Analysis of Geochemical Data Sampled on a Regional Scale

Knut Conradsen, Allan Aasbjerg Nielsen, Kristian Windfeld

AbstractWhen working with the analysis of geochemical data on a regional scale one will often encounter calibration problems arising from the chemical analysis of sampies that might be collected over several years in geographical units. A procedure for remo val of such geographical unit patterns is described. Stream sediment sampies are collected at 33992 sites and analyzed for the content of 26 elements. Semivariograms, the geostatisticai analogue af spatial autocorrelation functions, are estimated, modeled and applied in interpolating to a regular grid. The interpolation method used is kriging, in which the mean square prediction error is minimized taking the spatial correlation in to account. Heavy mineral sampies are collected at 3094 sites and analyzed for the content of several economically interesting minerals. Based on the nonparametric method, CART (classification and regression trees) probability maps for the accurrence af gold and cassiterite are canstructed using the krige d stream sediment data as predictors. This method is very useful for geologists when studying geological structures on a regional scale.
TypeBook [Chapter]
Book titleStatistics in the Environmental and Earth Sciences
EditorsAndrew T. Walden and Peter Guttorp
Year1992    pp. 283-300
PublisherEdward Arnold
AddressLondon, England
NoteInvited contribution
Electronic version(s)[pdf]
BibTeX data [bibtex]
IMM Group(s)Image Analysis & Computer Graphics