, W. K., 1970, Monte Carlo sampling methods using Markov chains and their applications: Biometrika, 57, 97-109
  12. Isaaks, E. H., and Srivastava, R. M., 1989, An Introduction to Applied Geostatistics: Oxford University Press, New York
  13. Isaaks, E. H., 1990, The application of Monte Carlo methods to the analysis of spatially correlated Data: Doctoral Dissertation, Stanford University, Stanford, CA
  14. Lines, L. R., and Treital, S., 1984, Tutorial: A review of least-squares inversion and its application to geophysical problems: Geophys. Prosp., 32, 159-186.
  15. Loredo, T. J., 1990, From Laplace to Supernova SN 1987A: Bayesian Inference in Astrophysics, Reprinted from Fougere, P. F. ed., Maximum Entropy and Bayesian Methods, Kluwer Academic Publishers, The Netherlands, 81-142
  16. Moraes, F. S., 1996, The application of marginalization and local distributions to multidimensional Bayesian inverse problems: Ph.D. Thesis, Colorado School of Mines, Golden, Co
  17. MPI Forum, 1995, MPI: A Message-Passing Interface Standard: University of Tennesse, Knoxville, TN
  18. Oh, S.H., and Kwon, B.D., 2001, Geostatistical approach to bayesian inversion of geophysical data: Markov chain Monte Carlo method: Earth Planets Space, 53, 777-791
  19. Pacheco, P. S., 1997, Parallel Programming with MPI: Morgan Kaufmann Publisher, Inc., San Francisco
  20. Tarantola, A., 1987, Inverse problem theory: Methods for data fitting and model parameter estimation: EIsevier, The Netherlands