Comparison of different approaches of soil sampling uncertainty determination.

The ISO/IEC 17025 standard requires that all significant contributions have to be propagated to the measurement uncertainty, including also sampling uncertainties. We evaluated soil sampling uncertainties for gamma-ray spectrometry by using in-house and split-sample methods. By in-house method, the sampling uncertainty was determined by comparing standard deviations of measurement results and average analytical uncertainty. With split-sample method, it was calculated using between-sample and between-analysis variances. If analytical uncertainty is reliably determined, the in-house method is recommended because fewer measurements are needed.