Estimation of Directional Vegetation Fraction Cover from TOA Spectral Data of AATSR

Among the key parameters acquired by remote sensing inversion, vegetation fraction cover is one of the crucial variables. Component temperatures inversion and leaf area index (LAI) inversion all have close relations with the vegetation fraction cover. The objective of this study is to develop a method to estimate the vegetation cover fraction from satellite observation. Traditional methods of inferring vegetation fraction cover from satellite remote sensing include spectral mixture analysis (SMA) and scaled normalized difference vegetation index (NDVI). Those methods often rely on a series of steps in the processing chain, including atmospheric correction, surface angular correction and so on. Generally, those procedures are very computationally demanding. In addition, the errors associated with each procedure may be accumulated and significantly affect to the accuracy of the final products. In this study, a new retrieval methodology is proposed to calculate vegetation fraction cover over mixed pixels directly from the AATSR spectral reflectance data at top-of-atmosphere (TOA). The method consists of extensive radiative transfer simulations under a wide variety of solar illumination and sensor view conditions, atmospheric profiles, aerosol types and conditions and vegetation canopy leaf angle distributions. The derivation of vegetation fraction cover from TOA observations requires several steps of processing.Important steps include, (1) Preparing for the model input variables: foliage and soil spectral data on red, green and near-infrared bands which measured from two kinds of vegetation and three kinds of soil in the field experiment; (2) Generating a database based on a canopy radiative transfer model, Scattering by Arbitrarily Inclined Leaves (SAIL), and a hybrid linear model with the spectral data combined with vegetation geometric construction data and observation geometric data; (3)Atmospheric correction that converts surface ensemble reflectance to TOA ensemb- - le reflectance based on a radiative transfer model, Second Simulation of the Satellite Signal in the Solar (6S); (4) Mapping the relationships between spectral directional ensemble reflectance and vegetation fraction cover through a nonlinear regress method. The correction coefficients of the surface vegetation fraction cover computed with AATSR are provided.vegetation fraction cover retrieval from TOA data of AATSR does not exceed by 6% at nadir view and 9.7% at forward view, respectively. The performances of input parameters on estimates of vegetation fraction cover are given compared with the "true" surface vegetation fraction cover. The aim of estimating vegetation fraction cover is to prepare for inversing component temperatures using AATSR data, in which process vegetation fraction cover is an important parameter.