CONSTRAINING ERRORS IN FIELD AND LAB GEOCHEMICAL ANALYSES USING A PXRF CALIBRATION FOR COLUMBIA RIVER FLOOD BASALT

Quantitative analyses of local basalt samples are possible using a calibration made for the EWU Geology portable XRF. However, the difference of measurements between rough or polished rock surfaces is not well constrained in the EWU lab or the literature. Ideally the goal is to establish constraints on the deviation in mass percent when using the Bruker Tracer III PXRF on freshly broken rough surfaces in the field. For this study, analyses were re-run on the samples used to create the calibration, with a focus on polished surfaces versus rough surfaces. As expected, flat surfaces have overall similar results for most major and trace elements. Rough surfaces show randomly distributed errors in the mass percentage results. No discernible patterns have emerged to compensate for the errors. Field use of the PXRF to identify or correlate basalt in outcrops therefore appears to be impractical. Analysis using cut and polished samples is required for accurate measurements. Geochemical identification using PXRF has proved to be helpful for identifying potential basalt aquifers and landslide blocks in eastern Washington.