Investigation And Application Of Atmospheric Correction Based On6S Radiative Transfer Model

With the development of remote sensing, the demand for quantitative remote sensing growing greatly and atmospheric correction become more important recently. The remote sensing imaging process is vulnerable to the influence of atmospheric absorption and scattering, so how to remove the impact of the atmosphere is the key to atmospheric radiation correction.6S radiative transfer model can accurately simulate the interaction process between electromagnetic waves and atmosphere, which has been widely used in the field of atmospheric correction after years of development. However, the calculation accuracy of the model is affected by the algorithm used in the model and the input atmospheric parameters. Therefore, this work develops from atmospheric absorption and scattering two aspects. First, the algorithm accuracy of atmospheric absorption in6S model will be assessed. Then the domestic regional aerosol model will be analyzed in order to point out the differences between domestic aerosol and predefined aerosol models in6S. Finally, the effects of the predefined aerosol model in6S will be studied aimed at providing reference for related research. The main work of this paper is as follows:1) The calculation accuracy of Malkmus model in6S model is low, to solve the problem, a modified Malkmus band model for transmittance calculation was proposed, which improved the accuracy about by1-2magnitude. When assessing the impact of transmittance calculation by6S for atmospheric correction, the results showed that the model affects the visible and near-infrared bands slightly with an average transmittance relative error of less than4%, while in the mid-infrared at3.5-3.9um, the relative error up to80%.2) Currently, atmospheric correction based on radiative transfer models are direct using the predefined aerosol model, without taking into account the differences between the predefined model and the actual aerosol model. To solve this problem, data of9AERONET stations in China was analyzed. The results show that northern areas susceptible to dust aerosols activities, while central and southern regions seldom affected by dust aerosols. The content of absorbent aerosols in the urban area is really high than in the non-urban area and the aerosol variations seasonally. In addition, the single scattering albedo and asymmetry factor of the predefined aerosol model and measured aerosol of AERONET are different.3) To validate the impacts of predefined aerosol model on atmospheric correction, different aerosol models are used for Landsat TM5image atmospheric correction with other atmospheric conditions be constant. Experimental results show that the predefined aerosol models for atmospheric correction in China may lead to errors.