Submitted to: Vadose Zone Journal
Publication Type: Peer reviewed journal
Publication Acceptance Date: 9/20/2004
Publication Date: 9/21/2005
Citation: Seyfried, M.S., Grant, L.E., Du, E., and Humes, K., Dielectric loss and calibration of the Hydra Probe soil water sensor. Vadose Zone Journal, 4:1070-1079, 2005 Interpretive Summary: Relatively recent advances in electronic technology have made it possible to monitor soil water content continuously. Data from this monitoring is being used to improve water management of golf courses, race tracks, watersheds and agricultural fields in addition to scientific research. A large number of soil water sensors have recently been developed for this purpose. Unfortunately, these sensors have not received independent testing and calibration. We developed two new calibration approaches for the Hydra Probe soil water sensor, one that improves sensor performance for soils in general, and another that accounts for specific soil properties that can be applied when those properties are known by the user. These findings are currently being incorporated by the manufacturer in a newer version of the instrument. These calibration approaches will be used by the Natural Resource Conservation Service, which currently uses the Hydra Probe in its SCAN (Soil Climate Analysis Network) and SNOTEL monitoring networks.
Technical Abstract: Widespread interest in soil water content (', m3m-3) information for both management and research has led to the development of a variety of soil water content sensors. The accuracy of these sensors depends partly on the quality of the calibration equation used. In most cases, manufacturer supplied calibrations have received little or no independent study. We investigated the calibration of the Hydra Probe soil water sensor, which is unusual in that it measures both the imaginary ('i) and real ('r) components of the complex soil dielectric constant. Our objectives were to: (i) quantify the inter-sensor variability, (ii) develop and test an optimal, general calibration equation and (iii) incorporate the effects of soil texture and/or 'i on the 'r(') calibration. We found that the inter-sensor variability measured in fluids was generally low. A general calibration based on mean soil calibration parameters, and a loss-corrected calibration that incorporates 'i, were developed using data from 20 highly variable soils. Soil texture had little effect on the 'r(') calibration. These calibrations were compared to manufacturer supplied calibrations on the 20-soil data set and an additional, independently collected, four-soil data set. Results from both data sets were similar and indicated that the calibrations developed in this study represent a substantial improvement over the manufacturer supplied calibrations. Calibration errors, expressed as an average ' difference across all soils, were 0.019 m3m-3 for the general calibration and 0.013 m3m-3 for the loss-corrected calibration. Similar performance with the independent data set supports the general applicability of these equations.