Evaluation of apparent surface reflectance estimation methodologies

Abstract The estimation of apparent surface reflectance values from imaging spectroscopy data requires a correction for the efTects of the intervening atmosphere. Four methods of estimating apparent surface reflectance have been evaluated, the empirical line method and three methods of radiative transfer modelling. To compare the results of these correction methodologies two high albedo targets, of identified composition, were selected. The empirical line method was found to be sensitive to errors in locating and spectral variations within the ground survey targets and also target height differences. The radiative transfer modelling techniques gave relatively similar results, reasonably close to the library spectra. This study indicates that radiative transfer modelling using only atmospheric information derived from the imaging spectroscopy data, while still not as sensitive as correction methods using additional ground and atmospheric information, can satisfactorily correct the atmospheric elfects involved in estimating apparent surface reflectance, allowing the identification of the major diagnostic absorption features.