There is an increased trend toward quantitativeestimation of land surface variables from hyperspectral remotesensing. One challenging issue is retrieving surface reflectancespectra from observed radiance through atmospheric correction,most methods for which are intended to correct water vapor andother absorbing gases. In this letter, methods for correcting bothaerosols and water vapor are explored. We first apply the clustermatching technique developed earlier for Landsat-7 ETM+imagery to Airborne Visible/Infrared Imaging Spectrometer(AVIRIS) data, then improve its aerosol estimation and incorporatea new method for estimating column water vapor contentusing the neural network technique. The improved algorithmis then used to correct Hyperion imagery. Case studies usingAVIRIS and Hyperion images demonstrate that both the originaland improved methods are very effective to remove heterogeneousatmospheric effects and recover surface reflectance spectra.
展开▼
