AML Oceanographic discusses the principles behind sensor accuracy in marine instrumentation, outlining how accuracy is defined, measured and reported for oceanographic sensors. Read more >>
Drawing on terminology from the International Vocabulary of Metrology (VIM3) and ISO 5725, the company explains that accuracy refers to the closeness of agreement between a measurement and a reference standard, while also distinguishing the related concepts of precision and trueness. Precision describes the consistency between repeated measurements, whereas trueness refers to how closely the average of those measurements aligns with a known reference value.
The article examines how different sources of error influence sensor performance in marine environments. Random effects such as thermal noise in electronic circuits are identified as examples of precision error, while systematic influences including biofouling on conductivity sensors are described as sources of bias that can progressively alter readings during deployments. AML Oceanographic also explains how standard deviation, or root mean square (RMS) error, is used to quantify precision error, noting that the company reports accuracy values at the 1σ level in accordance with statistical convention.
AML further details the concept of initial accuracy, defined as the condition immediately following calibration when sensor bias remains smaller than the associated precision error. At this stage, the precision component dominates the reported accuracy value, which is why AML lists this figure in its datasheets as calibration or initial accuracy. The company also distinguishes between repeatability, representing the precision of a single sensor, and reproducibility, which evaluates variation across multiple sensors of the same model.
To determine sensor accuracy, AML Oceanographic conducts testing against reference standards under controlled conditions using at least five sensors of the same model and a minimum of 15 repeated measurements at minimum, midpoint and maximum calibration levels. The resulting reproducibility values are reported as sensor accuracy in company datasheets, providing users with a representation of the variability they may observe when comparing sensors during field operations.
To find out more information, read ‘Five Questions about Sensor Accuracy, Answered’ here >>
