abstract: Determination of water content in porous materials, especially soil, is one of the main objectives of agrophysical metrology because of the importance of water in the majority of physical and chemical processes taking place in the soil. The development of Time Domain Reflectometry method for the determination of the bulk dielectric permittivity of soil enabled to determine indirectly its water content in fast, accurate, non-destructive and automatic way.
The increase of accuracy of reflectometric water content measurements requires further research and development of specialized measurement equipment to eliminate or correct the errors coming from the variability of soil material.
The objective of the presented study is to evaluate the influence of temperature on the bulk dielectric permittivity, b, of soil and the water content measurement error done by reflectometric method.
The presented objective is achieved by identification the reasons of the tem¬perature effect on b and its correction based on the performed measurements. The existing models based on the hypothesis of two opposite physical mecha¬nisms influencing the temperature effect on b of the soil were verified. One of them is the release of water particles bound to soil solids accompanied with the increase of their kinetic energy with the soil temperature increase. It causes the increase of the effective soil dielectric permittivity, which reflects the increase of the bulk dielectric permittivity determined by TDR. The other mechanism concerns the decrease of free water particles dielectric permittivity decrease with temperature increase.
The results of the research as well as the theoretical considerations prove that the soil electrical conductivity does not influence the observed temperature effect in the range below 2 dS•m-1. This confirms the results of earlier research con¬ducted by the author and the literature reports.
On the base of the collected TDR calibration curves for the tested mineral soils in various temperature it is concluded that the temperature effect of b is different for all the tested soils and it depends on water content. The dominant role play free water particles of the soil. The value of this effect may significantly affect the reflectometric soil water content measurement by the introduction 4% error referenced to the measured value. Additionally, the temperature effect of b is connected with the soil specific surface area, where the water particles are adsorbed and their mobility in the alternating electromagnetic field is limited, which decreases the resultant value of b measured by TDR meter. The increase of soil temperature causes the increase of the particles kinetic energy and after releasing from the solids they become more mobile in the external electromagnetic field and they increase the effective value of b of the soil.
The original scientific achievement coming from the investigations, particu¬larly from the analysis of the TDR calibration curves of the tested soils at various temperature was the discovery of the defined and unique value of soil water content, named the equilibrium water content, eq, where the temperature effect of the bulk dielectric permittivity minimizes. This means that the both opposite physical phenomena influencing the temperature change of the soil b are bal¬anced. For the soils with the water content below the value of eq the temperature effect of b is positive, i.e. b increases with the temperature increase, and for the soil with the water content above eq the temperature effect of b is negative, i.e. b decreases with temperature increase. It is shown that the calculated values of eq depend on soil specific surface of the tested soils. The equilibrium water content, eq, which is unique for each soil, may be used for dielectric determina¬tion of soil specific surface area.
The applied physical dielectric mixing models do not always produce data in agreement with the experimental values. The elements of the models responsible for the temperature effect of b require additional analysis because their verifica¬tion was negative. Correction of the temperature effect of b based on these mod¬els is not satisfactory, which is caused by the incompatibility of these models with the experimentally produced data.
The other correction of the temperature effect is empirical and it incorporates the trend lines fitted to the reflectometrically determined values of . This correction takes into account the values of the equilibrium water content, eq. For the soil water content below eq, the b increase caused by the release of bound water with temperature from the solids is corrected. For the soil water content above eq, the b decrease caused by the temperature effect of free water is cor¬rected. The absolute measurement error of soil water content determination by reflectometric method, resulting from the temperature effect of b, decreases almost three times after application of the empirical correction.
The performed research on the temperature effect of b involve the design and construction of the set of intelligent reflectometric sensors integrated with the sensors of soil electrical conductivity and temperature as well as a dedicated measurement system. The integration of the water content sensor with sensors of other physical and chemical properties of porous materials soil seems to be the future direction of agrophysical metrology development and the initiated work in this field will continue in the IA PAS, Lublin.