Measurement bias is an estimate of a systematic measurement error.1

The VIM defines a systematic measurement error as the component of measurement error that stays constant or varies in a predictable manner across replicate measurements. Bias is the estimate of that component, so it is a quantity derived from a reference and not an observed true error.1

Uncertainty components from systematic effects may be applied as corrections or, where an estimated effect is not corrected, included in the uncertainty. Averaging replicate readings does not remove a constant systematic error, which is why bias is treated separately from random scatter and from precision.1

In Earth observation, calibration is described as defining a system’s response to known and controlled inputs, and validation as independent assessment of derived products. These are the activities in which a bias would be found, though the CEOS page cited does not define bias itself.2

Footnotes

  1. BIPM and JCGM, International vocabulary of metrology: Basic and general concepts and associated terms (VIM), 3rd edition, JCGM 200:2012. https://www.bipm.org/documents/20126/2071204/JCGM_200_2012.pdf ↩ ↩2 ↩3

  2. CEOS, “Working Group on Calibration & Validation (WGCV)”. https://ceos.org/ourwork/workinggroups/wgcv/ ↩