Interpolation methods for geochemical maps; a comparative study using arsenic data from European stream waters

A geochemical map of As in water from the FOREGS Geochemical Atlas of Europe, performed using the Alkemia interpolation method based on moving weighted median (MWM), and a comparable map prepared by kriging are compared with an As map prepared with a new multifractal inverse distance weighted (MIDW) interpolation method using GeoDas (super TM) software. The colour scale classification of the MIDW interpolated map of As is based on the concentration-area (C-A) fractal method which allows images to be subdivided into components representing specific features on the ground related, for example, to geology. Conventional techniques, such as MWM and kriging, are shown to smooth out the local variability of the geochemical data. The problem is most serious in maps prepared by kriging which erroneously show large areas of Europe to have high levels of As in water. On the other hand, MIDW creates a geochemical map in which information about the local data structure is retained. This is essential in distinguishing anomalies from background values. The information provided by background and anomaly maps, using the MIDW and fractal filtering methods, are shown to give more reliable upper limits of background values.