Abstract
Continuous sensor measurements are becoming an important tool in environmental monitoring. However, the reliability of field measurements is still too often unknown, evaluated only through comparisons with laboratory methods or based on sometimes unrealistic information from the measuring device manufacturers. A water turbidity measurement system with automatic reference sample measurement and measurement uncertainty estimation was constructed and operated in laboratory conditions to test an approach that utilizes validation and quality control data for automatic measurement uncertainty estimation. Using validation and quality control data for measurement uncertainty estimation is a common practice in laboratories and, if applied to field measurements, could be a way to enhance the usability of field sensor measurements. The measurement system investigated performed replicate measurements of turbidity in river water and measured synthetic turbidity reference solutions at given intervals during the testing period. Measurement uncertainties were calculated for the results using AutoMUkit software and uncertainties were attached to appropriate results. The measurement results correlated well (R2 = 0.99) with laboratory results and the calculated measurement uncertainties were 0.8-2.1 formazin nephelometric units (FNU) (k = 2) for 1.2-5 FNU range and 11-27% (k = 2) for 5-40 FNU range. The measurement uncertainty estimation settings (such as measurement range selected and a number of replicates) provided by the user have a significant effect on the calculated measurement uncertainties. More research is needed especially on finding suitable measurement uncertainty estimation intervals for different field conditions. The approach presented is also applicable for other online measurements besides turbidity within limits set by available measurement devices and stable reference solutions. Potentially interesting areas of application could be the measurement of conductivity, pH, chemical oxygen demand (COD)/total organic carbon (TOC), or metals.