Open AccessTraceable measurements and calibration: a primer on uncertainty analysis
Author(s) -
Csavina Janae,
Roberti Joshua A.,
Taylor Jeffrey R.,
Loescher Henry W.
Publication year - 2017
Publication title -
ecosphere
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.255
H-Index - 57
ISSN - 2150-8925
DOI - 10.1002/ecs2.1683
Subject(s) - measurement uncertainty , calibration , uncertainty analysis , transparency (behavior) , repeatability , computer science , confidence interval , quality (philosophy) , propagation of uncertainty , data mining , environmental science , statistics , reliability engineering , mathematics , engineering , simulation , algorithm , philosophy , computer security , epistemology
Describing the quality of measurements is necessary to understand the level of confidence in any observation. Accuracy, precision, trueness, repeatability, reproducibility, and uncertainty are all used to describe quality of measurement, but the terms are inconsistently defined and measured and thus easily misunderstood. One purpose of quality parameters is for the comparison of observations, but when dissimilar methods for estimating quality terms are utilized, a comparison is misrepresented. A standardized approach to estimating uncertainty provides a basis for meeting measurement requirements and providing a level of confidence for observations. Here, we show the approach used by the National Ecological Observatory Network to estimate uncertainty of the calibration processes and measurements illustrated with an example of uncertainty assessment on a temperature sensor. Detailing the approach for uncertainty assessment provides the transparency necessary for network science and allows for the approach to be adopted in the scientific community. Reporting uncertainty with all measurements needs to become consistent and commonplace across disciplines.