An improved dense dark vegetation based algorithm for aerosol optical thickness retrieval from hyperspectral data

Aerosol optical thickness is a very important parameters in the atmospheric correction of the hyperspectral data. In this study, an improved dense dark vegetation (DDV) based algorithm is introduced to estimate the AOT@550nm from hyperspectral remote sensing data. A correction relationship between TOA and land surface reflectance at short wavelength near 2.13μm was introduced in order to reduce the assumption of the traditional DDV that the TOA reflectance is equal to the land surface reflectance at short wavelength near 2.13μm. Simulated hyperspectral data of Hyperion sensor were applied to the improved DDV algorithm. The retrieved AOT @550nm show a well correlation with the actual values and the correlation coefficients is larger than 0.99.