Prediction of the Gold-silver Deposits from Geochemical Maps - Applications to the Bayesian Geostatistics and Decision Tree Techniques

지화학자료를 이용한 금${\cdot}$은 광산의 배태 예상지역 추정-베이시안 지구통계학과 의사나무 결정기법의 활용

  • Hwang, Sang-Gi (Department of Civil & Geotechnical Engineering, Paichai University) ;
  • Lee, Pyeong-Koo (Geological & Environmental Hazards Division, Korea Institute of Geoscience and Mineral Resources)
  • 황상기 (배재대학교 토목환경공학과) ;
  • 이평구 (한국지질지원 연구원 지질환경재해 연구부)
  • Published : 2005.12.01

Abstract

This study investigates the relationship between the geochemical maps and the gold-silver deposit locations. Geochemical maps of 21 elements, which are published by KIGAM, locations of gold-silver deposits, and 1:1,000,000 scale geological map of Korea are utilized far this investigation. Pixel size of the basic geochemical maps is 250m and these data are resampled in 1km spacing for the statistical analyses. Relationship between the mine location and the geochemical data are investigated using bayesian statistics and decision tree algorithms. For the bayesian statistics, each geochemical maps are reclassified by percentile divisions which divides the data by 5, 25, 50, 75, 95, and $100\%$ data groups. Number of mine locations in these divisions are counted and the probabilities are calculated. Posterior probabilities of each pixel are calculated using the probability of 21 geochemical maps and the geological map. A prediction map of the mining locations is made by plotting the posterior probability. The input parameters for the decision tree construction are 21 geochemical elements and lithology, and the output parameters are 5 types of mines (Ag/Au, Cu, Fe, Pb/Zn, W) and absence of the mine. The locations for the absence of the mine are selected by resampling the overall area by 1 km spacing and eliminating my resampled points, which is in 750m distance from mine locations. A prediction map of each mine area is produced by applying the decision tree to every pixels. The prediction by Bayesian method is slightly better than the decision tree. However both prediction maps show reasonable match with the input mine locations. We interpret that such match indicate the rules produced by both methods are reasonable and therefore the geochemical data has strong relations with the mine locations. This implies that the geochemical rules could be used as background values oi mine locations, therefore could be used for evaluation of mine contamination. Bayesian statistics indicated that the probability of Au/Ag deposit increases as CaO, Cu, MgO, MnO, Pb and Li increases, and Zr decreases.

지화학 자료의 공간적 분포와 금은광산의 공간적 분포사이의 상관관계를 조사하였다. 활용된 자료는 한국자원연구소에서 발간된 지화학도 중 21개 원소에 대한 도면과, 현재까지 파악된 광산의 위치도면 및 1:100만 지질도이다. 지화학도는 250m 등간격의 격자형 화소로 제작된 도면 중 통계분석을 위하여 1km 간격의 자료를 추출하여 분석하였으며, 광산위치의 지화학 자료 역시 250m 간격의 화소에서 추출하여 분석을 수행하였다. 광산과 지화학자료의 공간적인 상관분석은 베이시안 중첩법과 의사결정나무 기법을 활용하였디. 베이시안 통계기법은 각 지화학도에 분포하는 원소의 화소값을 올림차순으로 정열한 후 자료의 개수가 자기 5, 25, 50, 75, 95, $100\%$에 해당하는 등급을 나누어 모든 지화학도를 6개의 등급을 갖는 도면으로 재분류 하였다. 자 등급에 속한 광산의 개수를 대상으로 광산이 발생할 확률이 계산되었으며, 이 확률을 취합하여 최종 사후확률이 계산되었으며, 사후확률로 광산이 배태될 예측 도면이 작성되었다. 금/은, 동, 철, 납/아연, 텅스텐광산 및 광산이 존재하지 않는 위치에 해당하는 지화학 자료와 암상을 기준으로 의사결정나무를 학습시키고, 학습된 결과를 전체 자료에 적용하여 예측도면을 작성하였다. 광산이 존재하지 않은 지역을 추출하기 위하여 지화학도의 화소를 1km간격으로 추출한 후 이들 중 광산과 750m이내에 있는 자료는 제외시키는 알고리듬을 활용하였다. 예측결과 베이시안 방법에 의한 광산의 위치 예측이 의사결정나무에 의한 예측보다 상대적으로 정확함이 확인되었다. 그러나 두 방법 모두 공히 기존의 광산위치를 적절히 예측하고 있어서 지화학 자료는 광산의 위치와 밀접한 관계를 갖고 있음이 확인되었다.

Keywords

References

  1. 강필종 (1995) 한국 지질도, 축척 1:1,000,000. 한국자원연구소, 성지문화사
  2. 신성천 (2001) 한국 지구화학 지도책(1:700,000), 한국지질자원연구원
  3. Asadi H.H. and Hale M. (2001) Apredictive GIS model for mapping potential gold and base metal mineralization in Takab area, Iran. Computer & Geosciences v.27, p. 901-912 https://doi.org/10.1016/S0098-3004(00)00130-8
  4. Bolviken, B. (1986) Geochemical atlas of northern Fen-noscandia, scale 1:4.000.000. Geological Survey of Sweden, 19pp, 155 maps
  5. Bonham-Carter, G.E, Agterberg, EP. and Wright, D.E (1989) Weights of evidence modelling: a new approach to mapping mineral potential. In: Agterberg, ER, Bonham-Carter, G.E, (Eds.), Statistical Applications in the Earth Sciences. Geological Survey of Canada Paper, 89-9, p. 171-183
  6. Bonham-Carter, G.E (1994) Geographic Information Systems for geoscientists, modeling with GIS. Pergamon Press, Oxford, 398pp
  7. Breiman, L., Friedman J.H., Olshen R. and Stone C. (1984) Classification and Regression Trees, Wad-sworth International Group, California
  8. Carranza, E.J.M. and Hale, M. (1999) Geological-constrained probabilistic mapping of gold potential, Baguio District, Philippines. Geocomputation 99, July 2528, Fredericksburg, Virginia, Conference Volume on CD-ROM
  9. Debes J.D. and Urrutia R. (2004) Bioinformatics tools to understand human diseases. Surgery 135, v. 6, p. 579-585
  10. Harris, J.R., L. Wilkinson, J. Broome, and S. Fumerton, (1995) Mineral exploration using GIS-based favour-ability analysis, Swayze greenstone belt, northern Ontario, in Proceedings of 1995 Canadian Geomatics, Ottawa, Ontario Canada
  11. Hwang, S.G., Nguyen Q.P. and Lee P.K. (2005) Reproducibility of a regional geological map derived from geochemical maps, using data mining techniques: with application to Chungbuk province of Korea. Environmental Geology, in press
  12. IGS. (1978) Geochemical atlas of Gt. Britain: Shetland Islands. Institute of Geological Sciences, London
  13. Kass, G.V (1980) An exploratory technique for investigating quantities of categorical data. Applied Statistics 29 v. 2, p. 119-127
  14. Lahermo, E, Vaananen E, Tarvainen T. and Salminen R. (1996) Geochemical Atlas of Finland, Part 3: Environmental geochemistry - Stream waters and sediments. Geological Survey of Finland, Espoo, 149p
  15. Meyer, W.T., Theobald EK., Bloom H. (1979) Stream sediment geochemistry. In: Hood EJ. (ed) Geophysics and geochemistry in the search for metallic ores. Geol Surv Can. Econ. Geol. Rep., v.31, p 411-434
  16. Quinlan, J.R. (1986) Induction of decision trees. Machine Learning 1, v.1, p. 81-106
  17. Quinlan, J.R. (1993) C4.5: Programs for machine learning. San Francisco: Morgan Kaufmann Publishers
  18. Shannon, C.E. and Weaver W (1949) The Mathematical Theory of Communication, University of Illinois Press
  19. Webb, J.S., Thornton I., Howarth R.J. and Thompson M. (1978) The Wolfson Geochemical Atlas of England and Wales. Clarendon Press, United Kingdom, 69pp
  20. Wright, D.E and Bonham-Carter G.E (1996) VHMS favourability mapping with GIS-based integration models, Chisel Lake-Anderson Lake area, in : Bonham-Carter, Galley, and Hall (eds.): EXTECHLA mul-tidisciplinary approach to massive sulfide research in the Rusty Lake-Snow Lake greenstone belts, Manitoba. Geolocical Survey of Canada, Bulletin, v. 426, p. 339-376