Study On Microwave Dielectric Property Of Coastal Soil

Remote sensing is the most important method to obtain the natural resources and environmental information of earth in21th century, meanwhile, it is one of the eight emphatic strategy sciences in the period of the eleventh five years in China. With the development of air-bone and space-bone microwave data source and the quality of the data, how to extract useful surface features from remote sensing data became the key work in microwave application.Soil microwave dielectric property is one of the most important content in study of microwave scattering mechanization. It is the basement of soil backscattering models and electromagnetism parameter reversion. Thus in this paper, the soil dielectric properties are studied from the theory and experiment aspects. Firstly, the traditional methods and tools are summarized to explain the technical framework of obtaining surface feature parameters by remote sensing. Then a series of studies of soil microwave dielectric properties as shown below are completed:.1) the soil dielectric models are listed and simulated to compute the relationship between the dielectric constant and frequency, moisture content, saline content, temperature, etc.2) the dielectric constant of coastal soil is measured by coaxial probe. The differences between the measured and simulated values and the factors leading to these differences are analyzed.3) the availability of the existing models for the coastal soil in our study area is investigated.The results of this study showed that:1) the responses of the soil sampled in our study area to microwave accord with the facts, but some small differences exist for the soil type.2) as a simple empirical model, Topp model is available for coastal saline soil with a good coincidence between measured and computed values.3)Dobson semi-empirical model has preferable availability in the real part, but has certain degree differences in the imagery part.4) the values of real part computed by Hallikainen model has good coincidence with the measured values for all the bands used in the experiment, but the imagery part only presents good coincidence at high frequency. Therefore these models only can be use for macroscopical dynamic monitor of soil moisture content at large area and must be improved for the soil saline content investigation. The improved Dobson model shows a good result in retrieving of the imagery part.