Evaluation of soil-water and salt content based on IEM in arid areas

Although the relations of backscattering coefficients and soil-water and salt content are the real problems about inversion, since the relation between soil dielectric constants and soil-water and salt is relatively clear 17,81, the relation of backscattering coefficients and dielectric constants is the key problem. In addition, a lot of physical geographical features in and areas, such as the ground surface with sparse vegetation, little soil water in dry seasons, are useful to retrieve physical parameters of the surface. According to these features, using IEM model, which was developed by Fung et al in 1992 [11] generated simulated backscattering coefficients in and areas. These data included H-H and VV polarizations and different seasons (wet and dry seasons) data. Different seasons data were achieved by changing dielectric constant. By analyzing these data, we found: in non-salinized areas, 1) there is a strong correlation IT, (> 0.9) between sigma(0)(wet) - sigma(0)(dry) (the difference between the dry- and wet-season SAR sigma(0)) and epsilon(wet) - epsilon(dry) (the difference between wet and dry seasons surface dielectric constants); and 2) dry-season epsilon(dry) (soil water) has little effect on the relation of sigma(0)(wet) - sigma(0)(dry) and epsilon(wet) - epsilon(dry), which ensures there is no need of specific information about dry-season soil water in inversion of wet-scason soil water. In salinized areas, 3) there is a good correlation between sigma(0)(wet) - sigma(0)(dry) and I(epsilon(wet)) - I(epsilon(dry)) (the difference between wet and dry seasons imaginary part of dielectric constants) that is related with soil salt content when is great. This may provide a way to monitor salinization although quantitative inversion is not feasible for imaginary part of dielectric constant is related with both soil water an salt content, it is hard to separate their effects.