Annotation:
A new approach to improving the type A uncertainty evaluation by cleaning of the collected data from unwanted influences which appears as non-periodical and periodical components identified in the data is presented in the paper. The approach refers to regularly in time sampled data. The non-periodical components are equivalent to trends while the periodical components are a type of disturbances of unknown a priori period. The cleaning process comply with the main stream of ISO GUM recommendation and can be recognized as good practice in uncertainty evaluation as the elimination of the influence like identified drift and periodic components are resulting in better approximation of the type A uncertainty. The proposed approach is discussed in the paper and the numerical result is presented as well