Statistics and GIS in environmental geochemistry - some problems and solutions

Citation
Cs. Zhang et O. Selinus, Statistics and GIS in environmental geochemistry - some problems and solutions, J GEOCHEM E, 64(1-3), 1998, pp. 339-354
Citations number
25
Categorie Soggetti
Earth Sciences
Journal title
JOURNAL OF GEOCHEMICAL EXPLORATION
ISSN journal
03756742 → ACNP
Volume
64
Issue
1-3
Year of publication
1998
Pages
339 - 354
Database
ISI
SICI code
0375-6742(199811)64:1-3<339:SAGIEG>2.0.ZU;2-T
Abstract
Statistics and geographical information system (GIS) are receiving more and more attention in environmental geochemistry. However, it is important to know the functions and limitations, the advantages and disadvantages of the se techniques for better understanding of their applications. Univariate st atistics is useful for mean calculation, identification of probability dist ribution and outlier detection. Multivariate analysis plays an important ro le in the study of relationships among variables. However, while dealing wi th regionalized variables in environmental geochemistry, the conventional s tatistics show their shortcomings as they are based on some kind of assumpt ions for random variables. Spatial analysis makes use of the spatial coordi nate information of the variables, and also takes the spatial correlation i nto consideration. However, these pure mathematical methods are still unsat isfactory as the nature of environmental geochemistry is far from being so simple. GIS provides visualization and some spatial analysis functions with much spatial information involved. An expert system is useful for classifi cation and prediction based on various types of information. However, the r ule base for expert systems in environmental geochemistry is too small, and needs to be developed. Problems and possible solutions with the applicatio n of statistics and GIS in environmental geochemistry are discussed. Exampl es are based on the authors' experiences in the Yangtze River basin, China, and in southeastern Sweden. Several ideas are discussed in this paper. A ' robust-symmetric mean' proposed by the authors is one of the best methods f or mean calculation. For the probability distribution of trace elements, th e widely accepted 'log-normal distribution' is only a special case of 'posi tively skewed distributions' which is more adequate. The combination of uni variate methods and PCA is used to detect outlying samples. Partial least s quare (PLS) regression, principal component analysis (PCA), cluster analysi s, discriminant analysis and expert systems may be used to differentiate an thropogenic anomalies from the natural background. Spatial correlations amo ng environmental geochemical variables are revealed by cross-variograms. An environmental information system, with the integration of statistics, GIS, expert systems and environmental models should be established to further t he study in environmental geochemistry, as well as to provide decision supp ort. (C) 1998 Elsevier Science B.V. All rights reserved.