An Airborne Imaging Spectrometer for Applications (AISA) hyperspectral imager was deployed on a manned aircraft flown at 1305-m altitude to collect data over optically shallow waters in the Florida Keys with the ultimate goal of mapping water quality and benthic habitats. As a first step, we developed a practical atmospheric correction (AC) approach to derive surface remote-sensing reflectance (
) from AISA measurements using radiative transfer simulations and constraints obtained from field spectral
measurements. Unlike previously published method, the AC approach removes the surface Fresnel reflection and accounts for aircraft altitude and nonzero near-infrared (NIR) reflectance through iteration over the pre-established look-up tables (LUTs) based on MODTRAN calculations. Simulations and comparison with concurrent
in situ
measurements show the feasibility of the approach in deriving surface
with acceptable uncertainties. The possibility of errors in the radiometric calibration of AISA is demonstrated, although a definitive assessment cannot be made due to lack of enough concurrent
in situ measurements. The need for noise reduction and the difficulty in carrying out a vicarious calibration are also discussed to help advance the design of future AISA missions.