Generalized refractive mixing dielectric model for moist soils

Abstract

This paper introduces a new soil dielectric model in the microwave band offering variation over a wide range of soil moisture, texture, mineral content, and wave frequency. The soil dielectric models commonly used in microwave remote sensing algorithms rarely apply intrinsic bound soil water (BSW) dielectric properties. However, models incorporating these properties could offer additional applicability and accuracy. This paper describes a technique developed for estimating the complex dielectric constant (CDC) for both the BSW and the free soil water (FSW) as separate water constituents present in moist soil. For this purpose, it has been shown that there exist two specific soil moisture regions where alteration in soil dielectric property is generated, exclusively, by either BSW or FSW increments, with their CDCs being constant values. The CDCs of the BSW and FSW are derived through straight line fitting with the help of a measured soil complex refractive index (CRI) as a function of moisture. The term CRI is understood here as a square root of the CDC. In order to distinguish between the two moisture regions, the maximum bound water fraction (MBWF) parameter was introduced, which also can be obtained through straight line fitting, using measured CRIs as a function of moisture over both specific moisture regions. Thus, obtained fittings are known in the literature as the refractive mixing dielectric model (RMDM). Finally, a method for deducing the Debye relaxation parameters for both types of soil water was proposed, with their CDCs being input parameters. It was shown that the soil water CDCs measured only at two frequencies are sufficient for deriving the Debye relaxation parameters. A set of equations representing the RMDM, the Debye formula, and the formulas proposed for deriving the Debye relaxation parameters is designated as the generalized refractive mixing dielectric model (GRMDM). In contrast to all previously known soil dielectric models, the GRMDM allows for CDC prediction beyond the frequency range where the initial soil CDC data are acquired.