SPATIAL FACTOR ANALYSIS OF STREAM SEDIMENT GEOCHEMISTRY DATA FROM SOUTH GREENLAND

This paper describes the application of min/max autocorrelation factor (MAF) analysis to irregularly sampled stream geochemical data from South Greenland. Kriged MAF images are compared with kriged images based on varimax rotated factors (VRF) from an ordinary non-spatial factor analysis, and the image patterns are discussed in a geological context. It is demonstrated that MAF analysis has a potential for objectively dividing multi-element data into spatial segments coinciding with large lithotectonic units.