Mathematical Geosciences

, Volume 42, Issue 6, pp 681–703 | Cite as

Hunting for Geochemical Associations of Elements: Factor Analysis and Self-Organising Maps

  • Gorazd Žibret
  • Robert Šajn


Two approaches, factor analysis (FA) and self-organising maps (SOM), have been used for the determination of geochemical associations in the two case studies evaluated here. In both case studies, different associations of elements, derived from different anthropogenic sources (smelters, ironworks, and chemical industry), are presented, together with natural associations of elements, all representing different geological environments. FA and SOM give similar results, despite some differences. Most similarities were achieved with the group of elements affected by strong pollution caused by smelting activities. The biggest difference between the two is that SOM can combine different groups into one, especially in the case of associations of elements connected with mild pollution (ironworking, chemical industry). The biggest advantage of SOM as opposed to FA is that SOM allow us to process variables, which are not normally distributed, or even of attributive nature. The use of such variables in SOM classification to prove the origins of associations of elements is also demonstrated here.


Self-organising maps Factor analysis Celje Mežica Heavy metals Compositional data processing 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Abdi H (2003) Factor rotations. In: Lewis-Beck M, Bryman A, Futing T (eds) Encyclopedia for research methods for the social sciences. Sage, Thousand Oaks, pp 978–982 Google Scholar
  2. Aitchison J (1986) The statistical analysis of compositional data. Wiley, New York Google Scholar
  3. Aitchison J, Barcelo-Vidal C, Martin-Fernandez JA, Pawlowsky-Glahn V (2000) Logratio analysis and compositional distance. Math Geol 32(3):271–275 CrossRefGoogle Scholar
  4. Alvarez-Guerra M, González-Piñuela C, Andrés A, Galán B, Viguri JR (2008) Assessment of Self-Organizing Map artificial neural networks for the classification of sediment quality. Environ Int 34(6):782–790 CrossRefGoogle Scholar
  5. Arias R, Barona A, Ibarra-Berastegi G, Aranguiz I, Elías A (2008) Assessment of metal contamination in dredged sediments using fractionation and Self-Organizing Maps. J Hazard Mater 151(1):78–85 CrossRefGoogle Scholar
  6. Astel A, Tsakovski S, Barbieri P, Simeonov V (2007) Comparison of self-organizing maps classification approach with cluster and principal components analysis for large environmental data sets. Water Res 41:4566–4578 CrossRefGoogle Scholar
  7. Badran F, Yacoub M, Thiria S (2005) Self-organizing maps and unsupervised classification. In: Dreyfus G (ed) Neural networks: methodology and applications. Springer, Berlin, pp 379–442 Google Scholar
  8. Bleeker EAJ, Van Gestel CAM (2007) Effects of spatial and temporal variation in metal availability on earthworms in floodplain soils of the river Dommel, The Netherlands. Environ Pollut 148:824–832 CrossRefGoogle Scholar
  9. Child D (2006) The essentials of factor analysis. Continuum, London, p 180 Google Scholar
  10. Covelli S, Piani R, Faganeli J, Brambati A (2004) Circulation and suspended matter distribution in a microtidal deltaic system: the Isonzo river mouth (northern Adriatic Sea). J Coast Res 41:130–141 Google Scholar
  11. Covelli S, Fontolan G, Faganeli J, Ogrinc N (2006) Anthropogenic markers in the Holocene stratigraphic sequence of the Gulf of Trieste (northern Adriatic Sea). Mar Geol 230:29–51 CrossRefGoogle Scholar
  12. Egozcue JJ, Pawlowsky-Glahn V (2005) Groups of parts and their balances in compositional data analysis. Math Geol 37(7):795–828 CrossRefGoogle Scholar
  13. Egozcue JJ, Pawlowsky-Glahn V, Mateu-Figueras G, Barcelo-Vidal C (2003) Isometric logratio transformations for compositional data analysis. Math Geol 35(3):279–300 CrossRefGoogle Scholar
  14. Filzmoser P, Hron K (2008) Outlier detection for compositional data using robust methods. Math Geosci 40(3):233–248 CrossRefGoogle Scholar
  15. Filzmoser P, Hron K, Reimann C, Garrett R (2009) Robust factor analysis for compositional data. Comput Geosci 35:1854–1861 CrossRefGoogle Scholar
  16. Frontasyeva MV, Galinskaya TY, Krmar M, Matavuly M, Pavlov SS, Povtoreyko EA, Radnovic D, Steinnes E (2004) Atmospheric deposition of heavy metals in northern Serbia and Bosnia-Herzegovina studied by moss biomonitoring, neutron activation analysis and GIS technology. J Radioanal Nucl Chem 259(1):141–147 CrossRefGoogle Scholar
  17. Gosar M (2008) Mercury in river sediments, floodplains and plants growing thereon in drainage area of Idrija mine, Slovenia. Pol J Environ Stud 17(2):227–236 Google Scholar
  18. Gosar M, Šajn R (2003) Geochemical soil and attic dust survey in Idrija, Slovenia. J Phys (Paris) 107:561–564 Google Scholar
  19. Gosar M, Šajn R, Biester H (2006) Binding of mercury in soils and attic dust in the Idrija mercury mine area (Slovenia). Sci Total Environ 369(1/3):150–162 Google Scholar
  20. Haykin S (1999) Neural networks—a comprehensive foundation, 2nd edn. Pearson, Singapore, p 842 Google Scholar
  21. Jeran Z, Jaaimovia R, Batie F, Smodi B, Wolterbeek HT (1996) Atmospheric heavy metal pollution in Slovenia derived from results for epiphytic lichens. Fresen J Anal Chem 354:681–687 Google Scholar
  22. Jeran Z, Jačimovič R, Batič F, Mavsar R (2002) Lichens as integrating air pollution monitors. Environ Pollut 120:107–113 CrossRefGoogle Scholar
  23. Jetter T (2009) MemBrain Neural Network Simulator V. Available via DIALOG.
  24. Kartal S, Aydin Z, Tokahoglu S (2006) Fractionation of metals in street sediment samples by using the BCR sequential extraction procedure and multivariate statistical elucidation of the data. J Hazard Mater 132:80–89 CrossRefGoogle Scholar
  25. Kohonen T (2001) Self-organizing maps, 3rd edn. Springer, Berlin, 501 Google Scholar
  26. Kooistra L, Salas EAL, Clevers JGPW, Wehrens R, Leuven RSEW, Nienhuis PH, Buydens LMC (2004) Exploring field vegetation reflectance as an indicator of soil contamination in river floodplains. Environ Pollut 127:281–290 CrossRefGoogle Scholar
  27. Kulahci F, Sen Z (2008) Multivariate statistical analyses of artificial radionuclides and heavy metal contaminations in deep mud of Keban Dam Lake, Turkey. Appl Radiat Isot 66:236–246 CrossRefGoogle Scholar
  28. Lacassie JP, Roser B, Del Solar JR, Herve F (2004) Discovering geochemical patterns using self-organizing neural networks: a new perspective for sedimentary provenance analysis. Sediment Geol 165:175–191 CrossRefGoogle Scholar
  29. Lee BH, Scholz M (2006) Application of the self-organizing map (SOM) to assess the heavy metal removal performance in experimental constructed wetlands. Water Res 40:3367–3374 CrossRefGoogle Scholar
  30. Maiz I, Arambarri I, Garcia R, Millan E (2000) Evaluation of heavy metal availability in polluted soils by two sequential extraction procedures using factor analysis. Environ Pollut 110:3–9 CrossRefGoogle Scholar
  31. Mele PM, Crowley DE (2008) Application of self-organizing maps for assessing soil biological quality. Agric Ecosyst Environ 126:139–152 CrossRefGoogle Scholar
  32. Pasternak GB, Brown KJ (2006) Natural and anthropogenic geochemical signatures of floodplain and deltaic sedimentary strata, Sacramento, San Joaquin Delta, California, USA. Environ Pollut 141:295–309 CrossRefGoogle Scholar
  33. Pawlowsky-Glahn V, Egozcue JJ (2001) Geometric approach to statistical analysis on the simplex. Stoch Environ Res Risk Assess 15/5:384–398 CrossRefGoogle Scholar
  34. Pawlowsky-Glahn V, Egozcue JJ (2006) Compositional data and their analysis: An introduction. In: Buccianti A, Mateu-Figueras G, Pawlowsky-Glahn V (eds) Compositional data analysis in the geosciences: from theory to practice. Special publications, vol 264. Geological Society, London, pp 1–10 Google Scholar
  35. Pawlowsky-Glahn V, Egozcue JJ, Tolosana-Delgado R (2007) Lecture notes on compositional data analysis. In: wePapers. Available via DIALOG.
  36. Poot A, Gillissen F, Koelmans AA (2007) Effects of flow regime and flooding on heavy metal availability in sediment and soil of a dynamic river system. Environ Pollut 148:779–787 CrossRefGoogle Scholar
  37. Reimann C, Filzmoser P, Garrett RG (2002) Factor analysis applied to regional geochemical data: problems and possibilities. Appl Geochem 17:185–206 CrossRefGoogle Scholar
  38. Reimann C, Filzmoser P, Garrett RG, Dutter R (2008) Statistical data analysis explained: applied environmental statistics with R. Wiley, Chichester Google Scholar
  39. Šajn R (1999) Geokemične lastnosti urbanih sedimentov na ozemlju Slovenije (Geochemical properties of urban sediments on the territory of Slovenia). Geological Survey of Slovenia, Ljubljana Google Scholar
  40. Šajn R (2005) Using attic dust and soil for the separation of anthropogenic and geogenic elemental distributions in an old metallurgic area (Celje, Slovenia). Geochem, Explor Environ Anal 5(1):59–67 CrossRefGoogle Scholar
  41. Šajn R (2006) Factor analysis of soil and attic-dust to separate mining and metallurgy influence, Meža Valley, Slovenia. Math Geol 38(6):735–747 Google Scholar
  42. Samecka-Cymerman A, Stankiewicz A, Kolon K, Kempers AJ (2007) Self-organizing feature map (neural networks) as a tool in classification of the relations between chemical composition of aquatic bryophytes and types of streambeds in the Tatra national park in Poland. Chemosphere 67:954–960 CrossRefGoogle Scholar
  43. Samecka-Cymerman A, Stankiewicz A, Kolon K, Kempers AJ (2009) Self-organizing feature map (neural networks) as a tool to select the best indicator of road traffic pollution (soil, leaves or bark of Robinia pseudoacacia L.). Environ Pollut 157:2061–2065 CrossRefGoogle Scholar
  44. Stafilov T, Šajn R, Pančevski Z, Boev B, Frontasyeva M, Strelkova L (2009) Heavy metal contamination of topsoils around a lead and zinc smelter in the Republic of Macedonia. J Hazard Mater. doi: 10.1016/j.jhazmat.2009.10.094 Google Scholar
  45. Tan HS, George SE (2005) Investigating learning parameters in a standard 2-D SOM model to select good maps and avoid poor ones. In: Webb GI, Yu X (eds) AI 2004: Advances in artificial intelligence. Springer, Berlin, pp 425–437 Google Scholar
  46. Tsakovski S, Kudlak B, Simeonov V, Wolska L, Namiesnik J (2009) Ecotoxicity and chemical sediment data classification by the use of self-organising maps. Anal Chim Acta 631:142–152 CrossRefGoogle Scholar
  47. Valls RA (2008) Why, and how, we should use compositional data analysis—a step-by-step guide for field geologists. Wikibooks, Toronto Google Scholar
  48. Van Griethuysen C, Luitwieler M, Joziasse J, Koelmans AA (2005) Temporal variation of trace metal geochemistry in floodplain lake sediment subject to dynamic hydrological conditions. Environ Pollut 137:281–294 CrossRefGoogle Scholar
  49. Vesanto J, Himberg J, Alhoniemi E, Parhankagas J (2000) SOM Toolbox for Matlab 5, Report A57. Helsinki University of Technology, Helsinki Google Scholar
  50. Wijnhoven S, van der Velde G, Leuven RSEW, Eijsackers HJP, Smits AJM (2006) The effect of turbation on zinc relocation in a vertical floodplain soil profile. Environ Pollut 140:444–452 CrossRefGoogle Scholar
  51. Yongming H, Peixuan D, Junji C, Posmentier ES (2006) Multivariate analysis of heavy metal contamination in urban dusts of Xi’an, Central China. Sci Total Environ 355:176–186 CrossRefGoogle Scholar
  52. Žibret G (2002) Geokemične lastnosti tal in podstrešnega prahu na območju Celja (geochemical properties of soil and attic dust in Celje area). BSc thesis, University of Ljubljana, Ljubljana, p 76 Google Scholar
  53. Žibret G, Rokavec D (2009) Household dust and street sediment as an indicator of recent heavy metals in atmospheric emissions: a case study in a previously heavily contaminated area. Environ Earth Sci. doi: 10.1007/s12665-009-0356-2 Google Scholar

Copyright information

© International Association for Mathematical Geosciences 2010

Authors and Affiliations

  1. 1.Geological Survey of SloveniaLjubljanaSlovenia

Personalised recommendations