Biological Systems Engineering, Department of


Document Type


Date of this Version



Published in Agricultural Water Management 231 (2020), 105984.

doi: 10.1016/j.agwat.2019.105984


Copyright © 2019 Elsevier B.V. Used by permission.


Interreplicate variability—the spread in output values among units of the same sensor subjected to essentially the same condition—can be a major source of uncertainty in sensor data. To investigate the interreplicate variability among eight electromagnetic soil moisture sensors through a field study, eight units of TDR315, CS616, CS655, HydraProbe2, EC5, 5TE, and Teros12 were installed at a depth of 0.30 m within 3 m of each other, whereas three units of AquaSpy Vector Probe were installed within 3 m of each other. The magnitude of interreplicate variability in volumetric water content (θv) was generally similar between a static period near field capacity and a dynamic period of 85 consecutive days in the growing season. However, a wider range of variability was observed during the dynamic period primarily because interreplicate variability in θv increased sharply whenever infiltrated rainfall reached the sensor depth. Interreplicate variability for most sensors was thus smaller if comparing θv changes over several days that excluded this phenomenon than if comparing θv directly. Among the sensors that also reported temperature and/or apparent electrical conductivity, the sensors exhibiting the largest interreplicate variability in these outputs were characterized by units with consistently above or below average readings. Although manufacturers may continue to improve the technology in and the quality control of soil moisture sensors, users would still benefit from paying greater attention to interreplicate variability and adopting strategies to mitigate the consequences of interreplicate variability.