THE NEED FOR A COMMON FRAMEWORK FOR COLLECTION AND INTERPRETATION OF DATA IN PLATINUM-GROUP ELEMENT GEOCHEMISTRY

Platinum-group element (PGE) concentrations and the distribution of th e metals in rocks serve as important tracers of mantle processes, as w ell as extraterrestrial input into crustal environments, but common st andards regarding the gathering and presentation of PGE data have neve r been formalized. Effective modelling assumes that concentration data are within acceptable levels of precision, yet the practices used in some studies to determine precision do not adequately assess precision and, as a result, the uncertainties on PGE concentrations and PGE rat ios are sometimes consistently underestimated. This article argues tha t replicate analyses of unknowns must be adopted more widely in order to overcome this problem. Related to the issue of uncertainties on PGE concentrations, is the issue of uncertainty associated with normalisa tion. Arguments have recently been put forward as to the significance of small positive or negative anomalies on chondrite normalized plots. At least four CI chondrite PGE datasets (of varying age and quality) are currently used for normalisation and significantly different patte rns can be derived simply by using one dataset rather than another. Th is article is intended to open a debate within the PGE research commun ity by asking whether more consistency needs to be applied in PGE anal ysis and in the subsequent interpretation of data. A rigorous assessme nt of the real uncertainties on PGE concentrations and the adoption of ct standard CI chondrite PGE dataset, in order to eliminate bias from normalisation, are suggested to be central to this.