ISSN:
1432-0495
Keywords:
Key words Background
;
Data distribution
;
Normal distribution
;
Log-normal distribution
;
Geochemistry
;
Environmental science
;
Robust geostatistics
Source:
Springer Online Journal Archives 1860-2000
Topics:
Geosciences
Notes:
Abstract The term "background" is discussed and a definition is suggested to put an end to the currently unsatisfying (non-)definition of geochemical or natural background. Based on the requirements stated in the definition, several simple and robust statistical methods are applied to different data sets (n〉50) from the atmosphere, pedosphere, and biosphere in order to explore their potential for the evaluation of a useful and robust background. Compared with the original data set both the calculated distribution, based upon the lower 50% of the values, as well as a 2σ-approximation of the normalised data set lead to promising and realistic results. Both methods seem appropriate for a fast and reliable evaluation of likely upper limits of background values. Nevertheless, even this robust method is not able to present absolute and doubtlessly correct background values. True quantification of any natural or geochemical background still requires a thorough investigation and is impossible without costly expert knowledge.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/s002549900084
