Advertisement

Using Mahalanobis Distance to Detect and Remove Outliers in Experimental Covariograms

  • David Alvarenga Drumond
  • Roberto Mentzingen Rolo
  • João Felipe Coimbra Leite Costa
Original Paper
  • 39 Downloads

Abstract

Experimental variograms are crucial for most geostatistical studies. In kriging, for example, the variography has a direct influence on the interpolation weights. Despite the great importance of variogram estimators in predicting geostatistical features, they are commonly influenced by outliers in the dataset. The effect of some randomly spatially distributed outliers can mask the pattern of the experimental variogram and produce a destructuration effect, implying that the true data spatial continuity cannot be reproduced. In this paper, an algorithm to detect and remove the effect of outliers in experimental variograms using the Mahalanobis distance is proposed. An example of the algorithm’s application is presented, showing that the developed technique is able to satisfactorily detect and remove outliers from a variogram.

Keywords

Mahalanobis distance Outliers Variogram 

References

  1. Ben-Gal, I. (2005). Outlier detection. In Data mining and knowledge discovery handbook (pp. 131–146).Google Scholar
  2. Costa, J. F. (2003). Reducing the impact of outliers in ore reserves estimation. Mathematical Geology, 35(3), 323–345.CrossRefGoogle Scholar
  3. Cressie, N., & Hawkins, D. M. (1980). Robust estimation of the variogram: I. Journal of the International Association for Mathematical Geology, 12(2), 115–125.CrossRefGoogle Scholar
  4. Dutter, R. (1996). On robust estimation of variograms in geostatistics. Robust statistics, data analysis, and computer intensive methods (pp. 153–171). Berlin: Springer.CrossRefGoogle Scholar
  5. Filzmoser, P. (2004). A multivariate outlier detection method. na.Google Scholar
  6. Genton, M. G. (1998). Highly robust variogram estimation. Mathematical Geology, 30(2), 213–221.CrossRefGoogle Scholar
  7. Hazewinkel, M. (2001). Chebyshev inequality in probability theory. Encyclopedia of mathematics. Berlin: Springer.Google Scholar
  8. Krige, D. G., & Magri, E. J. (1982). Studies of the effects of outliers and data transformation on variogram estimates for a base metal and a gold ore body. Journal of the International Association for Mathematical Geology, 14(6), 557–564.CrossRefGoogle Scholar
  9. Lebrenz, H., & Bárdossy, A. (2017). Estimation of the variogram using Kendall’s tau for a robust geostatistical interpolation. Journal of Hydrologic Engineering, 22(9), 04017038.CrossRefGoogle Scholar
  10. Mahalanobis, P. C. (1936). On the generalised distance in statistics. Proceedings of the National Institute of Sciences of India, 1936, 49–55.Google Scholar
  11. O’Leary, B., Reiners, J. J., Xu, X., & Lemke, L. D. (2016). Identification and influence of spatio-temporal outliers in urban air quality measurements. Science of the Total Environment, 573, 55–65.CrossRefGoogle Scholar
  12. Rousseeuw, P. J., & Van Zomeren, B. C. (1990). Unmasking multivariate outliers and leverage points. Journal of the American Statistical association, 85(411), 633–639.CrossRefGoogle Scholar
  13. Saw, J. G., Yang, M. C., & Mo, T. C. (1984). Chebyshev inequality with estimated mean and variance. The American Statistician, 38(2), 130–132.Google Scholar
  14. Srivastava, R. M. (2001). Outliers: A guide for data analysts and interpreters on how to evaluate unexpected high values. Contaminated sites statistical applications guidance document no. 12-8, BC, Canada, 4 pp. https://www2.gov.bc.ca/assets/gov/environment/air-land-water/site-remediation/docs/guidance-documents/gd08.pdf. Accessed 4 Aug 2018.
  15. Werner, M. (2003). Identification of multivariate outliers in large data sets. Ph.D. thesis, Citeseer.Google Scholar
  16. Ziegel, E. R. (1995). Gslib: Geostatistical software library and user’s guide.Google Scholar

Copyright information

© International Association for Mathematical Geosciences 2018

Authors and Affiliations

  • David Alvarenga Drumond
    • 1
  • Roberto Mentzingen Rolo
    • 1
  • João Felipe Coimbra Leite Costa
    • 1
  1. 1.Universidade Federal do Rio Grande do Sul (UFRGS)Porto AlegreBrazil

Personalised recommendations