The Study Of Inversion Of The Properties Of Sandy Land Basing On The Multiple-angle And Hyperspectral Reflectance

The sandy land system is significant for the ecology research of arid and semi-arid region because of its considerable productivity, complex biodiverse and abundant biomass. Unfortunately, the sandy land creates a negative feedback for the environment because the human absurd produce activities are increasing all the time, which lead to the degeneration of ecological functions cased by desertification. Thus, it is important for dynamically monitoring the change of sandy land area and the energy exchange between sandy land surface and atmosphere, these information are valuable for scientifically assessing the ecosystem recovery of sandy land and the radiation balance in the whole earth-atmosphere system. The data mentioned above can be received by remote sensing technique, it may provide dynamical, real time and wide image for the land surfaces, so that there are enough information for the research of the distribution of sandy land in space and the physical properties of sandy land in the worldwide scale.The primary basis of remote sensing technique on the sandy land surface study is the radiation response between the incident wave and sandy land surface, following the observation, treatment and analysis process to ensure the properties of sandy land, and draw on the real study of sandy land in final.As to the passive remote sensing, the main energy resource is the Solar radiation,the change of reflectance caused by the sandy land surface(such as: particle size,surface texture, moisture) is our only evidence for utilizing remote sensing to estimate the properties of sandy land. Until the hyperspectral and multi-angular remote sensing are available, the researchers can make a more accurate reversion for the properties and albedo of sandy land and comparing with traditional method. Meanwhile, there must be enough studies for understanding the hyperspectral reflectance and reflectance distribution, so that the inversion accuracy will be able to be used in the piratical application.In this paper, we focus on three points: first the bidirectional reflectance model which is basing on the radiative transfer theory has been analyzed, and discussed the physical implication of model. Moreover, we also assessed the application of the model on calculating the reflectance characteristic of sandy land basing on the laboratory and field measurements. Secondly, the comparison between modeled results andmeasurements results is used to ensure if the parameters of the model are suitable for describing the particle size, surface texture and moisture. For the study of reversing particle size by the reflectance, the semi-variance was used to quantitative the particle size of sandy land surfaces, then the relationship between the anisotropy measurements values and the particle size was established; the effect of particle size on the polarization was also included. Finally, we not only improved and verified the current empirical model, but also made use of the parameters of model for estimating the surface texture and moisture of sandy land basing on the multi-angular and hyperspectral reflectance in the field campaign. The results suggested that the reflectance, which was measured in the laboratory and field, combined with the parameters of model can be directly used for the reversion of sandy land properties.For inversing the particle size, the semi-variance is suitable for quantifying sandy land surface spatial characterization, and then the relationship between anisotropy measurements values and the semi-variance was used to estimate the particle size.Because the anisotropy measurements value is a ratio, we can eliminate the influence of measurement surrounding on the spectral reflectance. The method mentioned above will be an effective way for inversing particle size on the surface of sandy land.Moreover, the radiative transfer model can also inverse the particle size with a high accuracy. However, overfull parameters may lead to the unclear physical meaning of the model, and the model is semi-empirical in practice. Then, comparing the model value with measurements, we can find that the model is suitable for characterizing the distribution of reflectance from sandy land, but the errors are obvious in the forward scattering and backward scattering direction. In this paper, the relationship between the particle size and polarization was also found, which was a clear negative correlation for them. The polarization is required if one want to completely describe the scattering properties of sandy land surface and is useful for estimating the particle size of sandy land.For reversing the moisture, we can directly use the reflectance, because the relationship between the moisture and reflectance is linear and the correlation coefficient is high. By the way, it was found that there is an indirect way for quantifying the moisture basing on the bidirectional reflectance model, and the inversion accuracy was enough. Although the bidirectional reflectance model is available in the reversion of sandy particle size and sandy moisture, there should be a caution for reversing the surface texture due to the distribution of modeled reflectance is similar for all the samples. But the changes of surface texture yielded the inherentreflectance distribution in practice measurements. However, the parameters of model were also fitted to qualify the change of sandy land surface texture.Overall, the study of multi-angular and hyperspectral reflectance performed on the different characterization of sandy land is valuable for the researchers completely understand the radiation response characteristics between sandy land surface and electromagnetic wave. We combined the multi-angular information, spectral information with sandy land properties, modified the effectiveness of the current model and analyzed the relationship between the reflectance and particle size, moisture and surface texture. These results are not only important references for accurately estimating the sandy land surface using remote sensing technique, but also are useful for the development of the science correlated sandy land, and provide foundation for solving the global climate problem.