Author
 |
SCHWANK, MIKE - ETH-ZENTRUM |
|
Green, Timothy |
|
Submitted to: Proceedings of the International Soil Moisture Technology Conference
Publication Type: Abstract Only Publication Acceptance Date: 3/1/2007 Publication Date: 3/20/2007 Citation: Schwank, M., Green, T.R. 2007. Laboratory and Theoretical Investigations of a Ring-Capacitor Sensor. Proceedings of the International Soil Moisture Technology Conference. Honolulu, HI. March 20,2007. Interpretive Summary: Ring-capacitor sensors are used widely for real-time estimation of volumetric soil water content from measured resonant frequency (f) which is affected by the complex bulk soil permittivity. However, further investigation is required regarding the dependence of f on the real and the imaginary parts of permittivity. Laboratory experiments were performed to characterize the response of a Sentek EnviroSCANTM capacitance sensor over a full range of permittivity values from air to water and over a range of temperatures. The frequency responses were recorded leading to a new empirical relationship between normalized readings and real permittivity. Furthermore, total capacitances were measured using a vector network analyzer to explore the radius of sensitivity to a dielectric disturbance. Additional sensitivities of the sensor readings with respect to liquid salinity S and soil temperature T were investigated theoretically. The new model is based on an electrical circuit analogue approach in conjunction with a dielectric mixing model and a finite element model to compute electrical field distributions. Modeled sensor readings display sensitivities to both T and S, and the temperature gradients can go from negative to positive values. Finally, the results from the theoretical sensitivity study were combined with the empirical relationship to quantify the temperature effect on the permittivity values at different salinity and moisture levels. Temperature variations in the range 0 < T < 40 oC caused deviations in measured permittivities within a range of about ' 10 %. The combined experimental and simulation results improved our generic understanding of processes affecting instrumental sensitivities of capacitance sensors. The results are crucial for scientific applications of these sensor types used in natural soils, and the work provides a foundation for further inference of soil water content under field conditions. Technical Abstract: Ring-capacitor sensors are used widely for real-time estimation of volumetric soil water content ' from measured resonant frequency fr which is affected by the complex bulk soil permittivity ' = '’ + i'’’. However, the relationship fr(') requires improved investigation in terms of the dependence of fr on the real and the imaginary parts '’ and '’’. Electrical losses relevant in saline soils have high values of '’’, and soil moisture is correlated with both '’ and '’’. Laboratory experiments under nearly lossless conditions were performed to characterize the response fr('’) of a Sentek EnviroSCANTM capacitance sensor over a full range of '’ values from air to water and over a range of temperatures. The responses fr('’) were recorded while encompassing the sensor with water-dioxane mixtures at different mixing ratios. This led to a new empirical relationship N('’) between normalized readings and real permittivity '’. Furthermore, total capacitances Ctot were measured using a vector network analyzer. The readings were sensitive to changes in the dielectric properties within given annular distances. Additional sensitivities of the sensor readings with respect to '’’ via liquid salinity S and soil temperature T were investigated theoretically. The new model is based on an electrical circuit analogue approach in conjunction with a dielectric mixing model and a finite element model of the steady Maxwell’s equation to compute electrical field distributions E. The mixing approach estimates ' = '’ + i'’’ of solid-water-air mixture representing moist soil. Different values of '’ were used in simulations for computing E, from which the capacitive circuit analogue components were computed. Circuit resistances representing the electrical conductivity losses were calculated from '’’ and E. The circuit analogue was used to model fr('’, '’’) in an iterative manner to account for the frequency-dependent values of the circuit analogue components. Modeled sensor readings fr and N display sensitivities to both T and S, and the temperature gradients can go from negative to positive values. Finally, the results from the theoretical sensitivity study were combined with the empirical relationship N('’) to quantify the temperature effect on the permittivity values at different salinity and moisture levels. Temperature variations in the range 0 < T < 40 oC caused deviations in measured permittivities within a range of about ' 10 %. Although the investigation has been done with a specific sensor, the combined experimental and simulation results improved our generic understanding of processes affecting instrumental sensitivities of capacitance sensors. The results are crucial for scientific applications of these sensor types used in natural soils, and the work provides a foundation for further inference of soil water content under field conditions. |
