Measurement standards are organized in levels. At the top is the realization of the unit according to its definition; below it are standards increasingly close to everyday use, with growing uncertainty. Each level is calibrated against the one above, and that relationship forms the chain that underpins traceability.
The usual levels
- International standard: agreed by international convention and used as the basis for assigning values to others.
- National standard: officially recognized in a country; it is the reference for the other standards of that quantity.
- Primary standard: established with a method of the highest metrological quality, without depending on another standard of the same quantity.
- Secondary standard: its value is assigned by comparison with a primary standard.
- Reference standard: the highest-quality standard available at a location or organization, from which the measurements made there are derived.
- Working standard: used routinely to calibrate or verify instruments.
Why uncertainty grows on the way down
Each calibration incorporates the uncertainty of the higher standard and adds that of the comparison itself. This is why a working standard can never have a better uncertainty than the standard it was calibrated with. It is a useful rule of thumb: a standard chosen to calibrate an instrument must be clearly better than that instrument, not equal to it.
What the engineer sees in practice
A certificate shows which standards were used and where they come from. If a shop calibrates its pressure gauges with a reference gauge, that reference gauge in turn has its own certificate, and so on. Reviewing this chain avoids accepting a standard whose uncertainty is too high for the task.
When there is no clear hierarchy
For some complex quantities, such as certain material properties, the hierarchy relies on reference materials and interlaboratory comparisons rather than on a single physical standard. The principle is the same: a common reference and documented comparisons.