DERIVATION OF SCALED SURFACE REFLECTANCES FROM AVIRIS DATA

Analysis of high resolution imaging spectrometer data requires a thorough compensation for atmospheric absorption and scattering. A method for retrieving ''scaled surface reflectances,'' assuming horizontal surfaces having Lambertian reflectances, from spectral data collected by the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) (Vane, 1987) is presented here. In this method, the integrated water vapor amount on a pixel by pixel basis is derived from the 0.94-mum and 1.1 4-mum water vapor absorption features. The transmission spectrum Of water vapor (H2O), carbon dioxide (CO2), ozone (O3), nitrous oxide (N2O), carbon monoxide (CO), methane (CH4), and oxygen (O2) in the 0.4-2.5 mum region is simulated based on the derived water vapor value, the solar and the observational geometry, and through use of narrow band spectral models. The scattering effect due to atmospheric molecules and aerosols is modeled with the 5S computer code (Tanre et al., 1986). The AVIRIS radiances are divided by solar irradiances above the atmosphere to obtain the apparent reflectances. The scaled surface reflectances are derived from the apparent reflectances using the simulated atmospheric gaseous transmittances and the simulated molecular and aerosol scattering data. The scaled surface reflectances differ from the real surface reflectances by a multiplicative factor. In order to convert the scaled surface reflectances into real surface reflectances, the slopes and aspects of the surfaces must be known. For simplicity, the scaled surface reflectance is simply referred to as the ''surface reflectance'' in this article. The method described here is most applicablefor deriving surface reflectances from AVIRIS data acquired on clear days with visibilities 20 km or greater. More rigorous atmospheric radiative tranfer modeling is required in order to derive surface reflectances from AVIRIS data measured on hazy days.