Uncertainty-based measurement quality control

Optimal estimator for uncertainty-based measurement quality control

“…The purpose of setting a limit in the form of a MPU is to ensure that the measured value has an acceptable quality (precision) with reasonable measurement efforts. A low value of MPU may require highly accurate instruments and/or more observations, meaning high investment or high observation cost; it may also result in a high false rejection probability for small samples [3]. On the other hand, a high value of MPU may result in less precise measurement results; however, the false rejection probability associated with a high MPU may be low for small samples [3].…”

“…According to a simple acceptance decision rule, a measured value will be accepted if the expanded uncertainty U at a nominal coverage probability 1 -a (assumed to be 95 % in this paper) is not greater than a present maximum permissible uncertainty (MPU): U B MPU. Otherwise, the measured value will be rejected [3]. The purpose of setting a limit in the form of a MPU is to ensure that the measured value has an acceptable quality (precision) with reasonable measurement efforts.…”

“…An ADCP should be subject to its own CA to meet its velocity and depth measurement specifications. The quality of ADCP streamflow measurements is measured by the uncertainty of the transect discharge measurements and is controlled by a preset relative MPU [3].…”

“…The decision made based on U z is deterministic and assumed to be correct [3]. Thus, U z is considered as the ''gold standard'' for uncertainty estimation.…”

“…However, the decision made based onÛ will be uncertain. The risk in the decision making (i.e., the acceptance or rejection of a measured value) can be measured by the false acceptance probability or false rejection probability, which depends on the uncertainty estimator used [3].…”